an integrated model of buyer-seller relationships in the

i

An Integrated Model of Buyer-Seller

Relationships in the Australian Wine

Industry

Major thesis submitted in partial fulfilment of the

requirements for the degree of

Doctor of Philosophy in Sciences

Simon Alexander Somogyi

School of Agriculture, Food and Wine

University of Adelaide

Australia

February 2012

ii

Abstract

The study examined how communication elements and relational norms such as power

asymmetry influence relationship quality from the perspective of grape growers in their

relationships with wineries using an integrated model of the relationship between the

two actors (grape grower and winery) in the Australian wine industry.

First, a review of the literature identified a deficiency of research examining

communication between grape growers and wineries and the effect that power

asymmetry has on relationship quality. The literature review also identified that

relationship quality is measured both uni-dimensionally and multi-dimensionally.

Second, a qualitative exploratory study, involving in-depth interviews with grape

growers, examined how dimensionality of collaborative communication and power

asymmetry in the relationship (favouring the winery) influenced relationship quality.

Furthermore, the elements of collaborative communication were found to influence the

relationship quality, in particular the modality, formality, directionality and the non-

coercive abilities of communication. The exploratory study, combined with the

literature review, created a conceptual model based on a multidimensional measurement

of relationship quality and an alternative conceptual model based on a uni-dimensional

measurement.

Finally, the study involved a questionnaire administered to grape growers to test

quantitatively the conceptual models. The conceptual models were tested via Structural

Equation Modelling using Partial Least Squares Regression. The main results showed

that direct modes of communication (for example, face to face and direct email

communication) positively affected relationship quality, while non-direct modes (such

as seminars and newsletter) negatively affected relationship quality, and that the power

asymmetry led to decreased grape prices and lower relationship quality. The linkages in

the main conceptual model between satisfaction (an element of relationship quality),

and many of the relational dimensions, were insignificant. The reason for this was due

to the price per tonne that the grape growers received for their produce (grapes). The

estimation of the alternative model, based on a uni-dimensional estimation of

relationship quality, showed a greater fit of the data with less significant path

iii

estimations. Further analysis of the models showed a direct correlation between the

relationship quality and price of grape supplied, whereby the higher the price they

received, the higher the level of relationship quality they experienced.

The quantitative phase of the study also highlighted three clusters of respondents‟

relationships with wineries.

Firstly, there was an “unsustainable relationship”, whereby the respondents experienced

low levels of relationship quality, high power asymmetry favouring the winery, and a

very low price per tonne for their grapes. Respondents in this cluster were mainly

located in warm climate grape growing regions, and mainly dealt with large, publicly

owned wineries.

Secondly, an “OK relationship” was observed, whereby respondents experienced higher

levels of relationship quality and lower high power asymmetry favouring the winery

than the “unsustainable relationship” cluster. They received a higher price per tonne

than the “unsustainable relationship‟ cluster, were located in cool to warm climate

grape growing regions, and dealt with more small, privately owned wineries than the

“unsustainable relationship” cluster.

Thirdly, there was a “good relationship”, whereby respondents experienced the highest

level of relationship quality and the least amount of power asymmetry favouring the

winery, of the three clusters. This cluster also received the highest price per tonne of the

three clusters and was mostly located in cool climate wine growing regions. This cluster

dealt with more small, privately owned wineries than the other two clusters.

Wineries will need to take into consideration the results of this study, particularly the

dimensionality of communication and power asymmetry effects, when dealing with

grape growers.

iv

Statement of Declaration

I declare that this thesis does not incorporate without acknowledgement any material

previously submitted for the award of any other degree or diploma in any university or

other tertiary institution; and that to the best of my knowledge and belief, it does not

contain any materials previously published or written by another person, except where

due reference has been made in the text.

I give consent to this copy of my thesis, when deposited in the University Library,

being made available for loan and photocopying, subject to the provisions of the

Copyright Act 1968. I also give permission for the digital version of my thesis to be

made available on the web, via the University‟s digital research repository, the Library

catalogue, the Australasian Digital Theses Program (ADTP) and also through web

search engines, unless permission has been granted by the University to restrict access

for a period of time.

................................................................

Simon Alexander Somogyi

v

Acknowledgements

I would like to acknowledge the following people and parties for their assistance during

my study.

Firstly, I would like to acknowledge my supervisors, Dr Elton Li and Assoc Prof Johan

Bruwer who provided great advice and encouragement over the duration of the project.

It was their encouragement that aided me through the journey. I would also like to

thank Dr Amos Gyau who not only gave wise advice and comment but reinvigorated

my enthusiasm when things were not working as they should. I consider Amos not only

a colleague, but also a friend and confidant.

Secondly, this project would not have been possible without the assistance and advice

of many Australian wine industry stakeholders. In particular, I would like to thank

Lyndal Sterenberg of Morton Blacketer who not only shared her vast experience, but

also gave me access to respondents. I would also like to acknowledge Mark McKenzie

of Wine Grape Growers‟ Australia, Mike Stone formerly of the Murray Valley Wine

Growers‟ Association, Brian Simpson of the Riverina Wine Grapes Marketing Board,

Di Davidson and Sam Burton or Davidson Viticulture, Hamish Franks of Foster

Groups, and John Hahn and Elise Hayes of the Barossa Grape and Wine Association

who gave advice and access to respondents. Their assistance is greatly appreciated. I

would also like to thank the numerous regional grape growers‟ associations, too many

to list, who gave access to respondents. Their good humour and willing cooperation

toward me and the project was remarkable considering the harsh economic and social

issues facing their constituents. They cannot be thanked enough.

I would also like to acknowledge the anonymous grape growers who graciously gave

their time and effort assisting in the pilot phases of the study, including the

questionnaire design process. These individuals must be commended for their good

humour and patience when it appeared that I was bothering them. Thank you all very

much.

On a personal note, I would like to thank my mother and father, Lydia and Andrew, and

my sister Julia. They constantly encouraged me and were always there during the good

and bad times throughout the journey.

vi

I would also like show my immense appreciation of all the grape growers who

participated in this study. This project would not have existed without their

participation. I thank you from the bottom of my heart and hope you keep on fighting. I

would also like to thank Dr Vic Beasley who professionally edited this thesis.

And lastly I would also like to thank my fiancée, Justine, who not only performed the

task of proofreading this document, having to deal with my spelling and grammatical

foibles, but also provided me with support and encouragement during the bad times. I

cannot thank her enough for what she has given me.

vii

Table of Contents

Abstract ........................................................................................................................... ii

Statement of Declaration .............................................................................................. iv

Acknowledgements ......................................................................................................... v

Table of Contents .......................................................................................................... vii

Table of Tables ............................................................................................................. xiii

Table of Figures ............................................................................................................ xv

Chapter 1: Introduction ................................................................................................. 1

1.1 Chapter outline ................................................................................................... 1

1.2 History of the Australian wine industry ............................................................. 1

1.3 Current state of the Australian wine industry .................................................... 2

1.4 Research problem and objectives and thesis title .............................................. 5

1.5 Research Design and significance of the study .................................................... 7

1.6 Structure of the thesis ......................................................................................... 10

Chapter 2: Literature Review ..................................................................................... 11

2.1 Introduction ....................................................................................................... 11

2.2 Australian wine industry context for discussion of literature ............................ 11

2.3 Business to business interaction ........................................................................ 12

2.3.1 Industrial markets and inter-firm relational development .......................... 12

2.4 Business to Business Marketing ....................................................................... 14

2.4.1 B2B purchasing .......................................................................................... 14

2.4.2 Exchange relationships ............................................................................... 19

2.4.3 Relationship development and relationship marketing .............................. 22

2.4.4 Business to Business networks ...................................................................... 24

2.5 Relational norms ................................................................................................ 28

2.6 Communication ................................................................................................. 30

2.7 Relationship quality .......................................................................................... 32

viii

2.7.1 Trust as a dimension of relationship quality .............................................. 33

2.7.2 Satisfaction as a dimension of relationship quality .................................... 35

2.8 Power Asymmetry ............................................................................................. 36

2.9 Literature Discussion ......................................................................................... 38

2.10 Chapter conclusion ........................................................................................... 40

Chapter 3: Exploratory research methodology and results ..................................... 41

3.1 Chapter introduction .......................................................................................... 41

3.2 Exploratory research design ............................................................................... 41

3.3 Participant sample selection and interview format ............................................. 43

3.4 Structure of the interview format ....................................................................... 45

3.5 Research objectives ............................................................................................ 45

3.6 Audio transcription and data analysis technique ............................................... 46

3.7 Exploratory study results ................................................................................... 46

3.7.1 Research results related to uncovering the effect that collaborative

communication theory has on relationship quality ................................................. 47

3.8 Exploratory research findings and relevance to literature ................................. 53

3.8.1 Research results on communication modality and relevance to literature

and hypothesis development .................................................................................. 53

3.8.2 Research results on communication directionality and relevance to literature


3.8.3 Research results on non-coercive communication attempts and relevance to

literature and hypothesis development ................................................................... 54

3.8.4 Research results on communication formality and relevance to literature


3.8.5 Research results on power asymmetry and relevance to literature and

hypothesis development ......................................................................................... 55

3.8.6 Relationship quality and relevance to research results ................................ 56

3.9 Exploratory study research objectives overview .............................................. 56

3.10 Limitations of the exploratory study ............................................................... 57

3.11 Hypothesised model ......................................................................................... 57

ix

3.12 Alternative model .............................................................................................. 59

3.13 Chapter conclusion .......................................................................................... 62

Chapter 4: Descriptive and Causal Research Methodology ..................................... 63

4.1 Chapter outline .................................................................................................. 63

4.2 Quantitative research methodology design ........................................................ 63

4.3 Data collection method ...................................................................................... 64

4.3.1 Quantitative study sampling procedure and sample size ............................. 65

4.3.2 Administration of survey instrument ........................................................... 67

4.3.3 Questionnaire design .................................................................................. 69

4.3.4 Modification of questionnaire to online format .......................................... 70

4.3.5 Protection of questionnaire information against online fraud ...................... 72

4.3.6 Section 2: Scale items relating to research hypotheses .............................. 73

4.4 Data preparation and data analysis techniques .................................................. 78

4.4.1 Univariate Analysis ...................................................................................... 79

4.4.2 Multivariate Analysis ................................................................................... 79

4.5 Chapter summary ............................................................................................. 82

Chapter 5: Descriptive statistics of respondents and trading relationships ........... 83

5.1 Chapter outline .................................................................................................. 83

5.2 Section 1: Descriptive statistics of grower/winery relations. ............................ 84

5.2.1 Duration of relationship with winery .......................................................... 84

5.2.2 Volume of grapes supplied to winery ......................................................... 84

5.2.3 Value of grapes supplied to winery by respondents .................................... 85

5.2.4 Average price per tonne of grape supplied to winery ................................. 86

5.2.5 Other wineries supplied and the amount of grapes supplied to those

wineries. ................................................................................................................. 87

5.2.6 Business details of the winery that was supplied grapes ............................. 88

5.2.7 Summary of trading relations of grape grower respondents ...................... 92

5.3 Section 3: Descriptive statistics of respondents ................................................ 92

5.3.1 Size of the respondents‟ vineyards .............................................................. 93

x

5.3.2 Number of years respondents operating their viticultural business ............ 93

5.3.3 Number of people employed by respondents‟ businesses ........................... 94

5.3.4 Wine region location of respondents‟ businesses ........................................ 95

5.3.5 Technical viticultural qualifications of respondents .................................. 97

5.3.6 Summary of descriptive statistics of respondents ....................................... 98

5.4 Chapter Summary .............................................................................................. 98

Chapter 6: An integrated model of buyer-seller relationships in the Australia wine

industry .......................................................................................................................... 99

6.1 Chapter outline ................................................................................................... 99

6.2 Measurement model of constructs ..................................................................... 99

6.2.1 Evaluation of the outer model .................................................................. 100

6.2.2 Evaluation of the inner model ................................................................. 105

6.2.3 Results of the structural model .................................................................. 110

6.3 Consideration of structural model results ........................................................ 112

6.4 Alternative structural model estimation .......................................................... 115

6.5 Power, Satisfaction and Trust cluster analysis ................................................ 124

6.5.1 Cluster analysis methodology ................................................................... 125

6.5.2 Cluster 1: “Unsustainable Relationship” .................................................. 130

6.5.3 Cluster 2: “OK relationship” ..................................................................... 130

6.5.4 Cluster 3: “Good Relationship” ................................................................. 130

6.6 Chapter conclusion ........................................................................................... 132

Chapter 7: Discussion, conclusion and implications for further research ............ 133

7.1 Chapter outline ................................................................................................. 133

7.2 Summary of the research process ..................................................................... 133

7.3 Hypothesis discussion ...................................................................................... 135

7.3.1 H1: Direct modes of communication positively influence trust. ............. 135

7.3.2 H2: Direct modes of communication positively influence satisfaction ... 135

7.3.2.1 H1a: Direct modes of communication positively influence relationship

quality. .................................................................................................................. 136

xi

7.3.3 H3: Indirect modes of communication negatively influence trust ............ 136

7.3.4 H4: Indirect modes of communication negatively influence satisfaction. 136

7.3.4.1 H2a: Indirect modes of communication negatively influence relationship

quality. .................................................................................................................. 137

7.3.5 H5- Uni-directional communication from the winery positively influences

trust ....................................................................................................................... 137


satisfaction. ........................................................................................................... 138

7.3.6.1 H3a- Uni-directional communication from the winery positively

influences relationship quality. ............................................................................. 138

7.3.7 H7- Non-coercive communication attempts from the winery negatively

influence trust. ...................................................................................................... 138


influence satisfaction ............................................................................................ 139

7.3.8.1 H4a- Non-coercive communication attempts from the winery negatively

influence relationship quality. .............................................................................. 139

7.3.9 H9- Formality of communication from the winery negatively influences

trust ....................................................................................................................... 139


satisfaction. ........................................................................................................... 140

7.3.10.1 H5a- Formality of communication from the winery negatively

influences relationship quality. ............................................................................. 140

7.3.11 H11- Power asymmetry in the relationship, favouring the winery, is

decreasing growers trust in the winery. ................................................................ 141


decreasing growers‟ satisfaction with the winery. ............................................... 141

7.3.12.1 H6a- Power asymmetry in the relationship, favouring the winery is

decreasing grape growers perceptions of relationship quality. ............................ 141

7.4 Cluster analysis results discussion ................................................................... 142

7.4.1 “Unsustainable Relationship” cluster .......................................................... 142

7.4.2 “OK relationship” cluster .......................................................................... 143

xii

7.4.3 “Good Relationship” cluster ...................................................................... 144

7.4.4 Questionnaire item results discussion, by cluster ..................................... 144

7.4.5 Cluster results summary ............................................................................ 146

7.5 Research Question Summary ........................................................................... 146

7.5.1 Question 1: Which relational constructs constitute relationship quality? . 147

7.5.2 Question 2: Which elements of the grape grower/ winemaker relationship

affect grape growers‟ perceptions of relationship quality? .................................. 147

7.5.3 Question 3: Are there any commonalities between wine grape growers in

their perceptions of relationship quality? ............................................................. 148

7.6 Conclusion ....................................................................................................... 149

7.7 Study Limitations ............................................................................................ 150

7.8 Recommendations for further research ........................................................... 152

7.9 Study contribution ............................................................................................ 153

7.10 Study implications for the Australian wine industry ..................................... 155

Appendix 1: Questionnaire ........................................................................................ 156

Appendix 2: Cluster Analysis Results ...................................................................... 165

Appendix 3: IDI discussion questions ....................................................................... 175

Bibliography ................................................................................................................ 176

xiii

Table of Tables Table 2.1: List of relational norms ............................................................................... 29

Table 3.1: Location and size of grape grower participants‟ businesses ....................... 43

Table 3.2: Frequency of topic (code) discussion in in-depth interviews ...................... 47

Table 4.1: Grape grower associations and private organisations that provided access to

respondents ..................................................................................................................... 68

Table 4.2: Questionnaire scale times regarding the formality of communication ....... 75

Table 4.3: Questionnaire scale items regarding winery feedback ................................ 76

Table 4.4: Questionnaire scale items: non-coercive communication attempts ............ 76

Table 4.5: Questionnaire scale items regarding trust ................................................... 77

Table 4.6: Questionnaire scale items regarding satisfaction ........................................ 77

Table 4.7: Questionnaire scale items regarding power ................................................ 78

Table 4.8: Statistical criteria for model estimation via PLS ......................................... 82

Table 5.1: Years of contractual relationships between respondents and wineries ....... 84

Table: 5.2: Volume of grapes supplied to winery by grape grower respondents ......... 85

Table 5.3: Value of grapes supplied to winery by respondents .................................... 86

Table 5.4: Price per tonne of grapes supplied to the winery by respondents ............... 87

Table 5.5: Number of other wineries to which respondents supplied grapes ............... 88

Table 5.6: Percentage of grape production supplied to the other wineries .................. 88

Table 5.7: Ownership of the winery to which respondents supplied grapes ................ 89

Table 5.8: Size of the winery to which respondents supplied grapes ........................... 90

Table 5.9 Wine region winery was located in ............................................................. 90

Table 5.10: State wineries were located in ................................................................... 91

Table 5.11: Summary of the trading relationship of respondents and wineries ........... 92

Table 5.12: Size of respondents vineyards in acres ...................................................... 93

Table 5.13: Number of years respondents operation of business ................................. 94

Table 5.14: Number of people employed by respondents‟ businesses ......................... 94

Table 5.15: Wine region location of respondents viticultural businesses .................... 95

Table 5.16: State respondents were located in ............................................................. 97

Table 5.17: Viticultural qualification of respondents ................................................... 97

Table 5.18: Summary of descriptive statistics of respondents ..................................... 98

Table 6.1: Outer model evaluation of collaborative communication dimensions, trust,

satisfaction and power. ................................................................................................. 101

Table 6.2: Loadings and cross loadings of indicators and constructs ........................ 106

Table 6.3: Correlations of the latent variables and the AVE square roots ................. 109

xiv

Table 6.4: Results of the structural model .................................................................. 111

Table 6.4: Outer model evaluation of collaborative communication dimensions, trust,

satisfaction and power of alternative model. ................................................................ 116

Table 6.5: Loadings and cross loadings of indicators and constructs in the alternative

model ............................................................................................................................ 119

Table 6.6: Correlations of the latent variables and the AVE square roots ................. 122

Table 6.7: Results of the structural model for the alternative model ......................... 123

Table 6.8: Factor analysis and results of Trust, Satisfaction and Power dimensions . 125

Table 6.9: Questionnaire item mean, median and standard deviation score by cluster

...................................................................................................................................... 127

Table 6.10 Summary of cluster analysis results ........................................................ 131

xv

Table of Figures Figure 3.1: Conceptual model of grape grower perceptions of relationship quality in

the Australian wine industry ........................................................................................... 58

Figure 3.2 Alternative model based on uni-dimensional definition of relationship

quality and grape grower perception of collaborative communication and power

asymmetry ...................................................................................................................... 61

Figure 4.1: Questionnaire scale items regarding the mode of communication ............ 73

Figure 6.1 Conceptual model of grape grower perceptions of relationship quality in the

Australian wine industry .............................................................................................. 110

Figure 6.2 A graphical representation of the of main structural equation model results

...................................................................................................................................... 112

Figure 6.3 Alternative model based on uni-dimensional estimation of relationship

quality ........................................................................................................................... 115

Figure 6.4 Graphical representation of the alternative structural model results ........ 124

1

Chapter 1: Introduction

1.1 Chapter outline In this chapter the Australian wine industry, in particular the current state of grape

grower and winery relationships is discussed. The objectives of this study, including the

design of the research and research problems, are presented and a justification for using

the Australian wine industry as a context to test the research problems and objectives is

discussed. The chapter concludes with a summary of the composition of the thesis.

1.2 History of the Australian wine industry Wine in Australia has existed since European settlement of the country. Grape vines

were brought to Australia from Brazil by Captain Arthur Philip in the late 1700s and the

vines were planted around what is now Sydney and flourished there (Wine Australia,

2009). Grapes were then planted in areas such as New South Wales, Tasmania and

Victoria with mixed success; production mainly satisfied export demand, generally

from England. The discovery of gold in eastern Australia in the mid-1800s dramatically

increased the consumption of wine and, as a result, vines were planted widely (Culture

Portal, 2009).

The time period from the early 20th to mid-20th century saw two world wars, and the

resettlement of soldiers from those conflicts contributed to the rapid increase of wine

consumption, mainly driven by the consumption of fortified wines. The consumption of

fortified wines was derived from a cultural link with the United Kingdom. However, by

the 1960s and 1970s an influx of European migrants resulted in changing consumption

patterns. Table wine styles (for example red, white and sparkling wines) began to be

consumed and this was also aided by a more cosmopolitan view of life by Anglo-

Australians (Walsh, 1979). Young Anglo-Australians started to travel to European

countries and this spawned an appreciation of Mediterranean cuisine and associated

wine consumption patterns.

The Australian wine industry became hampered by an oversupply of grapes in the mid-

1980s, and 2500 acres of vines were removed; however, an export led boom in demand

2

for Australian wine (led by the UK and USA markets) in the late 1990s saw an

undersupply of grapes and consequently such removals in the 1980s were regretted

(Clancy pers comm. April 2009). A boom in production and export sales in the early

21st century created great wealth and prosperity for the industry, mainly led by

favourable taste preference of consumers in export markets and favourable exchange

rates (Stanford, 2007).

The preceding discussion has shown that the wine industry has gone through periods of

economic prosperity, specifically five periods (Osmond & Anderson, 1998). These

“booms” in economic prosperity are summarised chronologically as follows:

the first boom in the mid-1850s due to discovery of gold in Victoria and New

South Wales and aided by a trebling of the Australian population;

the second boom in the late 1880s due to domestic increases in consumption and

export growth, particularly to the British market;

the third boom in the mid-1920s led by the export of fortified wine to the United

Kingdom and aided by land development subsidies for grape production granted

by the federal government;

the fourth boom in the 1960s attributed to changing domestic consumer tastes

from fortified wine consumption to table wine consumption aided by a more

cosmopolitan view on life which resulted from Australians travelling overseas

and the migration of European migrants; and

the fifth boom in the late 1980s due to strong export demand from Europe and

North America; the North American consumption of Australian wine was aided

by favourable exchange rates, successful branding strategies and a focus on the

consumption of wine for health reasons.

(Osmond & Anderson, 1998)

1.3 Current state of the Australian wine industry The Australian wine industry has expanded markedly throughout the 20th century in

terms of the area under vine and the production of grapes. Winetitles (2010) states that

in 2009 the total area under vine was 162,550 hectares with a grape crush of 1.71

million tonnes. This is a decrease of approximately 7% from the 2008 vintage.

Winetitles (2010) lists 2420 companies that sell wine commercially, of which two

companies, Foster‟s Group and Constellation Wine Australia, account for

3

approximately 45% of all branded wine sales with the top 20 companies accounting for

90% of total sales. These figures indicate that the remaining 2400 producers compete

for 10% of the total sales of branded, bottled wine.

Evidently the Australian wine industry‟s sales have consolidated, with the largest wine

producers dominating sales. The increase in wine production volume has coincided with

a less than equal increase in sales, with a current wine inventory level of 2.1 billion

litres in 2006. The current stock to sales ratio of approximately 2:1 is unfavourable.

With a current stock inventory of 1.9 billion litres and estimates stating that a ratio of

1.7:1 is required (AWBC, 2007; ABS, 2009a), the Australian wine industry is

producing an excessive amount of grapes and an oversupply exists.

Approximately 60% of the wine produced in Australia is exported and consequently

export markets are of critical importance to the industry‟s well-being (Wine Australia,

2009). However, as previously mentioned, the effect of decreasing wine export volume

is compounded by the decreasing value per litre of exported wine and, therefore, has

resulted in a lower financial return for Australian wine producers. As such, in the year

to December 2009 the value per litre of exported wine decreased by 15% (Winetitles,

2010). Therefore wineries have experienced decreasing earnings, with the majority of

Australian wineries (under $20 million in revenue) receiving losses before tax in the

year to 2009 (Deloitte, 2009).

Such financial pressures experienced by the wineries are being passed onto grape

growers, who are in turn experiencing financial hardship. Part of the industry‟s hardship

has also been attributed to issues related to climate change. Frost, and particularly

drought, have caused a reduction in yields resulting in less income for the grape grower;

however, the lack of water has required grape growers to purchase water at ever

increasing prices, which has placed them under further cost pressures (Hayman et al.,

2007; Stone, pers comm., February 2010).

There have also been other issues relating to cost pressures affecting the 4500- 6500

grape growers in Australia and much of this is attributed to growers receiving lower

prices for their grapes (ABS, 2009b; McKenzie, pers comm., May 2009). While

statistics show that grape prices increased in the 2007 vintage (up to a 40% increase in

warm climate areas) with the reduced yield (due to frost and drought) increasing prices,

when viewed historically there has been an average decrease in price of 50% from the

2001 vintage (ABARE, 2009; McKenzie, pers comm., May 2009). This price reduction

4

is in contrast to the past; grape prices increased by 73% from 1987 to 1997 (Osmond &

Anderson, 1998).

Grape growers are currently experiencing poverty and this can be viewed against a

history which shows that grape growers have received lower prices for their grapes in

the past, particularly in the mid-1980s where a glut of grapes resulted in markedly

lower prices and the destruction of vines (IAC, 1995; Clancy pers comm., April 2009).

The current oversupply of grapes is also affecting wineries; to alleviate financial

pressures, some wineries have been cancelling, and not renewing, grape supply

contracts. As a result of the actions of certain wineries during this period, many

relationships between them and grape growers have become adversarial and have

resulted in inefficiencies which may harm the Australian wine industry (Speedy, 2006).

The adversarial nature of grape grower and winery relationship is not confined to the

Australian wine industry, nor to current times. In 1910-1911, riots occurred in

Champagne, France, due to grape growers‟ perceptions that the prices they were

receiving for their grapes were unfairly low (Phillips, 2000). The cause of the low

prices was attributed to a power asymmetry wielded by the Champagne houses, as a

result of there being a small number of houses and a large number of growers in the

region. This is also evident in current times where a power asymmetry favouring

Champagne houses is resulting in lower grape prices for their grape growers (Charters

& Menival, 2010). Furthermore, in recent times, particularly in Europe, there has been

conflict involving grape growers, wineries and retailers. For example, grape growers in

the south west of France have highjacked trucks, vandalised wine retail outlets, and

destroyed wine as they perceived that the low prices they received for their grapes was

a result of power wielded by wineries and the importation of cheap wine by wine

retailers (IAC, 1995; Quinn, 2008). The conflict is also evident in other European

countries such as Hungary and Kosovo where local grape growers, unable to find

buyers for their grapes, protested and took violent action against their respective

governments in order to gain better price terms (Farmers protest in Kosovo town turns

violent, 2010).

Therefore, the relationships between grape growers and wineries, not only in Australia,

have resulted in conflicts and potential inefficiencies. In Australia, the inefficiencies

and their effects could be compounded by strategic changes to wine industry policy by

the peak industry bodies. The industry is attempting to reposition itself to focus on the

5

production of quality wines (as opposed to volume production) and emphasising

regional branding (Hobley & Batt, 2005; Deloitte & WFA, 2006) hoping that a focus on

quality production will allow the wine industry to gain a strategic competitive

advantage (Chong, 2007).

Collaboration and long term relationships are crucial to the development of wine

products which meet appropriate quality specifications (CIE, 2004). Quality parameters,

while set by the purchasing winery early in the growing season, are controlled by the

grower with such elements as pH level, pest and disease control, grape sugar content

and berry size contributing most to wine quality (Spawton & Walters, 2003; Clancy,

2005). To obtain grapes of a certain quality parameter, the winery must engage in

relational activities that engender a higher level of relational quality for the grape

grower. Higher levels of relationship quality provide greater loyalty from the grape

grower to the winery and have the added effect of continued financial returns for the

grower. In light of the oversupply of grape and wine affecting the industry, and the

resultant lower grape price returns for grape growers, it is of interest to observe the

grape growers‟ perceptions of relationship quality.

Numerous wine industry and government publications have highlighted the need for

better relationships between grape grower and wineries. For example, the former

Industry Assistance Commission (now referred to as the Productivity Commission) in a

report to the federal government advocated improved relationships and better supply

chain coordination between grape growers and wineries to increase grape quality and

higher levels of trust between the two actors (IAC, 1995). Spawton & Walters (2003)

claim that better coordination of grape growers is required and that elements of these

relationships, such as communication, need to be improved. Chong (2007) advocates

that relationships between the two partners need to be developed further, particularly in

communication between the actors, and this notion is affirmed by Brown (2008) who

further comments that good communication is needed to maintain and enhance

relationships between the two.

1.4 Research problem and objectives and thesis title The rationale behind the research problem for this study was to conceptualise and

measure the relationship quality and its effect on other relational variables from a wine

grape seller‟s perspective in the Australian wine industry. In doing so, an integrated

6

model of the buyer-seller relationship in the Australian wine industry was created and

this notion is reflected in the title of this thesis.

Three research questions were devised for this study:

1. which relational constructs constitute relationship quality?

2. which elements of the grape grower/ winemaker relationship affect grape

growers‟ perception of relationship quality?

3. are there any commonalities among wine grape growers in their perceptions of

relationship quality?

Quality in a wine product is based on the quality of the grapes produced, with

approximately 60% of the work required to make a high quality wine derived from the

grapes (Scales, Croser and Freebairn, 1995). In order to obtain the grapes of a desired

quality, the winery must liaise with a grower during the growing season (from

approximately August to April in the Southern Hemisphere) and therefore much

emphasis is placed on the grower-derived inputs. Thus, it is of interest for the winery to

liaise appropriately with the grower.

This notion of grower-derived wine quality is of particular importance to the Australian

wine industry due to changes in the marketing and promotion of Australian wine to

emphasise quality and regionality (Henry, 2009). As a consequence, the suppliers of

wine grapes in this industry are becoming increasingly important in the supply chain,

and their needs and wants must be uncovered and satisfied. This study has attempted to

achieve this.

From an economic perspective, the wine industry is of great importance to the

Australian economy, further justifying its selection as a research subject. The wine

industry accounted for approximately $2.6 billion of domestic and export sales in 2009

(Winetitles, 2010). The number of wineries in Australia has also increased by

approximately 4.3% from 2008 to 2009, with the number of wineries having more than

doubled since 2000 (Winetitles, 2010). The wine industry directly employs 28,000

people and indirectly employs others in areas such as hospitality, retail and wholesaling

(DFAT, 2009). Currently there is no solid information available regarding the increase

or decrease in the number of grape growers; however, approximately 4500 to 6500

growers exist in the industry (ABS, 2009b; McKenzie pers comm., May 2009). Further

highlighting the industry‟s economic importance is the fact that it has a production

presence in all states and territories in Australia except for the Northern Territory

7

(Winetitles, 2010) and the industry has production entities (wineries and grape growers)

that are small, medium and large in size, both publicly and privately owned (Winetitles,

2010). It is evident that the Australian wine industry is of vital importance to the

Australian economy, particularly to its rural sector, and is therefore a significant area of

research.

Recent times have seen an upheaval in the Australian wine industry. Apart from the

issues previously discussed, grape growers have experienced decreasing grape prices

and as their future importance in the wine industry supply chain is being cemented by

marketing initiatives emphasising grower-derived inputs (e.g. quality and regionality), it

is of interest to investigate their perceptions of the relationship between the two actors;

this was an objective of this study (ABARE, 2009; Henry, 2009).

Furthermore, the marketing initiatives place greater importance on the grower in the

supply chain and, therefore, examining the relationship that wine producers have with

growers will allow wineries to tailor their grower liaison efforts to best satisfy grower

needs. The quality of the relationship which growers have with wineries is of

importance to wineries as the increasing importance of growers in the supply chain will

shift the emphasis to satisfying grower needs. As a result, the wine industry provides a

fertile area of research in any attempt to uncover supplier related perceptions of

relationship quality.

1.5 Research Design and significance of the study From an ontological perspective, the study involved interviewing and surveying

Australian wine grape growers about their perceptions of communication and power

asymmetry in the relationship they have with wineries. The study was designed

employing a two-step process, often referred to as a multi-method or mixed method

approach, whereby qualitative and quantitative methods were integrated into the study

(Carson & Coviello, 1996). From an epistemological perspective, the study utilised a

scientific, validity approach which was used to develop and test hypotheses in the

quantitative phase of the study (Wacquant, 1992; Cohen & Maldonado, 2007).

However, the qualitative phase of the study employed an interpretive, constructivist

perspective as this phase of the study explored concepts of relationships and required

interpretation by the researcher (Gall et al., 2003). The literature discusses three types

of research, namely exploratory, descriptive and causal (Kinnear et al. 1993). This study

8

contained these three types of research; this approach is common in agribusiness PhD

studies (see Storer, 2005; Hobley, 2007).

Firstly, an exploratory phase was deemed important as it allowed for the development

of a clearer understanding of the phenomena to be studied (Zikmund, 2003). This initial

stage of the research was deemed appropriate as the relationship between grape growers

and wineries, particularly related to the elements of communication and relationship

quality, had not been extensively investigated in the past. As these factors are complex,

an exploration was vital in order to gain an insight into their interactions (Zikmund,

2003). The exploratory research stage utilised qualitative research methods, namely in-

depth interviews (Ticehurst & Veal, 1999).

Descriptive research was also used in order to gain an understanding of the phenomena

such as frequencies and means, particularly the descriptive statistics of the respondents

and their trading relationships with wineries (see Chapter 5). The descriptive research

phase allowed for the validation of the sample against the sample frame, and the data

was captured via the use of a questionnaire.

While descriptive research has the purpose of describing phenomena and predicting

linkages between variables, explanatory, causal research was required to verify

assumptions that were made in the exploratory phases, such as the hypotheses that were

formulated, and was performed using structural equation modelling (SEM) utilising

partial least squares regression. The purpose of using SEM was that is has the ability to

test entire models (i.e. the conceptual models devised in the exploratory phase)

(Baumgartner and Homburg, 1996; Steenkamp & Baumgartner, 2000). The model

tested in the causal stage of the research involved various constructs (i.e. collaborative

communication element, power and relationship quality) which were operationalised in

a questionnaire using multiple questionnaire items derived from previous studies (Hair

et al, 2006).

The causal stage of the research process was used based on the understanding that

empirical research is required to understand and to extend business to business (B2B)

marketing theory (Medlin, 2001; Donaldson & O‟Toole, 2000; Plewa, 2005).

Furthermore, a large number of studies that utilise the grape grower and winery

relationship context, or the wine industry as a unit of analysis, are qualitative or

exploratory in nature rather than empirical or mixed method (qualitative and

quantitative combined) studies (see Hall, 2004; Benson-Rea, 2005; Rampersad, 2008).

9

After completion of the causal stage of the study, an exploratory phase was again

employed (namely cluster analysis) to uncover the nature of relationships between

grape growers and wineries and to categorise the relationships based on relationship

dimensions (Everitt, 1996; Janssens et al, 2008).

In summary, the study contained both a constructivist and positivistic epistemological

approach due to the three research methods employed: exploratory, descriptive and

causal. Firstly, an exploratory, qualitative phase allowed for a conceptual understanding

of the constructs investigated, and the production of a conceptual model and was

constructivist in nature as it was based on viewing and interpreting the grape growers‟

perspectives but not trying to measure them (Guba & Lincoln, 2005). The second phase

of the study was descriptive and employed quantitative methods whereby descriptive

statistics were obtained, mainly to validate the sample. The third phase of the study was

quantitatively causal whereby the conceptual model developed in the exploratory phase

was tested. Finally, an exploratory quantitative method was employed, via the use of

cluster analysis to uncover the nature of the relationships between grape growers and

wineries. As such, the final three stages of the research employed a positivistic,

epistemological paradigm due to the scientific nature of the data analysis that employed

the testing of hypotheses and the categorisation of data based on clusters (Babbie,

2004).

To develop instruments of measurement, such as the questionnaire, wine industry

experts, such as peak body leaders, viticultural consultants, winemakers and wine

industry commentators, helped in their development and validation. This was

particularly the case in the development of the questionnaire used in the descriptive and

causal stages of the study.

This study differs from previous studies that explored a similar context (wine industry)

as it uses a mixed-method approach, unlike studies that are qualitative in nature (see

Benson- Rea, 2005; Rampersad, 2008). This study employs a similar method and

context as that used by Hobley (2007) in that the relationship between grape growers

and wineries is explored from a B2B and a relationship marketing perspective;

however, this study extends Hobley‟s (2007) work by focussing on a particular element

of that study, namely communication elements as proposed by Mohr & Nevin (1990)

and Mohr et al. (1996) in their theory of collaborative communication. Furthermore,

this study is different from other studies that have investigated communication elements

between agribusiness buyers and suppliers (of which grape grower and winery

10

relationships are examples); for example, in Mohr & Nevin (1990) and Mohr et al,

(1996), collaborative communication elements are empirically tested as opposed to

using an inter-organisational information management system (IOIMS) which differs in

its perspective of communication.

1.6 Structure of the thesis This thesis is structured as follows:

Chapter 2 discusses the theory relating to buyer-seller relationships and business to

business interaction, particularly related to the dimensionality of relationship quality

and the relational norms that affect relationship quality.

Chapter 3 outlines the methodology of the data collection in the exploratory stage of the

study and presents its results. The chapter also presents the conceptual models that are

tested by later stages of the study.

Chapter 4 discusses the methods used in the descriptive and causal stages of the study

such as uni-variate (descriptive statistics) and multivariate statistical methods (structural

equation model utilising partial least squares regression).

Chapter 5 outlines the quantitative results of the descriptive stage of the study, mainly

concerning the trading relationships and business details of the respondents of the

study.

Chapter 6 discusses the results of the causal stage of the study, and presents the results

related to the conceptual models. It also identifies commonalities between the grape

growers in terms of relationship quality via cluster analysis.

Chapter 7 summarises the main findings of the study and provides a conclusion and

areas for further research.

11

Chapter 2: Literature Review

2.1 Introduction This chapter outlines the academic literature regarding the research problems and

objectives detailed in Chapter 1. The chapter commences with a brief summary of

Chapter 1, followed by a discussion of the literature regarding business to business

interactions, relational norms and relationship quality. The chapter concludes with a

summary of the discussion and a section introducing the following chapter.

2.2 Australian wine industry context for discussion of literature The Australian wine industry is currently undergoing a period of economic hardship.

Due to issues such as production oversupply, maturing markets, unfavourable exchange

rates in export markets and international retail consolidation, many wineries are

experiencing economic losses (Henry, 2009; Deloitte, 2009). The financial losses

experienced by the wineries are being passed onto grape growers through the lowering

of grape prices and the cancelling of contracts (Hobley & Batt, 2005; ABARE, 2009).

However, the wine industry is establishing a marketing strategy designed to mitigate the

negative economic effects which aims to bring prosperity to the industry. The strategy

aims to produce and promote quality wine and regionality in wine products. Both of

these dimensions are grape grower derived; therefore, the inefficient relationships that

exist will need to be rectified, and information regarding the grape grower perspective

of the relationship will require investigation. As grape growers will need to be engaged

in relationships in order to accomplish the strategic marketing objectives, information

will have to be obtained with respect to the grape growers‟ perception of the relational

dimensions such as relationship quality.

As growers have received lower prices for their produce (grapes) in these harsh

economic times, it is of interest to observe how grape prices affect relationship quality.

The issue of improving relationships between grape growers has been highlighted in

12

academic and wine industry trade literature, particularly in relation to elements of the

relationships, such as communication, between the two actors (IAC, 1995; Spawton &

Walters (2003; Chong 2007; Brown 2008)

Firstly, let us consider of the generic context of the research, namely business to

business interactions.

2.3 Business to business interaction The focus of this study is the interaction between grape growers and wineries and as

such, the general context of this study is business to business (B2B) interaction. There

are aspects of the interaction which can be discussed and the differences between B2B

and business to consumer (B2C) interactions which can be observed. The comparison

between the two is important as it gives a perspective between the two fields of study in

marketing. The main areas of the B2B interaction that will be discussed in this section

of the chapter are B2B purchasing, and relationship marketing in B2B markets. Firstly,

the differences between consumer (B2C) and industrial (B2B) markets will be

discussed.

2.3.1 Industrial markets and inter-firm relational development

Purchasing occurs in both business to business (B2B) (often called industrial markets)

and business to consumer markets (B2C) (often referred to as consumer markets).

However, there are many differences between the two. For example, in B2B markets,

organisations acquire goods and services that are resold to other industrial markets

(such as private businesses, governments or institutional markets such as schools and

hospitals) and in B2C markets the goods are sold for personal consumption by

consumers (Kotler et al., 2010). However, B2B and B2C markets do not work in

isolation. B2B markets create products that are ultimately used in B2C markets, with

the wine industry providing a clear example. Wine grapes, a B2B product as grapes are

made by a business (grape growers) and sold to a business (a winery), are transformed

into wine which is then sold in B2C markets to consumers. The demand by the

consumer will shape the overall nature of the product with firms striving to produce

products that are demanded by consumers (Hutt & Speh, 2010). The consumer demand

characteristic will be observed by the B2B actors, and therefore the nature of the

13

product produced in the B2B phase will be modified to meet the needs of the end

consumer.

While the purchasing decision process has been briefly discussed in this section, the

next section will involve a greater discussion of industrial (B2B) purchasing and‟ in the

first instance, interfirm relationship development.

Actors in B2B interaction develop relationships, and these relationships develop over

time; the development has been shown to occur in various phases. Wilson (1995)

discusses a relationship development framework similar to Dwyer et al. (1987). In the

first phase, “partner selection”, Wilson (1995) posits a more active firm pair than that of

the “awareness” phase of Dwyer et al. (1987), whereby the actors are already

conducting business with each other and a deeper relationship is sought by one or both

actors (Morris, 2005). The second phase, “defining purpose”, involves creating a set of

activities that are expected by each partner and is characterised by a higher level of

communication. The third phase, “setting relationship boundaries”, evolves by a

process that may not possess a legal or explicit nature. The fourth phase, “creating

relationship value”, involves obtaining benefits from the partnership that would have

been unattainable by each firm independently. It is in this phase that “relationship-

specific investments” assume a prominent role, with these assets being similar to

Thibaut & Kelley‟s (1953) relational norm theory in that cooperation and commitment

are both active in this phase. In the final phase, “relationship maintenance”, relational

elements such as trust and satisfaction become fixed and are omnipresent in the

relationship.

Therefore, in regard to B2B relationship development, incorporating social exchange

theory, there is a development phase where boundaries and duties are set and if the

expectations are met, the relationship grows (Thibaut & Kelley, 1953). As the

relationship develops further and commitment becomes greater between the two actors,

relational norms (such as trust and satisfaction) are engendered. Overall, the discussion

of relational development has one common element: the development of the

relationship requires that both members consider the exchange worthwhile for

commitment to the relationship to occur.

While B2B interaction, such as relational development, purchasing and relationship

marketing, have been discussed, a further investigation into B2B marketing, the context

of this study, is required and is the focus of the next section of the chapter.

14

2.4 Business to Business Marketing The focus of study into business to business (B2B) marketing has shifted over time.

Much of this change is due to the dynamic nature of firms and the fact that firms are

increasingly understanding the importance of buyer supplier management, as there is an

understanding that in order to create products and services to sell to buyers, firms must

manage their inter-firm relationships (Ulaga, 2001). This area of marketing has seen a

shift from a focusing on the exchange between firms, to an emphasis on relationships

and a focus on inter-firm networks. The discussion in this section of the chapter will

focus on these three areas. However, the main function of B2B marketing is the

purchasing of goods and as such the next section will discuss this concept.

2.4.1 B2B purchasing

As briefly discussed earlier, B2B purchasing of products differs greatly from B2C

purchasing. As this study is focused on B2B interactions, a more detailed discussion of

B2B purchasing will be undertaken.

The literature discusses industrial purchasing from numerous viewpoints. The main

perspective includes those of a function (Barnhill & Lawson, 1980; Anderson et al.,

1994; Trent & Monczka, 1998; van Weele, 2000), as a process (Robinson et al., 1967;

Ozanne & Churchill, 1971; Webster & Wind, 1972; Kelly, 1974; Bradley, 1977;

Barnhill & Lawson, 1980), and as a supply or value chain (OK Porter, 1985; Hines,

1993; Hines et.al, 2000; van Weele, 1994).

The role of purchasing as a function in a B2B context is to procure supplies (Lysons &

Gillingham, 2003). The term “function” is derived from the notion that many functions

within a business are coordinated to purchase a product. In relation to this, Barnhill &

Lawson (1980) discuss the operations function in purchasing supplies for a business as

revolving around the coordination of activities within a business towards purchasing,

and if this is done satisfactorily, then the business will excel during the exchange of

products. Barnhill & Lawson (1980) stress that the exchange is complex and involves

activities such as production, finance, distribution and promotion, and that each of these

activities is the responsibility of a separate division within a business that must

coordinate with the other elements in order for the purchasing function to be successful

and at lowest cost. Also, in relation to purchasing as a function, Leenders & Fearne

15

(1997) and Duffy (1999) discuss purchasing as involving various elements such as the

flow of materials and supplies, the organisation of inventory, the development of

supplier relationships, and the notion that the purchasing function should strive to

achieve the maximum gain at the lowest cost, which in turn gains the business a

competitive advantage. This concept is highlighted by Trent & Monczka (1998) who

discuss the procuring of resources for a business as involving functional groups within

the business that work to acquire products, and to strive to reduce transaction costs,

improve product quality, reduce lead times and use better technology in order to

ultimately gain greater customer satisfaction. The concept of purchasing as a process is

highlighted by Anderson et al. (1994) who comment that not only do firms strive to

maintain excellence in the functions involved in purchasing within the company, such

as the activities highlighted by Barnhill & Lawson (1980); they also comment that

purchasing also involves various networks outside of the business. Two firms

purchasing in a dyadic relationship are not only connected to each other via the

purchasing of goods, but also by the relationship with secondary, ancillary suppliers

who work with both the buyer and supplier to aid the purchasing process. This concept

is also discussed by Trent & Monczka (1998) who comment that purchasing is

increasingly becoming network oriented with suppliers, buyers and third party

providers linking together increasingly through electronic means to purchase goods.

Trent & Monczka (1998) stress that due to the complex nature of modern purchasing,

involving various actors, in order to improve the function of purchasing the purchasing

manager needs to continually monitor and appraise the various actors to improve the

functions and therefore, gain a competitive advantage.

The previous discussion has alluded to the fact that the role of purchasing as a process

within a firm is highly complex. This notion is further pointed out by Barnhill &

Lawson (1980) who comment that purchasing acts like a process in that a two way

action occurs where a flow of money and value is exchanged for a good, service or item

of value. Other early works in this area further discuss the specific processes involved.

For example, Ozanne & Churchill (1971) discuss the concept of the Industrial Adoption

Process which leads to the purchase of industrial products. This process involves

various elements such as:

(i) factors that activate the purchasing process, such as equipment capacities,

obsolescence and labour shortages;

16

(ii) Purchasing Directing factors which are factors that purchasing decisions are

based on such as lead times quoted by suppliers, product attributes and past

experiences;

(iii) duration of the buying process, such as the length of time from problem

awareness to purchase;

(iv) alternative evaluations of supplier products and cost benefit analyses; and

(v) the use of information to make decisions about which product to purchase.

The early works in industrial purchasing heavily focus on the specific processes

involved, such as those commented on by Ozanne & Churchill (1971) shown above.

Kelly (1974) further discusses the process, particularly the decision making process

involved, and relates this closely to that which is performed in consumer decision-

making behavioural processes (Schiffman et al. 2001). This process involves the

recognition of a need, a search for alternatives, an alternative evaluation and a decision

on a product; however, Kelly (1974) states that a difference occurs in that the approval

process for purchasing is far more involved as various people in the buying centre are

required to authorise the transaction as opposed to an individual making a decision

(Rosenboom, 2004).

Webster & Wind (1972) comment further in relation to an organisation‟s decision

making process in that the purchasing of goods and services is based around four

elements: the environment in which the process is occurring, the abilities of the

organisation in terms of its technology and management structure, the buying centre of

the organization, including its structure and leadership style, and the individual

participants in the decision making process. These elements combine to affect the

decision making process, particularly the nature of the process including its duration

and complexity. This process is more developed by Kelly (1974) who likened the

process to consumer purchasing behaviour and did not highlight the specific differences

that occur in industrial purchasing.

Bradley (1977) further developed the notion of industrial purchasing, stating that, as

opposed to consumer purchasing, numerous people are involved in the process, and that

the transaction involves issues such as delivery terms and after sales services such as

technical support which may occur in consumer markets but is less of a concern.

Bradley (1974) also mentions that the type of product purchased will influence the

purchasing process such that a spectrum exists from routinely purchased products that

require little alternative evaluation to buying centre considerations of a capital product,

17

such as plant equipment and buildings that require great scrutiny and effort in the

decision-making and purchasing process. Bradley (1974) also discusses the purchasing

process and echoes the works of Ozanne & Churchill (1971), Webster & Wind (1972)

and Kelly (1974) when stating that when a purchasing need is felt by the company, a

shortlist of suppliers and products is made, contracts are awarded and a product is

purchased.

Much of the work focusing on purchasing as an industrial process is prescriptive in that

it talks specifically about the individual processes involved. Later work focusing on

industrial purchasing concentrates more on the functions involved. In summary,

purchasing by firms is a strategic process involving various units of a company and

decision making processes. These processes involve consideration as to the function

and profitability that a product will offer and the notion that a relationship between the

firms involved is complex and requires effort to establish and maintain.

The discussion has also highlighted the fact that industrial purchasing processes involve

many people within a company, generally described as a buying centre or team. This

team is highly complex and skilled in tasks required in the purchasing process, such as

alternative evaluation, negotiation and the procurement of the product. It can be

surmised that the abilities and talents of this team gains the company a strategic

competitive advantage (Rosenboom, 2004). From a wine industry perspective, many of

the firms, whether they be grape grower or winery, are small in size; many are

considered SMEs (Winetitles, 2010) and as such, the decision making process may only

be made by one person; for example, the owner of the business may make the decision.

In this case, the grape grower business or winery owner may be the only decision maker

and therefore the process of purchasing may differ in complexity from those discussed

in the literature, which tend to involve large corporations.

The notion of purchasing involvement in a supply chain can be discussed from a value

or supply chain perspective. In Porter (1985), the value chain perspective is discussed

as having many activities such as human resources involvement, and technology and

facilities supported by activities such as logistical functions (both outbound and

inbound) that ultimately result in product acquisition and value gained by the customer.

Effectively, Porter‟s (1985) premise is that material management (which includes

purchasing of materials for manufacture reasons) adds value and that if managed

appropriately will gain customer satisfaction. Hines (1993) adds to Porter‟s value chain

system by discussing an Integrated Materials Value Pipeline which shows numerous

18

pipelines of activities that exist in a supplier‟s network that aid the procurement of

product. Hines (1993, pg. 13) is clear in pointing out that the concept of value raised by

Porter (1985), namely that the “...value built into a company‟s products is the result of

activities required to design, produce, market, deliver and support that product” and that

these activities are based on the human capabilities of the company. Hines (1993) adds

that the problem with Porter‟s (1985) value chain model is that is focuses too much on a

firm‟s profit and not enough on customer satisfaction, and that the Porter (1985) model

does not fully show the interconnectedness of the firm‟s value chain, such as the

interconnectedness of human resources functions, materials and engineering research

development and marketing that are used to create values which Hines (1993) proposes

in the Integrated Materials Pipeline. The premise is that these sections of the firm are

driven by the needs of the consumer. Effectively Hines (1993) discusses the value chain

perspective from the consumer and then “up” the chain, whereby the function of the

firm, including the purchasing process, are fashioned to gain the maximum level of

customer satisfaction. Therefore, Porter‟s (1985) and Hines‟ (1993) perspectives of

B2B purchasing both put forward the notion that a “chain” or process, starting from

design and raw product, and ending at the consumer, will involve some sort of

purchasing and that this purchasing, between firms, is important to the ultimate success

of the firms.

Furthermore, Porter‟s value chain model (Porter, 1985; Porter 1990) has been discussed

and elaborated on in a wine industry context (Spawton & Walters, 2003). Spawton &

Walters (2003) discuss wine as a valuable product and discuss the way in which the

wine supply chain can be coordinated to create value which will give wine consumers

satisfaction. The processes in the Porter value chain, adapted by Spawton & Walters

(2003), include grape production facets such as the coordination of grape growing with

wine making parameters such that grapes of a specific quality are obtained in order to

create a wine that gives consumer value. Spawton & Walters (2003) note that relational

norms such as communication between the two actors will aid in the creation of

consumer value, which should be a basis for a sustainable competitive advantage of the

Australian wine industry.

The purchasing literature, as discussed above, has concentrated on three perspectives of

purchasing. Early literature discusses purchasing mainly from a process perspective that

appears very logistical in nature. Later literature discusses the intricacies of the

functions performed in purchasing, including the idea of customer satisfaction being an

19

important factors in purchasing, as proposed by value chain literature. The recent

literature tends to discuss purchasing more from a network perspective, whereby

individual firms have various networks of suppliers that aid and facilitate the

purchasing of products. The preceding literature has shown this evolution of thought

from process to function and then relationship networks.

2.4.2 Exchange relationships

Early B2B marketing theory focuses heavily on the concept of relational exchanges

whereby firms exchange products and people in order to be profitable and therefore to

gain consumer value (Dwyer et al. 1987). Bagozzi (1975) adds that an exchange is a

direct transfer of tangible entities between two parties. The basic premises that

exchange relationships are important to the firm are that:

i) The exchange serves as a focal event between two or more parties that aid in

product transfer;

ii) The exchange allows for the individual firms to identify the roles they play

in the exchange which allows them to recognise weaknesses in the roles and

better them, thereby aiding the exchange;

iii) The exchange allows for the product that is to be exchanged to be examined

for faults or benefits of the product to be realised; and

iv) The exchange can be observed so that the parties in the exchange can make

judgements as to whether the exchange was successful or otherwise.

(Dwyer et al. 1987)

While exchanges have numerous benefits to firms, as discussed above, they are

mechanical in nature and have processes similar to industrial purchasing, discussed

earlier in this chapter. This is shown by Frazier (1983) who highlights three processes

in exchange relationships:

initiating processes whereby a product needing recognition is made and partner

search (to fulfil at need) is initiated;

an implementation process where the product flows between the two companies

and therefore, an interaction between the firms occurs; and

a review process whereby the exchange is evaluated in terms of the benefits of

the product obtained and whether the goals were obtained.

20

Bagozzi (1975) discusses the nature of exchanges and has categorised three types of

exchanges:

i) A restricted exchange which contains two parties in a reciprocal relationship

whereby A gives to B for example a buyer purchases a product from a

supplier;

ii) A generalised exchange which is considered to be an univocal, reciprocal

relationship whereby there are at least three actors in the exchange but some

do not benefit directly from the exchange. For example, a grape grower

supplies grapes to a winery who then transforms it to wine and sells the wine

to a retailer. The label on the wine bottle lists the grape growers details and

as such the grape grower gains value and benefit due to consumer

recognition (on the basis that consumers value this information); and

iii) A complex relationship which is a mutual relationship between at least three

parties with a direct relationship between each, such as a supplier-

manufacturer-distributor relationship.

Bagozzi (1975) discusses the fact that within the exchange is a “medium”, some form

of communication, which allows information to flow between each party. Furthermore,

Bagozzi (1975) introduces the concept of social marketing which is a precursor to a

discussion on relationships whereby relational norms occur through social interaction

between the parties during the exchange, and these strengthen the relationship and aid

in relational continuity. This concept of social marketing is similar to Thibaut &

Kelley‟s (1953) relational norm concept imbedded within social exchange theory in that

cooperation and commitment and other relational norms are imbedded in the exchange

via social interaction.

Frazier (1983) also discusses a structure for inter-firm exchange. Frazier (1983)

comments that the relationship has elements or “sub-processes” such as achieved

influence, goal compatibility, role satisfaction, manifest conflict, conflict resolution

and, finally, cooperation and effort, and these can be linked to the relational norm

concept developed by Thibaut & Kelley (1953). The review process is an assessment

of the benefits or losses achieved by each firm as a result of the exchange. Similar to

the expansion phase in Dwyer et al.‟s (1987) theory is Frazier‟s (1983) model which

shows the expansion phase as involving a great level of interaction between the two

actors which results in continuing relations which create satisfaction for each partner.

The concept of a structured, mechanical exchange is further developed by Weitz (1981)

21

who observed the exchange, not from a firm‟s perspective but from the personnel

involved in the exchange, such as sales persons. Weitz (1981) discusses the personal

characteristics of the sales person, such as demeanour and selling ability, which will

affect the exchange; he also discusses the communicative abilities which aid the success

of the exchange, in line with Bagozzi‟s (1975) social marketing concept. The rapport

developed between the individuals involved in the exchange aid in the development of

relational norms such as role satisfaction and conflict resolution which ultimately aid

exchange success, similar to those proposed by Thibaut & Kelly (1953) and Frazier

(1983).

Lambe et al. (2001) offers a comprehensive review of social exchange theory as applied

to B2B relationship literature. They discuss the following four premises of social

exchange theory:

1) exchanges result in economic and/or social outcomes;

2) the outcomes are evaluated over time to substitute exchanges to determine how much

dependence is required on the exchange;

3) positive outcomes over time increase a firm‟s trust in their trading partner and

commitment to that exchange; and

4) positive exchange interactions over time produce new relational exchange norms that

govern the exchange relationship.

Blau (1964, 91) defines social exchange as “voluntary actions of individuals that are

motivated by the returns they are expected to bring and typically do, in fact, bring from

others.”, meaning that interactions are motivated by the notion that further benefits will

occur to the actor if they keep interacting in a positive manner. The interactions are

motivated by benefits such as common norms, roles, or goals and these elements act as

incentives for the social interaction. A network of social relationships and group

structures then begins to emerge. Finally, group norms and expectations become more

solidified (Morris, 2005).

Blau (1964, p 92-93) distinguishes social from economic exchange by arguing that

economic exchange entails specific obligations while social exchange “involves the

principle that one person does another a favour, and while there is a general expectation

of some future return, its exact nature is definitely not stipulated in advance”. As the

future obligations are not specified, trust in the exchange partner is necessary for social

22

exchange. Such actions help to create a relationship that is long-term, as social bonds

between the actors become strengthened by remaining connected to each other as well

as through a long period of trusting that others will discharge their own obligations

(Blau, 1964; Morris, 2005). Similar to the discussion of Blau (1964) are the comments

of Thibaut & Kelley (1959) who suggest that the creation of relational norms may serve

in the place of contracts or other legal mechanisms. The elements of relational exchange

in marketing channels are strengthened by norms of role integrity, relationship

preservation, and harmonization of conflict (Brown et al., 2000).

It can be shown that many of the most significant postulations from exchange theory

provide insight into interactions among firms. The tenets of social exchange theory state

that interactions involve trust and that, as the interactions increase, relationship

continuity is engendered. This notion can be further explained by Dwyer et al. (1987)

who present a model to illustrate buyer-supplier relationships along the transaction-

relational continuum. Dwyer et al. (1987) comment that when the levels of net

expected benefit are high in absolute terms for both partners, “bilateral relationship

maintenance” occurs and so both actors work to maintain the relationship.

Much of the B2B marketing literature regarding relational exchange theory has been

posited in earlier times, and criticisms of exchanges in fully understanding inter-firm

interaction have been made. Dwyer et al. (1987) believe that the relationship aspect of

exchanges has been neglected in the literature, particularly the dyadic perspective of

relationships, and that there needs to be a greater emphasis on investigating the benefits

and the effect of ongoing relationships. The concept of ongoing, dyadic relationships is

therefore, the area of discussion in the next section.

2.4.3 Relationship development and relationship marketing

B2B marketing literature discusses the ways in which firms exchange products and

socially interact in exchange episodes. These premises were discussed earlier. Criticism

has been made that the relationships between actors has been neglected. For example,

Ravald & Gronroos (1996) discuss the shift in discussion to the focus on relationships

whereby inter-firm loyalty enhances profitability and a long term relationship is

engendered. In “close” relationships, the buyer, rather than just evaluating the product

being exchanged, evaluates the relationship. Future purchases from a supplier are not

purely based on the attributes of the product being offered, or whether the product is

23

exactly what is required, but also on whether the buyer wishes to maintain the

relationship. Ravald & Gronroos (1996) focus on relational behaviour, not just during

the length of the relationship, but during the episodes of the relationship such as when

the product is being purchased. Ravald & Gronroos‟ (1996) premise is that the

purchasing decision during this episode is not only driven by the core product benefits

but also by the willingness to maintain the relationship.

The concept of B2B marketing focusing on relationships rather than exchanges is

further discussed by Dywer et al. (1987) who comment that exchanges do not fully

conceptualise inter-firm interaction and that relationships are rooted in the idea of

relational contracts whereby the effect that is associated with the social interaction

between firms is important and both parties make an effort to maintain a relationship

that contains healthy social interaction. Dwyer et al. (1987) further discuss the concept

of relationships by describing a process whereby relationships develop. Dwyer et al.

(1987) discuss five stages of relationship development: awareness of a need to create a

relationship, exploration for firm partners based on product needs and relational

compatibility, expansion of the relationship into further products and subsequent orders,

commitment to the relationship, and dissolution of the relationship. Anderson (1995)

critiques the relationship development process of Dwyer et al. (1987) by stating that,

although the process involves stages, it is very linear in fashion. Anderson (1995)

argues that relationship development is a continuous process, but is remembered by

managers and business owners as a series of exchange episodes that involved personal

experience. Each of these episodes gives the firm a positive or negative appraisal of the

firm, and after each episode the firm can decide whether to continue the relationship at

the same level of collaboration, to broaden it, or to cease it. This approach focuses on

the exchange process discussed previously whereby the firms focus on exchanges and

the benefits derived from them (Bagozzi, 1975; Frazier, 1983).

Further to the discussion of Dwyer et al. (1987) and Anderson (1995) is the premise

posited by Wilson (1995) that relationships develop over time; however, the

development contains various relational variables within the process. Wilson (1985)

discusses a 5 stage process which contains:

1. partner selection which is based on variables such as reputation, social bonds,

mutual goals, trust and power;

2. a definition of the purpose of the relationship which is based on trust, social

bonds and mutual goals;

24

3. setting relational boundaries based on adaptation, power, mutual goals;

4. the creation of relational value based on cooperation, commitment, structural

bonds; and

5. relationship maintenance, based on commitment, mutual investment and

adaptation.

Based on the notion of relationship and relationship development is the concept of

relationship marketing. The tenet of relationship marketing is that greater cooperation

between buyers and suppliers creates competitive success (Morgan & Hunt, 1994). The

relationship marketing process involves four stages:

1. deciding on customer accounts (involving considerations for profit potential);

2. developing account-specific offerings (i.e. product offers specific to the

partner);

3. implementing relationship strategies; and

4. evaluating relationship strategy outcomes (for performance and changes in

customer needs). (Hutt & Speh, 2010).

The tasks involved in the relationship marketing process include the problem of

allocating resources to different relationships and managing interactions within each

relationship (Håkansson et al. 1976; Ford, 1980). In such a way, a winery would

develop different ways of interacting with growers in the process of obtaining grapes.

Concepts of relationship marketing are more than just cooperation between two parties

and refer to relationships in a more personal and less transactional manner. Relationship

marketing is also grounded in social exchange theory, the premise of which is that

parties enter into long-term relationships in order to gain additional benefits.

Relationship marketing and relationship development literature has shown that

interaction between firms is more than episodic or discrete for the purpose of obtaining

product and the interaction is also social in nature.

Further discussion has also considered the interaction between firms which is fashioned

like a network. This is discussed in the next section.

2.4.4 Business to Business networks

Firms interact with each other in order to gain products which can be sold onto other

supply chain members and ultimately are consumed by end users. Prior discussion has

25

focused on the exchange and relational perspective of inter-firm interaction. However,

inter-firm interaction can also be discussed from a network perspective. This

perspective has been advocated by the Industrial Marketing and Purchasing (IMP)

group who view inter-firm interaction from a network perspective (Håkansson &

Snehota, 1995; Simon et al., 2003). Put simply, the IMP perspective is rooted in the

notion that firms are interconnected, particularly in terms of the activities performed by

each actor, and the resources which are utilised and obtained to facilitate the

relationship (Simon et al., 2003). The interconnectedness of all the firms creates a

network that not only facilitates the movement and purchasing of product but also the

psychosocial interaction between the firms (Simon et al., 2003). The IMP perspective

also posits that networks are an efficient form for organising business activities and that

there is something to gain from operating in a network rather than being a “lone ranger”

in the market. This notion is further developed by Geersbro & Ritter (2010) who

comment that due to the network-like behaviour of the inter-firm interaction, the

relationship is not under the control of one firm in the interaction but by bilateral

interaction between firms. However, business networks can enable or hinder firm

performance (Håkansson & Ford, 2002) as the network allows actors to gain

connectedness and share networks that allow for the creation of efficiencies that lead to

customer satisfaction, but can also hinder efficiency due to the potentially large number

of parties in the network which can cause conflict to arise because the connectedness

becomes too complex and the interests of individual firms are forgone in the interests of

the network.

The IMP perspective also discusses the notion of “ingredients” that create and sustain

the network, similar in fashion to the exchange and relationship process development of

Frazier (1983) and Dwyer et al., (1987). Ford et al., (2003) state that networks consist

of three variables:

1. bonds between actors, which evolved over time through activities such as

purchasing and buying, and such as social bonds rooted in personal

relationships. These bonds are enhanced due to the interdependence that results

from close knit networks;

2. resource ties, such as investment in resources to aid the networks such as IT

systems to improve communication across the network .These resources are

embedded in the network and are adapted for the purposes that are needed

(Lusch & Brown, 1996); and

26

3. activities of network members that are fashioned in a way to gain maximum

benefit to the network as a whole, such as which firms will be involved in

activities such as logistics, information dissemination and manufacture.

The benefit of the network perspective to individual firms can be viewed in the

following way: by having a high quality relationship with other firms in the network,

investing in resources to gain access to another firm, and by being effective and

efficient in the activities they perform in the network, a firm can gain a strong “network

position” that will lead to its success and profitability (Ford et al., 2003).

This is particularly true in the wine industry where the strength of the grape grower, in

growing grapes, and the strength of the winery, in processing those grapes into wine,

ensures that a wine product is made that results in a monetary compensation to the

grower (for their grapes). Therefore, the grape grower gains from the relationship by

gaining monetary compensation and the winery gains by having a product which is fit

for market, can be sold, and thus the winery also gains monetary compensation.

The obtaining of benefits from a mutually rewarding relationship is, of course, not

automatic. Management of supplier/buyer relationships is necessary to gain from the

relationship. Links between the buyer‟s operations and those of the supplier can be

adapted to improve efficiency and performance. Furthermore, firms may choose to

combine resources such as facilities, equipment or operations in order to strengthen ties

with a trading partner (Ford et al. 2003). Over time, the development of actor bonds

may create continuity of the relationship (Wilson, 1995). These actor bonds have most

commonly been characterised by relational elements such as commitment. Commitment

is an implicit or explicit pledge of relational continuity between exchange partners that

occurs at an advanced stage of the relationship (Dwyer et al. 1987). The literature

discussing commitment focuses on the notion that it only exists in successful

relationships that are high in levels of satisfaction and contain solidarity and cohesion

(Dwyer et al. 1987; Gundlach et al. 1985).

Commitment to a relationship is not only denoted by the level of investment, over time,

made by each party, such as investments in capital items that facilitate the transfer of

goods and services such as logistics systems, but also the amount of time given to

maintaining the relationship and the salespersons allocated to the relationship (Ford,

1984; Dwyer et al., 1987; Gundlach et al., 1995). Relationships that have high levels of

commitment exhibit behaviours from the actors, referred to as relational norms

27

(Gundlach et al. 1995). For example, relationships that are high in commitment exhibit

lower levels of relational norms such as opportunistic behaviour and higher levels of

adaptation (Mohr & Spekman, 1994; Gundlach et al., 1995). Therefore, actors that are

involved in highly committed relationships not only make a high level of investment in

terms of time and resources but also forgo short term goals for long term benefits and in

turn are less likely to engage in opportunistic behaviour (Mohr & Spekman, 1994;

Gundlach et al., 1995).

Business networks can provide other benefits through the social interaction that occurs

between members of the networks (Benson-Rea, 2005). The interaction between

members of these networks can provide each member with information that allows each

firm to find new markets which in turn allows for information concerning new products

and assists with new product development (Blankenburg-Holm et al., 1996). As such,

these social interactions can aid the businesses in finding new markets and market

expansion, thereby assisting in business profitability and sustainability.

While business relationships differ depending on whether they are exchanges, dyadic

relationships or part of a network, the relationship outcomes will ultimately affect each

other (Mandjak & Simon, 2004). This empirical study addressed a gap through the

development of a theoretical model to conceptualise and measure the effect that specific

relational norms have on relationship quality created through trading relationships

between buyers and sellers of wine grapes in Australia. The model viewed the

suppliers‟ (grape growers‟) perspective of the relationship, as it is considered that this

actor plays the most vital role in the production of wine.

This study focused on a specific relational norm of communication (namely, a theory of

collaborative communication) and how it affected quality in the relationship.

Development of the theoretical framework entailed the selection of relational norms

which reflected the interaction between the two actors and observed the effect that

environmental influences (such as power asymmetry) had on exchange behaviour

(Håkansson, 1982). A number of exploratory in-depth interviews with grape growers

were conducted to ensure that the selected relational norm (collaborative

communication) and power asymmetry were relevant to trading relationships in the

Australian grape and wine industry.

28

The next section of this chapter discusses the relational norms evident in business to

business relationships which were tested on grape growers in the exploratory stage of

the study to gain an understanding of their appropriateness for the conceptual model.

2.5 Relational norms Although many researchers have used various ideas for conceptualising relational

behaviour constructs, the relational contracting theory is relatively comprehensive

(MacNeil, 1978; Dwyer et al., 1987; Heide, 1994). MacNeil (1978) posits that formal

contracting is but one of the mechanisms to govern business relationships and that

exchange partners will develop joint expectations about what behaviours are

appropriate in order to complete formal arrangements (Heide, 1994). The relationship is

thus governed by certain expected behaviours, namely relationship norms (Thibault &

Kelley, 1959; Heide & John, 1992). Furthermore, the general property of the relational

norm is the prescription of behaviours that aim at maintaining a relationship and their

rejection of behaviours that promote individual goal seeking (Heide & John, 1992). In

evidence of this, Ivens (2004, p 301) has argued that “…every norm refers to a potential

behaviour and the norm framework may be used as a structuring scheme for research on

relational behaviour”. The literature has highlighted numerous relational norms and

Table 2.1 shows a number of relationship norms as summarised by Ivens (2004) from

other relationship literature.

Table 2.1 exhibits a synopsis of the literature regarding relational norms. The table

provides an overview of the various norms and it would be of interest to observe which

set of norms is applicable to the Australian wine industry and the grape grower/ winery

relationship. Furthermore, the relational norms are “building blocks” of the relationship

and as such must be viewed in terms of their effect on a whole relationship. Of

particular interest is the relational norm of communication.

29

Table 2.1: List of relational norms

Norm/behaviour Description

Cooperation The coordination tasks which are undertaken jointly and

individually to pursue common and/or compatible goals and

activities undertaken to develop and maintain the

relationship (Young & Wilkinson, 1997; Leonidou et al

(2002, 2006); Woo & Ennew, 2004)

Social bonds A personal relationship resulting from the economic

exchange that can be linked to social bonds which are a

“glue” that holds the individuals together (Turnbull &

Wilson, 1989; Bendapudi & Leone, 2002)

Communication Readiness to proactively provide all information useful to

the partner (Mohr & Nevin, 1990; Heide & John, 1992;

Lusch & Brown, 1996; Mohr et al. 1996)

Solidarity The preservation of the relationship particularly in

situations in which one partner is in a predicament

(Kaufmann & Stern, 1988; Achrol, 1997)

Flexibility Actor‟s readiness to adapt to an existing agreement

(implicit or explicit) or to new environmental conditions

(Nordewier et al. 1990)

Conflict resolution The use of personal, friendly and informal mechanisms to

resolve conflicts (Kaufmann, 1987)

Cultural fit Understanding of partners‟ attitudes and behaviours and

appropriate interpretation of actions (Gyau & Spiller, 2007)

As previously discussed, the two wine industry actors must liaise during the growing

season to create a grape product fit for their purpose. To do so, the two actors must

communicate to convey the necessary information, particularly in regard to grape

parameters such as sugar content, berry size and residual chemical content that results

from the use of pesticides and herbicides (Clancy, 2005). Communication is performed

via various modes (e.g. face to face, electronic, telephone, seminars, newsletters etc). It

30

is of interest to observe how these different modes and their frequency influence grape

growers‟ perceptions of relationship quality. Furthermore, it has been highlighted in

wine industry trade literature that communication between grape growers and wineries

is important (Spawton & Walter, 2003; Chong, 2007; Hobley, 2007; Brown, 2008).

Discussion on the literature regarding communication is warranted and is the topic of

discussion in the next section.

2.6 Communication Communication is important in establishing objectives and coordinating activities to

meet those objectives (Mohr et al., 1996). Much of the literature discusses the effect

that specific dimensions of communication have on the relationship (Mohr et al., 1996)

and the openness of the information (Heide & John, 1992), and is of particular interest

to this study whereby relational norms have an effect on relationship quality.

Communication frequency is a dimension of communication that requires observation.

Daft & Lengel (1984) suggest that the modes of communication differ in their fertility

and their ability to convey information, and that richer modes of communication (such

as face to face) allow for more tailored communication (specific to the circumstance)

and allow for immediate feedback. Daft & Lengel (1984) also discuss written and

electronic forms of communication as being less “rich” and more useful in

communicating large amounts of homogeneous information. Given these observations it

is of interest to observe how these various modes and dimensions affect relationship

quality.

From a winery perspective, these communication modes have varying degrees of cost.

Cannon & Homburg (2001) comment that richer modes are more costly, but also

concede that the effectiveness and efficiency of communication must be matched with

the mode. With respect to the efficiency of communication, Daft & Lengel (1984) state

that complex, unstandardised information is best communicated by rich modes (for

example face to face) as opposed to less important more “mechanical” information that

is best transmitted via less rich modes such as written or electronic. Furthermore, less

rich modes of communication (such as written or electronic) can supplement the rich

forms (for example face to face), particularly if the supplementary information is self-

explanatory and does not require a rich description from the sender (Cannon &

Homburg, 2001). In regards to the specific frequency of the mode of communication,

31

an increase in frequency produces a greater volume of communication to be

transmitted, hence improving the understanding of the problem faced by the supplier

(O‟Neal, 1993).

Associated with the notion of communication is information sharing. This is defined as

the extent to which the supplier shares information with the buyer and this can lead to a

fruitful relationship with the buyer if the information leads to a lowering of costs and

greater relational efficiencies that result from understanding future plans of the supplier

and the coordination of production development (Anderson & Narus, 1990; Cannon &

Homburg, 2001).

Further to the notions of frequency and modality of communication is the multi-

dimensional nature of collaborative communication as proposed by Mohr & Nevin

(1990). Mohr & Nevin (1990) propose that communication has facets beyond

frequency and modality and includes such aspects as the formality of the

communication (whether communication is formal or informal), its bi-directionality

(whether the communication flows only from one actor or from both i.e. in both

directions of the relationship), and indirect influences of communication (whether the

communication indirectly affects the activities of the partner). From a wine industry

context, the importance of communication between actors has been identified as an

issue (Spawton & Walter, 2003; Chong, 2007; Brown, 2008). Hobley (2007) has shown

that communication is an important relational dimension in the Australian industry; this

is also an important issue in other countries‟ grape grower/ winery relationships.

Redondo & Fierro (2007, pg 86) discuss this issue from the Spanish wine industry

context where it was found that increased communication makes the relationship

“continual” and resulted in greater levels of satisfaction for the actors. Brown (2008)

added that communication between grape growers and wineries is of major importance

to the success of individual grape relationships, with Chong (2007) further adding that a

greater understanding and refinement of communication and information systems is

needed in the industry.

This discussion regarding communication has illustrated the multi-dimensional nature

of the communication construct, including the nature of information sharing, its overall

effect on the relationship, and its importance in the Australian wine industry context.

Therefore it would be of interest to identify the modes, frequency, formality, bi-

directionality and influence of communication from a grape grower‟s perspective and to

observe how these dimensions influence relationship quality.

32

2.7 Relationship quality While relationship norms and variables can be considered building blocks of the

relationship, the quality of the relationship is also an important factor which can be

observed. Relationship quality refers to a supplier‟s perception of how well their

relationship fulfils his expectations, predictions, goals and desires (Gyau & Spiller,

2007). According to Wong and Sohal (2002), relationship quality conveys a customer‟s

impression about the whole relationship and, as such, is manifested in several distinct

but related constructs. However, there seems to be no consensus among researchers on

the set of constructs or variables that constitute relationship quality (Crosby et al., 1990;

Ceceres & Paparoidamis, 2007; Gyau & Spiller, 2007).

In spite of the fact that researchers conceptualize relationship quality with dissimilar

dimensions, they appear to concur generally that relationship quality measures actors‟

awareness of how well their relationships with their partners fit, and is often connected

to a firm‟s ability to sustain their relationships in the long-term. Ceceres &

Paparoidamis (2007, p 837) affirm that “…there is general agreement in the relationship

marketing literature that the quality of the relationship between the parties involved is

an important determinant of the permanency and the intensity of the relationship and

the consequent success of relationship marketing practices”.

Studies involving relationship quality draw heavily upon the social psychology

literature. Unlike relational norms and elements (such as social norms, flexibility and

shared goals) which are uni-dimensional constructs measured in a uni-dimensional

fashion (e.g. the trust construct is measured via latent variables concerning trust); the

literature discusses relationship quality as a multi-dimensional higher order construct

that consists of trust and satisfaction. Crosby et al. (1990) discuss the relationship

quality as being comprised of trust and satisfaction and argue that if a partner can be

relied upon to fulfil his duties in the interest of the relationship, then satisfaction will

occur. Wray et al. (1994) and Lagace et al. (1991) affirm this notion, and add that trust

helps in allowing tensions to be resolved which results in satisfaction for the partner.

This alludes to a notion that trust is an antecedent of satisfaction and relationship

quality, although the literature has not confirmed this notion. Kim & Cha (2002) and

Kim et al. (2006) for instance, conceptualise the relationship quality construct as

indicative of the level of satisfaction and do not discuss the influence of trust; however,

their comments oppose those of Dwyer et al. (1997) that commitment is not a measure

33

of relationship quality but a predictor or outcome of it, whereby trust in a partner and

satisfaction in the relationship leads to commitment. Therefore, Kim & Cha (2002) and

Kim et al. (2006) allude to the fact that relationship quality is an antecedent of

commitment and further add that relationship quality is a higher order construct that

reflects the strength of the relationship. Other researchers, such as Gummeson (1987),

Leuthesser (1997), Dorch et al. (1998), Naudé & Buttle (2000) and Parsons (2002),

further argue that relationship quality is comprised of trust and satisfaction. However,

Scheer & Stern (1992) and Leuthesser (1997) empirically tested relationship quality as

a uni-dimensional construct whereby the construct of relationship quality consists of the

latent variables of trust and satisfaction. Given the framework that has been adopted for

study, and the prevailing emphasis in the literature linking trust and satisfaction to

relationship quality, the author has conceptualised relationship quality as a measure of

trust and satisfaction.

The preceding literature discussed whether relationship quality is a multi-dimensional

construct comprised of trust and satisfaction; however, there have been two instances

where relationship quality has been judged to be a uni-dimensional construct (see

Scheer & Stern, 1992 and Leuthesser, 1997). Crosby et al. (1990), Dorch et al. (1997),

Kim & Cha (2002) and Kim et al. (2006) empirically test relationship quality, mostly

via SEM and other multi-variate regression techniques, and use trust and satisfaction as

separate constructs; they discuss whether higher levels of trust and satisfaction in their

model correspond with higher levels of relationship quality.

This notion has been applied to this study and is discussed further in Chapter 6.

However, an alternative estimation of relationship quality, based on a uni-dimensional

measurement, has also been performed in this study and is further discussed in Chapter

3, section 3.12.

Regardless of the estimation technique, the literature has shown that relationship quality

is comprised of trust and satisfaction.

2.7.1 Trust as a dimension of relationship quality

Trust is defined by Zaheer et al. (1998, p 21) as the principle that the business partner

“can be relied upon to fulfil obligations and behave in a predictable manner”. However,

trust is not attainable in the short-term. Blau (1964) commented that trust is the result of

34

repeated exchanges between two organizations. Houston & Gassenheimer (1987, p 10)

affirm this statement and add that trust between two parties “…leads to a long term

relationship”. Trust also decreases risk, particularly as it can act as an “…information

resource that reduces the threat of information asymmetry and performance ambiguity”

(Batt, 2003 p 66). Trust also results from the expertise, reliability or intentionality of the

partner, and can be built by the competence, honesty, dependability and likability of the

partner (Batt, 2003). From an SME context, trust has been shown as an important

ingredient in the creation of partnerships, strategic alliances and networks (Brusco,

1986; Smitka, 1991; Powell, 1996). Additionally, Sako (1997) viewed trust from three

perspectives, namely contractual, competence and goodwill trust. Contractual trust is

concerned with the extent to which parties can carry out their contractual obligations.

Competence trust relates to the understanding of professional and technical standards,

and goodwill trust denotes that the relationship has a degree of fairness related to

practices. Adding to Sako‟s (1997) discussion of trust perspectives, Kumar et al. (1995)

discuss that trust has two elements:

1. trust in the partner‟s honesty and the belief that the partner will stand by his

word and fulfil his obligations and is sincere; and

2. trust in the benevolence of the partner in that the actor is interested in the

welfare of the partner‟s firm and won‟t work to take actions that will negatively

affect that firm.

Trust in relationships also has many benefits for each firm. Relationships that contain

trust will also be better able to manage conflict within the relationship and a greater

degree of adaptability to the other firm‟s requests will occur (Mohr & Spekman, 1994).

Once trust is evident in a relationship, the actor understands that joint efforts will lead

to outcomes that exceed what could have been achieved if each firm acted solely in

their own interests (Mohr & Spekman, 1994). Conversely, a lack of trust in a

relationship can lead to decreased relational norm effectiveness such as a decrease in

communication quality and the ability to jointly solve problems when they occur.

Trust is also developed in a relationship over time and it has been shown that trust is an

antecedent of commitment (Morgan & Hunt, 1994). Furthermore, evidence that trust is

being developed in a relationship is shown if:

1. an actor is willing to customise their equipment and processes to the other

actor‟s requirements;

35

2. actors are willing to share confidential information; and

3. in line with the discussion of Morgan and Hunt (1994), are willing to engender a

long term relationship.

(Doney & Cannon, 1997)

2.7.2 Satisfaction as a dimension of relationship quality

Satisfaction refers to a positive affective state resulting from the appraisal of all aspects

of a firm‟s working relationship with another firm. Satisfaction is important to the long

term success of the firms involved in a working relationship and therefore encourages a

long-term relationship and relational continuity (Oliver, 1980, Anderson & Narus,

1990; Ganesan, 1994). Various dimensions of the BS relationship have been discussed,

particularly how intensity affects the relationship quality. If the intensity is low, the

relationship quality is poor; however, satisfaction is a consequence of a positive

relationship. Batt (2003) describes satisfaction as occurring when performance exceeds

expectations. Oliver (1980) further describes satisfaction as a result of an evaluation

between the partner‟s performance and the firm‟s expectations. Further studies show

that satisfaction positively enhances trust (Mackenzie & Hardy, 1996) with Geyskens et

al. (1999) arguing that if the channel members are highly satisfied, the partners believe

them to be more trustworthy. However, satisfaction‟s influence on trust is not easily

attained. Batt (2003, p 69) states that “…satisfaction with an exchange will lead to some

initial trusting behaviours, but as satisfaction increases, trust will increase”. Fornell

(1992) further adds that satisfaction is evident in quality relationships and is cumulative

over time and based on experiences.

Satisfaction has been discussed as a function of expectations in the partner firm‟s

performance (Oliver, 1980). The perception of the performance leads to post-purchase

satisfaction; however, Anderson & Narus (1990) warn that satisfaction as an area of

academic research is fraught as it is a highly subjective construct. Anderson & Narus

(1990) discuss satisfaction as being linked to perceptions of influence; if a firm believes

they have greater influence over their partner, they experience higher levels of

satisfaction. This appears to relate to power asymmetry whereby the actor which has the

higher level of power has greater satisfaction, although Anderson & Narus (1990) do

not comment on this. Anderson & Narus (1990) do add that relational norms will affect

satisfaction, that conflict and disagreements between firms will block goal attainment

36

and lead to decreased satisfaction, and that cooperation and mutual goal attainment

positively affect satisfaction.

As previously discussed in wine industry trade literature, relational norms such as

power asymmetry are having a great effect on grape grower and winery relationships in

the Australian wine industry and, as power asymmetry is a relational norm, this will

have an effect on relationship quality. This is highlighted in the next section.

2.8 Power Asymmetry As noted earlier, wineries in the Australian wine industry have been cancelling supply

contracts and not maintaining business relationships due to an oversupply of grapes.

The cancellation of contracts is a result of power asymmetry.

Power asymmetry is not uncommon in the wine industry. Discussed in Chapter 1, there

are incidences of power asymmetry disrupting and potentially harming grape grower

and winery relationships such as is evident in Kosovo and France where this has led to

protests and violence (Phillips, 2000). Power asymmetry in these countries was

attributed to a lowering of the grape harvest as a result of pressure further up the wine

supply chain, such as supermarkets discounting imported wine products or decreasing

exchange rates. The exchange rate decreases resulted in decreased revenue for wineries,

which in turn created cost pressures for wineries; they alleviated these pressures by

offering lower grape price (Phillips, 2000).

However, there is also evidence in the wine industry of the effects of the power

asymmetry being felt by the wineries; in this case the grape grower holds the power in

the relationship. This appears to be the case in wine regions where grape products are in

high demand. This phenomenon is documented by Redondo & Fierro (2007) whose

study on the Somontano wine region in north western Spain examined a region where

grape produce was in high demand and therefore where grape growers attained power

over the wineries and could demand higher prices. Charters & Menival (2010) showed

a power asymmetry favouring grape growers due to wineries wishing to maintain high

quality in their products from the Champagne region; grape growers had a level of

power over wineries and could dictate terms. This was exacerbated by the fact that there

was a shortage of grape growers in Champagne due to geographical restriction.

Therefore, in the study of Redondo & Fierro (2007), grape growers gained power due to

37

a scarcity of their product, while in the study of Charters & Menival (2010) grape

growers‟ gained power because there was only a small population of grape growers, and

the wish of wineries to maintain quality standards was only possible by rewarding grape

growers who produced high quality grapes.

While power and power asymmetry is documented in the wine industry, it is also

evident in academic literature. By definition, power is the ability of one actor to

influence another to act in a manner that he/she would not have otherwise chosen

(Emerson, 1962). Cox et al. (2003) contribute to the discussion by arguing that

buyer/seller relationships are driven by the power maintained by one organisation

which is willing to take whatever action is necessary to maintain that dominant position.

Cox et al. (2003) further identify four main power structures in the literature:

dominance, interdependence, independence and dependence (Cox et al. 2003). Thibaut

and Kelley (1959) explore the issue of both coercive and constructive conflict in

impersonal relationships, and this was applied in a business to business context by Ford

(1984), who argued that all inter-organisational relationships exhibit conflicts and

cooperation simultaneously and that the two are not mutually exclusive. Power has also

been attributed to conflict in a distribution channel with the nature and origin of the

power that the channel member possesses influencing possible conflict (Gaski, 1984).

However, while creating conflict, power has also been viewed as a “moderating

power”, alluded to by Ford (1994), whereby the power allows for conflicts and

cooperation to exist simultaneously. Reve & Stern (1979) mention that power is used to

organise the channel member and also to ensure that conflict stays manageable. The

power in the channel is confirmed by Seyed- Mohamed & Wilson (1990) who mention

that the greater the degree of threats made by the buyer, the greater the amount of

disturbance that exists in the relationship. Operationally, the seller wishes to have his/

her products purchased; thus, if the seller is dependent on the buyer, the buyer will have

the power in the relationship thereby creating a power imbalance (Wilson & Vlosky,

1998). Anderson & Weitz (1989), Ganesan (1994) and Varadarajan & Cunningham

(1995) pointed out that in a situation of power imbalance, the party with the higher level

of power will try to exploit its advantage in such a way that the other party becomes

dissatisfied with the relationship. This is prevalent in the Australian wine industry,

where the winery has the power to accept or decline the supply of grapes from grape

growers and can use their dominant position to demand certain requirements from grape

38

growers. This is exacerbated in a scenario where the seller has a limited number of

buyers to select from, which is evident in the current Australian grape grower/winery

relational circumstance. Therefore, the relational norm of power asymmetry is affecting


The preceding discussion regarding relational norms, in particular communication,

power asymmetry and relationship quality, has been linked to current dilemmas in the

Australian wine industry.

2.9 Literature Discussion This chapter has discussed literature in the B2B marketing and inter-firm behaviour

area. It has also combined both academic and wine industry trade literature in order to

understand the background to a problem. The purpose of the chapter is to highlight the

nascent literature in the domain of business to business marketing and purchasing which

occurs between grape growers and wineries, the context for this study.

The chapter has shown that unlike purchasing in consumer markets, the decision

making process used to identify which products should be purchased is far more

complex in business markets. Furthermore, the volume of product purchased in B2B

markets is far greater than in B2C markets and the time and effort exerted in B2B

purchasing is great due to the importance of the process to firms (Kotler et al., 2010).

Similarities can be shown between purchasing decision processes in consumer and

business markets, though a greater level of time and resources is used in the latter

(Moriarty, 1983; Johnson & Lewin, 1994; Ford et al., 2002). With regard to the actual

purchasing of product in a B2B context, firms coordinate various intra-firm departments

and activities (such as logistics, finance, manufacturing); however, what sets it apart

from consumer markets is the level of decision making and the process of decision

making whereby buying centres are used to deliberate over and coordinate the process

(Ozanne & Churchill, 1971; Webster & Wind, 1972; Bradley, 1974), the main aim

being to gain consumer satisfaction (Porter, 1985). In effect, purchasing has a process

and function perspective whereby processes are performed to gain product, and many

functions of the business are coordinated to facilitate this process.

However, this concept of purchasing is highly focused on intra-firm activities and does

not comprehend fully the concept that firms interact in the purchasing process. Earlier

literature discusses inter-firm relationships as social exchanges whereby the exchange

39

has numerous elements conceptualised as relational norms such as cooperation,

solidarity and social bonds that are built by the parties and maintained in order to

maintain the relationship and to reduce transaction costs and resource misallocation that

results from discrete relationships (Thibault & Kelley, 1959; Wilson, 1995). This

discussion is based on the concepts of relationship marketing and social exchange

theory which are generally based on the notion of a dyadic relationship whereby two

parties interact in a relationship.

The concept of a relationship in B2B markets is further developed by the IMP group

who argue that relationships are more than dyadic and involve networks in and around

the firms; this can involve numerous firms which support the activities of the

relationship (Hakansson & Snehota, 1995; Ford et al., 2003; Simon et al., 2003;

Geersbro & Ritter, 2010). As numerous firms are involved in the network, each makes

high levels of investment of capital resources and time to establish and maintain

themselves in the network. In well run networks, relational norms such as cooperation,

communication, social bonds and flexibility are in existence at high levels and help to

maintain the network with actions such as opportunistic behaviour being discouraged as

it impedes other firms within the network and so is detrimental to it (Ford et al., 2003;

Simon et al., 2003).

Of interest to this study is the relational norm of communication. It has been described

as the glue of a relationship, and has been discussed by wine industry literature as being

highly important to grape grower and winery relationships and to the wine industry as a

whole (Spawton & Walter, 2003; Chong, 2007; Hobley, 2007; Brown, 2008). From an

academic literature perspective, Mohr & Nevin‟s (1990) collaborative communication

theory states that communication has various elements such as formality, frequency,

non –coercive communication attempts and bi-directionality, and that this theory has

not been tested in a wine industry context.

While exchanges, relationships and relationship networks act to facilitate the flow of

product through the marketing channel, the outcome of the relationship is important,

particularly the perception of the quality of the relationship that an actor has. The

chapter has discussed the theory of relationship quality and shown that a high quality

relationship contains high levels of trust and satisfaction and vice versa (Gummeson,

1987; Leuthesser, 1997; Dorch et al., 1998; Naudé & Buttle, 2000; Parsons, 2002).

While the discussion of the B2B literature has contained various theories and concepts

of importance to the wine industry, trade literature has shown the relational norm of

40

communication (Spawton & Walter, 2003; Chong, 2007; Hobley, 2007; Brown, 2008)

and power asymmetry (Phillips, 2000; Redondo & Fierro, 2007; Charters & Menival,

2010) affect the wine industry and should be observed. This is the underlying premise

of this study.

Mohr & Nevin‟s (1990) theory of collaborative communication has not been studied in

the wine industry context and from an academic perspective how the elements of

collaborative communication affect relationship quality or the affect that power

asymmetry has on relationship quality, has not been studied either.

In summary, based on concepts taken from academic and wine industry trade literature,

this study will attempt to examine how collaborative communication elements and

power asymmetry affect relationship quality. This has not been studied previously

either in an academic context or in a wine industry context, and has been highlighted in

the wine industry literature as being important.

2.10 Chapter conclusion This chapter detailed literature pertaining to B2B marketing, particularly related to

communication, power asymmetry and relationship quality. The next stage of the study,

the exploratory stage, involves qualitatively testing (by way of in-depth interviews with

grape growers) these concepts to understand the effect these elements (i.e.

communication, power asymmetry and relationship quality) have on their perception of

the relationship.

The exploratory stage is followed by the descriptive and causal stages in which the

results from the exploratory stage are quantitatively tested (via a questionnaire).

However, before the results of the exploratory stage can be presented, the methodology

of how the data collections were performed will be discussed and both provide the topic

of the next chapter.

41

Chapter 3: Exploratory research methodology and results

3.1 Chapter introduction The first chapter of this dissertation outlined the Australian wine industry rationale for

the study, including the current economic state of the industry, and defined research

objectives and problems related to this study. The previous chapter (Chapter 2)

provided a theoretical background to the research objectives and problems by

highlighting the relevant literature in relation to B2B interactions and marketing.

Further investigation was required to verify the effect that power asymmetry and Mohr

& Nevins‟ (1990) collaborative communication theory had on relationship quality and,

as a result, an exploratory research study was required to observe this and create a

conceptual model. The opening part of this chapter describes the exploratory research

study design, including the methodology used and the objectives of the research study.

The concluding sections of this chapter relate to the analysis and results of the study,

including the limitations of the study, the definition of hypotheses, and the presentation

of a theoretical model.

3.2 Exploratory research design A qualitative research study was used to explore the effects that relational norms, such

as communication, and relational elements, such as power asymmetry (as discussed in

Chapter 2), have on relationship quality, utilising the South Australian and Victorian

wine industries as a context. Such an approach is supported by discussions in the

literature that state that this type of methodology is appropriate when a researcher

wishes to understand and further develop hypothetical issues raised in the literature, and

to make sure that they are applicable to the business context to be examined (Zikmund

& Babin, 2007; Leedy & Ormrod, 2010). The qualitative research approach and the

information gleaned was critical in understanding the appropriate questions and scale

items to be used in the descriptive and causal stages of the study, the quantitative phase

of the project. As such, the exploratory, qualitative research approach allowed the

author to ascertain whether the issues highlighted in the literature were applicable to the

wine industry context, and they allow both for more concrete assumptions to be made

42

and also for the creation of a conceptual model to address the research problems and

objectives.

The qualitative methodology employed in this stage of the study was in-depth

interviews (IDI). IDI involves conducting one-on-one interviews with a small number

of respondents to uncover their opinions on issues that are raised by the interviewer

(Boyce & Neale, 2006). This method yields richer information than other methods such

as quantitative methods (survey based methods), and other qualitative methods such as

focus group interviews (Leedy & Ormrod, 2010, Malhotra et al. 2006). The

methodology is also more flexible than other methods such as quantitative ones, as the

issues and questions that are posed by the interviewer can be open ended and the

interviewer has the option to further explore topics as they see fit (Malhotra et al. 2006).

The “one-on-one” nature of the IDI method also allows for confidential information,

which can be problematic in other less confidential qualitative methods (such as focus

groups) that involve interviewing numerous people at one time, to be discussed

(Malhotra et al. 2006). The participants of the IDI in this study were discussing private

business relationships that may have involved confidential information (such as contract

issues and legal issues surrounding abuses of power asymmetry) and as such, the

personal, confidential nature of IDI made them appropriate for this stage of the study.

Thirteen grape growers were recruited for the IDI and were based in South Australia

and Victoria. The location of the growers was based on their close proximity to the

author. The participants were recruited on the basis of the size of their operation and the

nature of the region (in South Australia and Victoria) in which their grape growing

businesses were located. This allowed for a participant base with differing production

sizes, ranging from professional grape growing businesses to “hobby” style grape

growing businesses whose proprietors were less involved in the business by way of

time commitment and financial investment. The location and size of the participants are

illustrated in Table 3.1.

43

Table 3.1: Location and size of grape grower participants‟ businesses

Location Size

McLaren Vale, SA < 10 acres

McLaren Vale, SA 30-40 acres

McLaren Vale, SA 10-20 acres

McLaren Vale, SA < 10 acres

Clare Valley, SA 10-20 acres

Adelaide Hills, SA 20-30 acres

Adelaide Hills, SA < 10 acres

Adelaide Hills, SA < 10 acres

Barossa Valley, SA 40-50 acres



Yarra Valley, Vic < 10 acres

Yarra Valley, Vic 10-20 acres

3.3 Participant sample selection and interview format Of major importance in the planning stage of IDI was the selection of the participants

(Malhotra et al. 2006). The business details of grape growers were obtained from the

various grape and wine region industry associations (for example, Barossa Valley

Vignerons Association, Adelaide Hills Wine Growers Association). The regional

associations provided the details of the grape growers (the size of vineyards, location,

telephone numbers) and a selection was made to gain a broad cross section of growers

with differing production size and quality foci, thus allowing observations to be made

on relationships with differing types of wineries (for example, publicly and privately

owned, high quality and lower quality production wineries). Even though the results of

IDI are not able to be generalised due to the small sample size, efforts were made to

maximise the extent to which the participants were representative of to that of the grape

grower population in Australia (Boyce & Neale, 2006; Hobley, 2007).

44

The participants, as detailed in Table 3.1, differ in their production foci and quality foci.

They are based in warm and cool climate regions with sizes that would indicate that

they produce grapes that are lower or higher in quality. This allowed for responses that

would be more easily generalised to the target population (i.e. grape growers in

Australia). The participants were recruited by telephone, as this was deemed an

appropriate way to make contact with the most suitable interview candidates due to

their disparate location which made personal contact via travelling and meeting

unsuitable (Malhotra, 2006). The reason for choosing the participant, the duration of the

interview, the structure of the questions, the type of questions to be posed in the

interview, and details of how the information obtained would be kept in confidence,

were all discussed in the initial telephone call. A time and a place for the interview was

agreed upon, with all but two interviews taking place in the participant‟s place of

business. The other two interviews took place at the home of the participant in a quiet

and secure room.

In terms of the participants‟ activities and responsibilities within the grape growing

business, the participants recruited had to be the principal owner or manager of the

business with responsibility for the decisions regarding which winery to sell their

grapes to, and to have managerial responsibility over the other employees of the

business. Therefore, in all cases, the participants were either the owner of the grape

growing business or the managing director of the business. In terms of Webster &

Wind‟s (1972) criteria for roles in decision-making in a business, the participants had to

satisfy the „decider‟ criteria for decision making. Consequently, the grape grower

(supplier) selected as a participant in the study were required:

to have significant experience with dealing with wineries in the trading of wine

grapes; and

to be involved in a grape growing business that is generally representative of the

wine-grape growing population.

After consultation with wine industry experts (Davidson, pers. com., December 2008,

Mckenzie, pers. com., December 2008), it was decided that owners or managing

directors of the grape growing business were the appropriate participants for the

exploratory study.

45

3.4 Structure of the interview format Each IDI commenced with a statement by the interviewer that the information gleaned

from the interview would to be kept in strict confidence. The participant was also asked

to consent to the use of an audio recorder that was used to record the interview for later

analysis. Participants were also reassured that the audio recording would not be

exhibited to any other persons and would be used strictly for the study. All the

participants agreed to the requests regarding the confidential nature of the recording.

Each IDI‟s duration was 45 to 90 minutes. The variations in the time of each interview

were mainly due to the time away from work that the participant could allow; however,

all topics were discussed by each of the participants.

While the nature of IDI was unstructured and free flowing, questions were devised to

give the interviewer a direction from which to inquire. Structured questions were posed

to the respondent to allow for the information required for the research objectives and

questions to be attained (Malhotra et al. 2006). As such, the questions were based on

issues highlighted in the literature.

A copy of the questions are shown in Appendix 3. All the questions were open-ended in

style, allowing participants to answer in their own style and to express opinions.

However, prompt questions were also devised so that the participants‟ answers did not

go too off the topic and become irrelevant to the study (Cavana et al. 2001; Malhotra et

al. 2006). The questions in the interview were based around the research objectives

which are discussed in the next section.

3.5 Research objectives The objectives of the exploratory, qualitative study were to explore the nature of the

effect of Mohr & Nevin‟s (1990) collaborative communication theory, and the effects

of power asymmetry on perceived relationship quality of grape growers. While the free-

flowing nature of the interviews was crucial to gathering relevant information,

questions (as discussed in the previous section) were posed based on the literature, as it

was deemed that communication was an important relational norm between grape

growers and wineries in the Australian wine industry. Furthermore, it was identified in

the literature that a power asymmetry favouring wineries was affecting grape growers‟

perceptions of relationship quality. The questions were based on the overall objectives

46

which were required to aid in the development of a conceptual model. Therefore, the

research objectives were to:

1. uncover the extent to which Mohr & Nevin‟s (1990) collaborative

communication elements affect wine grape growers‟ perceptions of relationship

quality; and

2. identify the extent to which power asymmetry was affecting grape growers'

perceptions of relationship quality.

3.6 Audio transcription and data analysis technique Audio recordings were converted into a digital audio file by the recorder and the audio

file was uploaded into a computer. The audio file was then uploaded into computer

software named HyperTRANSCRIBE. HyperTRANSCRIBE allowed the researcher to

play a section of the audio file into headphones and then type the words into a word

processing document. Notepad software was used as the word processing document.

Each speaker in the audio file was identified in the transcript. For example, if the

interviewer was named Frank, an “F” was placed in front of the words Frank spoke.

This allowed easy identification of who was speaking in the final transcribed word

processing document. This process was followed for all thirteen interviews.

Following transcription, the word processing documents were analysed using the

computer program HyperRESEARCH (Version 2.8). HyperRESEARCH is a computer

program used to analyse and highlight words, sentences and phrases from transcription

and categorise them (Hesse-Biber et al. 1991). HyperRESEARCH allowed for themes

to be coded and aggregated so that reports could be made for each of them. For

example, if a participants‟ discussion included the comment “I trust the winery”, this

section of the transcript would be coded as “trust” and similar discussion would also be

coded as such. Following coding of each IDI, a report was made of all the coded

discussion, allowing the researcher to see all the discussion regarding each code.

3.7 Exploratory study results The results from the exploratory phase of the study were coded and analysed in the

HyperRESEARCH computer software program. As previously mentioned, the software

allowed reports to be produced, including a report on the frequency of codes. Table 3.2

47

illustrates the frequency of discussion of topics in the IDIs. The purpose of the table is

to illustrate the level of discussion on each topic and, how it related to the research

objectives discussed in 3.5.

Table 3.2: Frequency of topic (code) discussion in in-depth interviews

Topic of discussion Frequency of comment

( number of times)

Communication 62

Trust 38

Satisfaction 22

Power asymmetry 16

Winery issues 15

Discussion about preceding vintage

(grape harvest)

8

The table illustrates that topics related to the relational norm of communication had the

highest level of discussion, with discussion regarding relationship quality dimensions

(i.e. trust and satisfaction) accounting for the second and third highest topics of

discussion. However, the specific comments that were made by participants must be

viewed in terms of the research objectives detailed in section 3.5. The specific

comments, as they relate to the research objectives, are outlined in the next section.

3.7.1 Research results related to uncovering the effect that collaborative

communication theory has on relationship quality

As illustrated in section 3.7 the relational norm of communication was of crucial

importance to grape growers, evident in their frequency of discussion in Table 3.2. The

research results affirm the literature (Morgan & Hunt, 1994; Mohr & Nevin, 1990;

Mohr, et al. 1996) which has commented that communication positively influences the

buyer-supplier relationship. Some representative comments (i.e. representative of

similar comments by other participants) include the following:

48

Participant 6:

“ I think that communication is without and in this scenario (in the current relationship

with their winery) and it is the most valuable thing you have got in the relationship”

Participant 1:

“ I just like good honest communication……I know times have been hard and wineries

need to make cutbacks but good honest communication is what is needed”

Participant 11:

“One guy rejected our fruit because he said it had too much MOG (material other than

grape) in the bins……I reckon he just didn’t have the space to take the fruit…..I would

have felt better if he’d been honest....I would have trusted him more”

These comments affirm the notion that the honesty and completeness of information aid

the relationship, and therefore enhance relationship quality. Of interest was the

discussion relating to communication‟s influence on trust and, therefore, relationship

quality. A representative comment included:

Participant 9:

“ We just entered into a 5 year contract with these guys and it is fabulous…..it’s a

small company and the GM (General Manager) popped around the other morning for

breakfast…..it was great to be able to talk with him…I have a good feeling about these

guys and I really trust them to do what is best for me”

This comment relates to the concept of trust and that communication positively

influences trust, which is determined as a measurement of relationship quality.

Participant 2:

“ They had the winemaker come over……then the viticulturalist……then the grower

liaison…..they all seemed to be saying different things………confused the hell out of

me…I don’t know how much I trust these guys are all telling the truth”

49

Daft & Lengel (1984), comment that certain modes of communication provide richer,

more complete information. Therefore, it was of interest to observe, in the causal stage

of the study, how specific modes of communication influence relationship quality.

As a result, comments regarding communication modality were linked to the creation of

contract and further linked to trust. Such a representative comment included the

following:

Participant 4:

“ The last guys that set up the contract were great…..they just came to the house (the

house of the grape grower) and we discussed it and we liked it so we signed……when it

finished (the contract) we changed to a different winery….they just emailed us their

terms and asked us for our terms….would have been better if they just came and talked

to us”

Participant 7:“ Would’ve been better to work out the contract face to face than over the

phone or the fax”

These comments are linked to the nature of the contract creation and allude to the

notion that face to face communication, as opposed to electronic computer modes,

created more trust in the winery and more satisfaction. This is also links to Daft and

Lengel‟s (1984) discussion that rich modes of communication (i.e. face to face) are

better than less rich modes. The comments are also linked to the formality of

communication as proposed by Mohr & Nevin (1990) and Mohr et al. (1996) where

written modes of communication create more trust as opposed to word of mouth modes

of communication. It appears that these comments from growers seem opposed to the

views expressed in the literature. Thus, it can be surmised that formal communication

decreased trust and satisfaction.

Further comments were made by respondents with respect to the bi-directionality of

communication. These comments included:

Participant 11:

“feedback from the winery was good”

Participant 6:

“winery talked to us a lot…..but didn’t do much listening…..they weren’t interested in

what we had to say”

50

Participant 10:

“they (winery) did all the talking…….rarely listened to what I had to say…..it didn’t

like this….made me feel like what I had to say didn’t mean anything”

Participant 3:

“ they (winery) never cared about what we said….i didn’t trust them…..yeah that didn’t

make me feel good”

Participant 7:

“We never talked to the winery…….they told us what they wanted and we gave them the

grapes…..i didn’t really care as long as I got paid”

Participant 12:

“Just gave us MOG and brix and other measurements………didn’t say much until

picking….didn’t bother me too much”

Participant 10:

“gave us the specs (grape specifications) and we did the spray diary (which catalogues

the spraying of chemical for export requirement) and that was it…..didn’t ask them

much

Participant 1:

“winery talked a lot…..we only contacted them a few times…..it did feel good….them

checking up on us’

These comments allude to the fact that the winery was producing much of the

communication. While respondents were given information regarding grape

specifications and the use of spray diaries to record the spraying of chemicals for wine

export requirement, the respondents‟ communication input appeared to be minor. It can

therefore be surmised from the comments that the communication in the relationship

was almost exclusively being transmitted from the winery and that bi-directionality of

communication was not evident. In addition to this, there was no decrease in trust or

satisfaction from respondents. Therefore, the uni-directional communication from the

winery positively influenced trust and satisfaction

51

Proposed by Mohr & Nevin (1990) is the concept of non-coercive communication

attempts. This concept is based on the premise that communication can indirectly make

an actor take a course of action, without being directly asked to do so. It can be

surmised that the characteristics (for example, demeanour, wit or charm) of the actor

transmitting the information can make another actor take action without specifically

being asked to do so. A representative comment by a respondent alluded to this notion:

Participant 3:

“…..he’s (the winery representative) a big, imposing guy whom I’ve never trusted…I

didn’t want to piss him off…..he didn’t ask us to complete the survey (a survey

regarding the vineyard details)…but I did I because I didn’t want to piss him off….I

didn’t want to do it otherwise”

This comment alludes to the notion of non-coercive communication attempts whereby

this participant performed a task, without specifically being asked, because of wanting

to please the winery representative (a person that was not trusted); however, it appears

that the action did not create satisfaction. Non-coercive communication attempts are an

effect of communication whereby without an explicit instruction to an actor, the actor

obeys by the communication transmitter‟s wishes due to factors such as intimiation,

reputation and body language affets (Mohr & Nevin, 1990). In term of the respond‟s

comments, it can be concluded that the non-coercive communication attempts

negatively influenced trust and satisfaction.

Much information that was gleaned from the discussion related to relational norms that

affect the grape growers‟ perspective of the relationship. Many of the comments were

centred on communication and how its various aspects affected trust and satisfaction.

However, discussion also occurred regarding how industry issues were affecting

participants‟ perceptions of the relationship.

Many comments were made by respondents that the wine industry was suffering

economic hardship. The hardship manifests itself in many ways, particularly in the use

of power. Such comments included the following:

52

Participant 2:

“We know times are tough…..we were around in the 80s (difficult period) but they

(winery) shouldn’t treat us as though we are stupid……I really don’t like it”

Participant 3:

“There are too many companies (wineries) just squeezing us too hard…..it’s difficult

(the current industry scenario) but they could be a little more honest….now I don’t trust

them as much”

Participant 10

“ the wineries sometimes take advantage of the fact that we have no alternative market

for our grapes…also because of oversupply of grapes..”

Participant 11

“ there are often threats of rejection of our grapes “

These comments illustrate that participants have knowledge of issues related to

industry oversupply and they believe that some wineries were exploiting the oversupply

of grapes scenario to better suit their circumstances. This may be a result of a power

asymmetry in the relationship which favours the winery, and this coercive power is

leading to decreased trust as proposed by Morgan & Hunt‟s (1994) extended KMV

model of relationship marketing. Furthermore, the climatic conditions that lead to

power asymmetry are evident in their assessment of grape quality and lead to the

rejection of the grapes. This action highlights the notion that wineries have the greater

power in the relationship and wish to take whatever action is necessary to maintain it, as

discussed by Cox et al. (2001). This scenario leads to conflict and disturbance, as

discussed by Gaski (1984) and Seyed- Mohammed & Wilson (1990). It can be surmised

that this power imbalance, and the resulting conflict, diminished relationship quality.

Specific comments were specifically made that the coercive power of wineries was

diminishing the level of trust and satisfaction in the relationship. This concept is evident

in comments whereby the participant mentioned that “now I don‟t trust them as much”

and “I really don‟t like it”. In the light of these comments, it can be surmised that power

is negatively influencing trust and satisfaction.

53

3.8 Exploratory research findings and relevance to literature The exploratory research results have uncovered numerous dimensions of relational

norms and relationship quality between grape growers and wineries. The aim of the

exploratory research study was to provide a result which, in view of the literature,

would allow for the creation of a conceptual model for testing via quantitative methods

in the causal stage of the study. Therefore, a discussion of the results and the literature

follows.

3.8.1 Research results on communication modality and relevance to literature

and hypothesis development

Discussion in the literature regarding communication modality suggests that face-to-

face forms of communication positively influence satisfaction (Daft & Lengel, 1984).

Mohr & Nevin (1990) and Mohr et al. (1996) further discuss frequency, which is

manifest in modality, as influencing satisfaction; however, Mohr & Nevin (1990) do not

distinguish between the modalities of communication and frequency. The results of the

exploratory study suggest that face-to-face or direct modes of communication positively

influence trust and satisfaction, while non face-to-face (or indirect) modes decrease

trust and satisfaction. However, the results do not distinguish whether they only

influence trust or satisfaction and therefore, relationship quality. Thus, in light of the

results and the literature, the following hypotheses were formulated:

H1. Face to face (direct) modes of communication positively influence trust

H2. Face to face (direct) modes of communication positively influence satisfaction

H3. Non Face to face (non direct) modes of communication negatively influence

trust

H4. Non Face to face (non direct) modes of communication negatively influence

satisfaction

3.8.2 Research results on communication directionality and relevance to

literature and hypothesis development

The literature defines communication as bi-directional; thus communication flows in

both directions i.e. from buyer to supplier and from supplier to buyer (Mohr & Nevin

54

1990; Mohr et al. 1996). Mohr & Nevin (1990) comment that bi-directionality does

influence satisfaction but did not test the effect of directionality on satisfaction. The

exploratory research results of this study suggest that bi-directionality of

communication is minimal in the relationship and is uni-directional from the source of

the buyer (winery). The results do suggest that the uni-directional communication does

positively influence participants‟ trust and satisfaction in the relationship and therefore,

relationship quality. In view of the literature and the research results, the following

hypotheses are posited:

H5. Uni-directional communication (feedback) from the winery positively

influences trust.

H6. Uni-directional communication (feedback) from the winery positively

influences satisfaction.

3.8.3 Research results on non-coercive communication attempts and relevance to

literature and hypothesis development

Non-coercive communication attempts is an element of collaborative communication as

posited by Mohr & Nevin (1990). Mohr & Nevin (1990) discuss how it affects the

relationship, but did not observe how the dimension affects trust or satisfaction and

merely combined the notion with other elements of collaborative communication in a

summated scale. The results of the exploratory study suggest that non-coercive

communication attempts negatively influence trust and satisfaction and as a result,

relationship quality. Therefore, in view of the literature and the research results, the

following hypotheses are posited:

H7. Non-coercive communication attempts from the winery negatively influence

trust

H8. Non-coercive communication attempts from the winery negatively influence

satisfaction.

55

3.8.4 Research results on communication formality and relevance to literature

and hypothesis development

Mohr & Nevin (1990) discuss the concept that formality of communication aids the

relational partner in clearly defining and understanding what is expected in the

relationship. While not directly testing the effect of formality of communication on trust

and satisfaction, Mohr et al. (1996) do comment that it has a positive effect on

satisfaction but do not comment on its effect on trust. The result of the exploratory

study suggests that formality negatively influences trust and satisfaction, which is in

conflict with the discussions of Mohr & Nevin (1990) and Mohr et al. (1996).

Therefore in view of the literature and the research results, and particularly in light of

the discussion above where communication from the winery seems uni-directional, the

following hypotheses are posited.

H9. Formality of communication from wineries negatively influences trust

H10. Formality of communication from wineries negatively influences satisfaction

3.8.5 Research results on power asymmetry and relevance to literature and

hypothesis development

The literature suggests that power asymmetry affects the relationship and creates

disturbances and conflict, and decreases the level of trust and satisfaction in the

relationship (see Cox et al. 2003, Gaski, 1984, Seyed Mohammed & Wilson, 1990).

The results of the exploratory study suggest that a strong power asymmetry (favouring

the winery) exists in the Australian wine industry. The results illustrate that this power

asymmetry affects the level of trust and satisfaction in the relationship. Therefore, in

view of the literature and these results, the following hypotheses are posited:

H11. A power asymmetry favouring the winery is decreasing grape growers‟

perception of trust in the relationship.

H12. A power asymmetry favouring the winery is decreasing grape growers‟

perception of satisfaction in the relationship.

56

3.8.6 Relationship quality and relevance to research results

The literature on relationship quality does not define a clear measure of the construct.

For example, Crosby et al. (1990), Wray et al. (1994), Kim & Cha (2002) and Kim et

al. (2006) for instance, operationalised the relationship quality construct as indicative of

the level of satisfaction. Others such as Leuthesser (1997), Dorch et al. (1998), Naudé

& Buttle (2000) and Parsons (2002) discuss relationship quality‟s relevance to trust,

satisfaction, commitment, opportunism and customer satisfaction. The framework

proposed for this study defined relationship quality as a measure of trust and

satisfaction; however, the literature does suggest that satisfaction positively enhances

trust (Mackenzie & Hardy, 1996) with Geyskens et al. (1999) arguing that if the

channel members are highly satisfied, the partners believe them to be more trustworthy

and Batt (2003, p 69) stating that “…satisfaction with an exchange will lead to some

initial trusting behaviours, but as satisfaction increases, trust will increase”. The

literature does suggest that satisfaction positively influences trust, but no concrete link

was found in the exploratory study linking trust and satisfaction. However, the

following comment was made by one respondent:

“they (winery) never cared about what we said…I didn’t trust them…..yeah that didn’t

make me feel good”

This comment seems to suggest that trust leads to a sense of feeling good, or

satisfaction. However, this link between the two constructs was not evident in other

participants‟ discussions. The results and the hypotheses formulated suggest that trust

and satisfaction exist in the relationship (and are influenced by the various

communication elements and power) and therefore allow for an observation of

relationship quality, but there appears to be no link between them. In view of the

exploratory study results, no link between trust and satisfaction can be said to exist.

3.9 Exploratory study research objectives overview The exploratory study had two research objectives. These were:

1. to uncover the extent to which Mohr & Nevin's (1990) collaborative

communication elements affects wine grape growers' perceptions of relationship

quality; and

57

2. to identify the extent to which power asymmetry is affecting grape growers‟

perceptions of relationship quality.

In relation to the first of these, it was found that collaborative communication affects

trust and satisfaction and therefore relationship quality. In relation to the second of

these, it was uncovered that power asymmetry in the relationship, favouring wineries,

affects growers‟ perceptions of relationship quality.

3.10 Limitations of the exploratory study The exploratory study provided information regarding the relationship that grape

growers have with wineries. The information was used to validate a theoretical model

(Figure 3.1) of grape grower perceptions of relationship quality in the Australian wine

industry. However, the exploratory study has a key limitation. The participants‟

businesses were located in South Australia and Victoria and, while these states

comprise the major grape growing areas of Australia, other states in Australia contain

other grape growing areas that were not represented by participants for the exploratory

study sample. Furthermore, participants were located in only five different wine regions

and 13 grape growers participated in the study, which is a relatively small sample

(Malhotra et al, 2006). However, the size of production and quality of production of the

participants in the exploratory study do allow for some generalisations to the Australian

grape growing industry.

Furthermore, this thesis has employed a mixed method approach, and as the exploratory

phase of the study is smaller in size (number of participants and scope of analysis) than

the causal and descriptive stage, this research stage is justified in terms of size, unlike a

triangulation study where both quantitative and qualitative methods are similar or equal

in size (Cavana et al, 2001; Cresswell, 1994). Similar studies in these areas of research

have also employed similar participant sizes in the qualitative phases (Hobley, 2007;

Plewa, 2008).

3.11 Hypothesised model The preceding literature and research results have provided many hypotheses for

examination in the causal stage of the study. These hypotheses are combined

graphically into a model which is shown in Figure 3.1

58

Figure 3.1: Conceptual model of grape grower perceptions of relationship quality

in the Australian wine industry

59

3.12 Alternative model In Chapter 2 it was shown that relationship quality was measured as a multi-

dimensional, higher order construct consisting of trust and satisfaction. Authors such as

Crosby et al. (1990), Dorch et al. (1997), Kim & Cha (2002) and Kim et al. (2006)

empirically tested relationship quality, mostly via SEM and other multi-variate

regression techniques, using trust and satisfaction as separate constructs, and they

discuss whether higher levels of trust and satisfaction in the model correspond with

higher levels of relationship quality. This proposition has been used as the basis of the

exploratory study, as participants were asked how they perceived the various

dimensions of collaborative communication and power asymmetry based on trust and

satisfaction.

However, Scheer & Stern (1992) and Leuthesser (1997) empirically tested relationship

quality as a uni-dimensional construct whereby the construct of relationship quality

consists of latent variables of trust and satisfaction. SEM literature discusses whether

this alternative (or 2 step model estimation) can be performed in order to observe which

model best fits the data concerned (Joreskog & World, 1982; Anderson, & Gerbing,

1988; McDonald & Ho, 2002). In this instance, it would be of interest to observe a

model which estimated relationship quality as a uni-dimensional construct as opposed

to a multi-dimensional one, thereby satisfying a theoretical and methodological

concern.

As such, the constructs exhibited in Figure 3.1 would directly affect relationship quality

in the alternative model as opposed to the multi-dimensional effect shown in Figure 3.1.

The hypotheses would remain the same, although each independent variable in the

model (i.e. power, collaborative communication elements) would affect the singular

dependent variable (i.e. relationship quality). Furthermore, the results of the exploratory

study showed that each element of collaborative communication and power had the

same effect on trust and satisfaction and would therefore affect relationship quality the

same as the uni-dimensional construct consists of the two factors. For example, power

was shown to negatively affect trust and negatively affect satisfaction in the literature

review and in the exploratory study. Therefore, in the alternative model, power would

negatively affect relationship quality, as relationship quality consists of trust and

satisfaction. Thus the alternative model hypotheses would be:

H1a. Face to face (direct) modes of communication positively influence


60

H2a. Non face to face (non direct) modes of communication negatively influence


H3a. Uni-directional communication (feedback) from the winery positively

influences relationship quality.

H4a. Non-coercive communication attempts from the winery negatively influence


H5a. Formality of communication from wineries negatively influences relationship

quality.

H6a. A power asymmetry favouring the winery decreases grape growers‟

perception relationship quality.

Figure 3.2 graphically illustrates the alternative model.

61

Figure 3.2 Alternative model based on uni-dimensional definition of relationship

quality and grape grower perception of collaborative communication and power

asymmetry

62

3.13 Chapter conclusion This chapter has discussed the exploratory study, including the qualitative study

methodology, the findings and hypothesis development, and has concluded with a

conceptual model derived from the hypotheses. The next chapter will discuss the

descriptive and causal research methodology, which will be used to test the conceptual

model and its various hypotheses.

63

Chapter 4: Descriptive and Causal Research Methodology

4.1 Chapter outline This chapter outlines the methodology employed in the descriptive and causal research

stages of the study. The chapter begins with a discussion of how the data was collected

during these research stages, and the techniques used to analyse the resulting data.

The chapter summarises the scale item measures used in the descriptive and causal

research stage and gives a discussion on the source of the scale items. The statistical

procedures used to analyse the data are presented, and the chapter concludes with a

summary of the methodology.

4.2 Quantitative research methodology design The research hypotheses and conceptual model that were devised in Chapter 3 are

tested by the methodology outlined in this chapter. This study employed a positivist

epistemological perspective whereby a scientific, validity approach was used to test the

hypotheses and the model devised in Chapter 3 (Wacquant, 1992; Cohen & Maldonado,

2007). To gain the descriptive and causal results, structured equation modelling was

used to test the model and thereby confirm the hypotheses (Hair et al., 2006). As such,

a quantitative research method was employed.

A survey was used to acquire the descriptive and causal information from the sample

population. The survey method was an appropriate means to collect the large numbers

of responses required for hypothesis testing, is simple to administer, and relatively

undemanding for the respondents to complete (Malhotra et al., 2006; Hair et al., 2006).

The survey instrument consisted of a structured questionnaire with questions placed in

a predetermined order. The questions posed to the respondents were, in the main,

quantitative in nature and were devised to gain the information required to test the

hypotheses in the causal stage of the study, but also to validate the study through

descriptive statistics (discussed in Chapter 5) (Malhotra et al., 2006).

64

4.3 Data collection method Preceding the creation of the questionnaire instrument, a large amount of time was

invested in examining the academic and wine industry trade literature and the

exploratory research results to ensure that the developed scale item measures were

suitable to measure what was required in the descriptive and causal stages of the study.

The constructs (i.e. collaborative communication elements, power and trust and

satisfaction) depicted in Figure 3.1 were operationalised in the questionnaire using

multiple measures that had been utilised in previous studies.

The scale items in the survey were modified numerous times to improve the efficacy

and content of the questions used. This process went for five rounds so that each

questionnaire item was clear and easy to understand by the intended respondent. The

questionnaire was pilot tested on 15 respondents to gain insight into whether the

respondents understood the questions and could successfully complete the

questionnaire. The small number of pilot test respondents was due to the sample

population size (4500- 8000 grape growers in Australia), and as pilot study respondents

were precluded from the main data collection phase for validity reasons, a large pilot

study response would have restricted the number of potential respondents in the final

sample.

On completion of the questionnaire, the study investigator met with the pilot test

respondents and discussed each questionnaire item, asking if they understood what the

question was asking of them (i.e. was it easy to understand) and querying if the

question could be posed a different way. The questionnaire contained 63 questions

which gave a pilot study respondent ratio to scale items of 4.2: 1 which is considered

acceptable according to Cavana et al., (2001). In addition, the questionnaire was

examined by wine industry professionals, including heads of grape grower associations

and wine industry experts, to gain their opinions of the efficacy of the scale items and

the overall effectiveness of the questionnaire. Each of the wine industry experts

examined the questionnaire and, on completion, was queried as to whether the

questions were appropriate for the intended sample frame (i.e. would the questions be

understandable to grape growers) and if any questions should be discarded or new

questions devised. This second stage was of critical importance as it improved the first

section of the questionnaire related to specific questions about grape contracts such as

price per tonne, respondents‟ vineyard acreage, and overall crop price. The wine

65

industry professionals generally commented that the contents of the questionnaire were

sound.

4.3.1 Quantitative study sampling procedure and sample size

The survey population for the quantitative research study were contracted grape

growers. These were independent wine grape growers currently supplying wineries;

only wine grape growers who supplied wineries with grapes, as opposed to making

their own wine, were eligible to complete the questionnaire. In consultation with wine

industry experts, it was deemed appropriate to survey non-wine making grape growers

because including those that make and market wine may give distorted results. They

(wine making grape growers) would be in the business of making wine; as such they

may contract other grape growers to obtain grapes, and this may bias some of their

opinions regarding wineries and may affect their questionnaire results.

The survey population was from all states in Australia where wine grapes are grown

and included Western Australia, South Australia, Victoria, Tasmania, New South

Wales and Queensland and the survey was administered from March to July 2009.

Separately from the data collected to test hypotheses and the conceptual models, the

survey was also designed to obtain information regarding the business and the

demographic characteristics of respondents, and this was used to cluster respondents by

their responses (Malhotra et al., 2006; Hair et al., 2006).

The survey population included independent wine grape growers who could be

classified according to the varying natures of their business structures. The

respondents‟ business structures included large investor-owned vineyards, managed

investment schemes, small part time producers, and long term, grape growing families.

Therefore, the business structure falls in line with the grape grower classifications of

PIRSA (Primary and Resources South Australia) wine divisions (PIRSA, 2006).

The census, and therefore the size of volume grape growers in Australia was also

determined and played a part in determining the representativeness of the primary study

respondents in relation to the target population. However, statistics on the number of

grape growers in Australia are not accurate. The Australian Bureau of Statistics (ABS)

(2005) and Hobley (2007) reported that 8,347 individual grape growing establishments

exist in Australia for the purpose of winemaking, drying and fresh fruit consumption. A

breakdown of establishments that specifically grow grapes for wine making was not

66

available; however, Hobley (2007) reported that 90% of the grape growing

establishments grow grapes for the production of wine. Furthermore, the study sample

frame included independent wine grape growers who supplied wineries and the ABS

figures do not discriminate between independent grape growers and wineries that grow

grapes for their own wine production. Further investigation by the author revealed that

the recent industry economic downturn had reduced the number of establishments and

that the number of grape growers in the sample frame may be as low as 4500

(Mckenzie pers. comm., 2009).

The wine grape industry does not have a database of contact information for the sample

population and thus wine industry bodies such as wine grape grower associations and

private companies that have contact with grape growers were used to obtain grape

grower information. Therefore, by necessity the descriptive and causal stages of the

study (relying on quantitative data) had to rely on a non-representative sample.

However, the descriptive and causal research studies wished to achieve

representativeness of the wine grape growers in terms of the size and geographical

location of the businesses, particularly in terms of the state where production was made.

Wine grape grower associations and private companies were willing to provide

assistance in terms of giving direct access to their grape growers via electronic

distribution of the questionnaire. The wine grape grower associations (e.g. in the

Murray Valley, Riverina, Barossa, McLaren Vale, Adelaide Hills, Tasmania, King

Valley, Granite Belt and Hunter Valley) assisted by electronically distributing the

survey to their constituents and provided comments of endorsement. Private companies

(including grower liaison companies and large wineries) also provided access to their

growers using a similar method, including wine industry news services. Assistance

from the associations and companies provided a good regional, state and production

(quality of grape production focus) representation in the final sample. The

representativeness of the final sample is discussed in Chapter 5. The survey was

completed in less than four months.

A large number of respondents was desirable due to the large number of relational

variables devised in the conceptual model and the use of multivariate statistics such as

structural equation modelling (Hair et al., 2006). In general, the number of respondents

required depended on numerous factors such as:

the level of precision of results (confidence interval);

67

the acceptable risk in predicting the level of precision;

time and cost constraints;

size of the actual population; and

variability in the population.

(Ticehurst & Veal, 2000; Cavana et al. 2001; Malhotra et al. 2006)

4.3.2 Administration of survey instrument

Two methods were used for the administration of the survey instrument to grape

growers. The main survey methodology was via online administration. The survey was

uploaded onto an online survey administration portal which allowed the respondents to

complete the survey via an internet web browser. The web administration of the survey

instrument was deemed an efficient and cost effective way of accessing respondents

due to the geographically disparate nature of respondents‟ places of residence and the

cost issues related to paper administration of the survey where paper surveys are

delivered to the respondents and then self-completed and returned. The online methods

reduced the amount of postage and paper expense that would normally be associated

with paper administration. Respondents were also able to complete the questionnaire at

their own pace and convenience and were able to save their responses online for later

completion.

The two methods of administration were as follows:

1. Firstly, the grape grower associations and private grower liaison companies were

sent a web URL link to the web site of the survey. The associations and companies then

sent a group email to the grape growers on their databases and the grape growers were

then able to complete the survey.

2. Secondly, the details of the web URL and a description of the study were posted on

the web site of grower liaison companies and on the web site of various Australian wine

industry news web sites, frequently viewed by grape growers.

A prize of $2000 of viticultural services was offered to respondents to motivate them to

respond. This was deemed necessary to increase the response rate due to the lengthy

68

nature of the survey and due to the prevalence of survey fatigue in the Australian wine

industry, as it is widely surveyed.

This type of approach was deemed appropriate by similar studies in the Australian wine

industry (Boyce & Neale, 2006; Hobley, 2007).

Table 4.1 illustrates the grape grower associations and private organisations that

provided access to, and assistance in, obtaining responses. This table indicates that

grape grower associations from all wine grape growing states (i.e. Tasmania, South

Australia, Victoria, New South Wales, Queensland) assisted in administering the survey

and therefore in making responses from all states achievable

Table 4.1: Grape grower associations and private organisations that provided

access to respondents

Grape grower associations Private companies

Adelaide Hills Wine Region Inc. Morton Blacketer

Coonawarra Vignerons Association Davidson Viticulture

Goulburn District Vignerons Association Constellation Wines Australia

Hunter Valley Wine Industry Association Orlando Wyndham

King Valley Vignerons Fosters Wine Estates

Barossa Grape and Wine Association Wine Biz Online

Wine Industry Tasmania Australian Grape Grower and Winemaker

Great Southern Wine Producers Association

Granite Belt Wine Growers Association

Swan Valley & Regional Winemakers' Association Inc.

McLaren Vale Grape, Wine & Tourism

Murray Valley Winegrape Growers Association

Riverina Wine Grape Marketing Board

69

While a non-probability sample existed for the study, efforts were taken to reduce the

level of self-selection bias associated with web-based surveys (Zikmund & Babin,

2007) by observing whether the respondents matched the sample frame. Associated

with the $2000 prize, respondents had to enter their personal business details (name of

person completing the survey, telephone number, business address). A random sample

of 30 respondents were contacted from the list and asked whether they had completed

the survey. In all 30 cases, the respondents matched the characteristics of the sample

population thereby allaying the problems associated with self-selection bias and false

representation.

4.3.3 Questionnaire design

Numerous issues were considered when devising the design of the questionnaire. The

most important factors were, as discussed by Cavana et al., (2001), the research

objectives, the sample size, the method of distribution, the sample frame, the data input

method and the type of analysis. The questionnaire contained two types of questions.

Unstructured questions, which were open-ended in nature, were mainly used to gain

information regarding the respondents‟ business details; structured, specific response

questions were used mainly in the form of multiple choice and scaled questionnaire

items.

The questionnaire used two types of scales. Firstly nominal scales were used to

describe the differences in a characteristic of the respondent, for example, 1= small

winery contracted, and 2= medium winery contracted. However the majority of scales

used were interval in nature; this was required for the multivariate analysis (Hair et al.,

2006). In the interval scale questions respondents were asked to indicate on a seven

point Likert scale their agreement or disagreement with a statement. A seven point

Likert scale was used in the majority of the scale items that were adapted for this study

because it was important to maintain consistency (Hair et al., 2006).

The questionnaire was divided into three sections. The survey commenced with a brief

introduction to the survey, explaining who was eligible to complete the survey, and an

outline of the prize incentive.

The questionnaire asked the respondents to focus on the most important relationship

they had with a winery when answering the questionnaire items. This was directed in

consultation with wine industry experts because respondents may have had numerous

70

relationships with different wineries and it was not deemed appropriate for them to

answer the questionnaire items for each relationship as that would have taken a great

deal of time to complete and cause fatigue (Hair et al., 2006). For example, a

respondent may have their grapes contracted to three wineries and it would have been

an onerous task for them to complete three questionnaires, one for each relationship, so

they completed one questionnaire focusing on their most important relationship. The

ramifications of this are discussed further in Chapter 6 and Chapter 7.

Section 1 contained questions where growers had to discuss the business details of the

contract they had with a winery. These questions included the length of the contract

with the winery, how many tonnes of grapes were supplied, the dollar amount of the

grapes supplied, the price per tonne of the grapes, and whether the respondent supplied

any other wineries. The questions in Section 1 allowed for the comparison of responses

based on the contracting relationship that existed between the respondent and a winery.

The responses also allowed for variables that could be used for clustering purposes.

Section 2 contained the bulk of the questions regarding the research hypotheses and

objectives. Section 3 was designed to gain information relating to the details of the

respondent‟s business. This included questions regarding the size (in acres) of the

vineyards of the respondent, the years the respondent had been producing grapes and

been in the grape growing business, the number of people who worked for the

respondent‟s business, the wine region the respondent‟s business was in, whether the

respondent was contracted to a winery in terms of the winery size (small, medium or

large), and the ownership of the winery (publicly or privately owned). As in Section 1,

Section 3 responses could be used as clustering variables for later analysis. The

questions in Section 3 could also be used to observe the location, size and general

business “demographics” of the respondents.

4.3.4 Modification of questionnaire to online format

A vast amount of time and effort was devoted to modifying the questionnaire into the

online format. While the efficacy and validity of the questionnaire items was tested in a

paper format (via internal modification and testing on a pilot sample of grape growers

and wine industry experts), the online modification was a further process that was

necessary in order for the instrument to be easy to complete and understandable by the

sample population.

71

An online survey provider was hired to host the questionnaire. The service they

provided included hosting the web pages that contained the questionnaire and providing

mechanisms for collating the data and downloading it into data analysis software. The

provider also had mechanisms to ensure data protection and minimise fraud. The

questionnaire was uploaded, in an electronic form, into the provider‟s web site. From

there it was modified to be aesthetically pleasing and easy to read. The questionnaire

was aesthetically modified via html to change the appearance and size of words, by

making them bold or underlining them, or by increasing or decreasing font sizes in the

questionnaire items. The Likert scales were also modified to highlight terms (for

example, AGREE) and to fit the scale into a web page. This process was performed to

make the items easier to read.

When modifying questionnaire items and the scale to fit into a computer screen,

consideration was given to the size and resolution of the screen. Time and effort was

spent on this issue as it was deemed important that survey response errors, such as false

or non-responses, be minimised (Ritter & Sue, 2007). For example, if a respondent had

a computer with a large screen (e.g. 21 inches in diameter) with a high resolution (e.g.

800-1000 horizontal pixels), the words on the screen would be in a very small font and

the Likert scale would be long on the screen. Conversely a respondent with a small

screen and a small resolution (12 inches in diameter and 400-600 horizontal pixels)

would view the questionnaire in a large font size and the Likert scale would fall off the

screen requiring the respondent to scroll across to fill in the scale; this would have

created respondent fatigue and possibly created inaccurate results (Ritter & Sue, 2007;

Zikmund & Babin, 2007).

After consultation with an IT expert (Matthews pers. comm., February 2009), it was

found that most respondents would have a screen approximately 14-17 inches in

diameter with a 600-800 horizontal resolution .Therefore, the font size of questionnaire

items and scale length was modified with these parameters in mind so that the font was

large enough to read and the scale did not require scrolling across the screen to

complete.

The online questionnaire was divided into six html web pages so that questionnaire

items and scales could be displayed properly, in a vertical fashion. The respondents did

not need to scale the screen vertically to a great degree. Utilising six pages allowed the

respondent to avoid becoming too fatigued by replying to questions with too many

items on one html page. After completing each page, the respondent clicked a button (at

72

the bottom of the page) and was directed to the next html, and so on, until all the pages

were completed.

After all aesthetic modifications were completed, the online questionnaire was further

pilot tested on five grape growers who were then met by the investigator. As in the pilot

testing of the initial phase of the questionnaire, a small number of pilot testers were

used so as not to exclude a large number of potential respondents from the final

questionnaire response. The investigator questioned the respondents to gauge whether

they believed the layout of the questions was appropriate and easy to understand, and

whether the size and style of the questionnaire items and scale was appropriate for their

computer screen.

After this process was completed, the questionnaire was ready for deployment to the

sample population. A copy of the questionnaire is shown in Appendix 1. This copy is a

replication of the online version in a paper form, and as such, the representation is not

identical as certain elements of the online format (drop down menus, borders) cannot be

reproduced on paper. However, its question and scale content is identical.

4.3.5 Protection of questionnaire information against online fraud

Time and effort were taken to mitigate against the effects of online fraud. The main

concern was mitigating against accidental multiple responses and the hacking

(reprogramming of the questionnaire or questionnaire responses for fraud reasons),

particularly in relation to false responses to gain further entries into the questionnaire

incentive (Wright, 2005). The protection of the questionnaire and respondent

information from fraud was performed in two ways.

Firstly, the online survey provider had an IP (internet protocol) address registration

system. All internet enabled computers have an IP address and the provider logged each

IP address of the respondents who completed the questionnaire. This allowed

respondents to finish completing the questionnaire at a later date if they did not fully

complete it in one attempt by returning to the question they last completed. The IP

logging also allowed the rejecting of respondents if they had already completed the

survey. If an attempt to complete the survey a second time from the same computer was

made, the respondent was greeted with a message stating they had already completed

the questionnaire and was denied further access to it. To combat the issue that

respondents could complete the questionnaire at an additional time using another

73

computer, which therefore had a different IP address, respondents were asked to give

their contact details to win the prize; therefore, it was easy to observe if they had

completed the questionnaire a second time, and all responses that did not contain

contact details were rejected from the final valid response set. These measures were

deemed appropriate to mitigate against such issues (Rolland & Prakash, 2005; Wright,

2005).

Secondly, to further combat fraud, the responses were manually screened by the

investigator to observe if any responses were too similar or contained information that

was irregular or blatantly incorrect (for example, an irregular response might state that

they had received $72,000 per tonne for their grapes, which is blatantly not possible).

All suspicious responses were deleted from the final valid set of responses. This

technique aided in combating “hacking” fraud by online miscreants.

This thorough process revealed that no intentional fraud occurred and, in all, 48

responses were deleted from the final valid response set due to irregular or blatantly

wrong responses, no contact details being given, or due to suspicious responses, leaving

396 valid responses. No major fraud or hacking was encountered in the online survey

process. It could be reasoned that all rejected responses were due to mistakes and

confusion rather than wilful fraud.

4.3.6 Section 2: Scale items relating to research hypotheses

The focuses of the scale items in the questionnaire were to test the research objectives

and hypotheses. Section 2 of the questionnaire contained all the questions relating to

the research objectives and hypotheses.

The first questions in Section 2, regarding the frequency and mode of communication,

are show in Figure 4.1.

Figure 4.1: Questionnaire scale items regarding the mode of communication

For each of the following methods, over the 2009 Vintage growing season (August 08- May 09), please estimate the frequency (the number of times) with which the winery communicates with you via these various methods. Please type in the "number of times" as a number, e.g. "4" rather than "four". If you did not communicate via a certain method, please put "0"

74

Face to face interaction with winery people (number of times) (Required) Telephone interaction (telephone calls) with winery people (number of times)

(Required) Written letters and all written correspondence (non-electronic e.g. no email)

(number of times) (Required) Direct Email, from a winery representative to you (number of times) (Required) Seminars [e.g. Grower Days (winery - growers meetings)] (number of times)

(Required) General newsletters from the winery (number of times) (Required) Other (number of times)

The instructions accompanying the questions and the nature of the modalities of

communication were gleaned from scale items obtained from Mohr et al., (1996) and

adapted from Cannon & Homburg (2001). The specific modalities (e.g. computer,

seminars, etc.) were added after consultation with Australian wine industry experts as

the modalities needed to be relevant to the industry. Further scale items from Prahinski

& Fan (2007), Kwon & Suh (2004), Claycomb & Frankwick (2004), Morgan & Hunt

(1994) Redondo & Fierro (2005), Petersen et al., (2005), Lusch & Brown (1996) and

Heide & John (1992) were used as comparison scale items, mainly to gain examples of

the wording of questions in relation to communication. Formality of communication

was the topic of the next set of questions designed to test the research hypotheses. The

questions were as follows:

75

Table 4.2: Questionnaire scale times regarding the formality of communication

These questions are based on the Mohr and Nevin (1990) collaborative communication

elements. They were derived from that study and the Mohr et al.,(1996) study and

wording was adjusted to be relevant to the wine industry, and done in consultation with

The next set of questions in Section 2 related to the feedback produced from wineries.

A summated scale was used for these questions, based on the positive and negative

feedback obtained from the winery. The positive and negative feedback responses were

added together to produce a single scale item in the data analysis stage. Therefore, if a

respondent answered “1” for negative winery feedback and “4” for positive winery

feedback, the summate was “3” (i.e. -1+4=3). The scale items were derived from the

Mohr & Nevin (1990) and Mohr et al., (1996) studies, as the feedback elements are a

basis of collaborative communication.

The next few questions are regarding the feedback that the winery provides to the

growers.

Please indicate how strongly you agree on the following statements When working with this winery, formal communication channels are followed (i.e. communication is formal, regular and structured) versus casual informal, word –of-mouth modes).

The terms of our business contract with the winery have been written down in detail. The winery‟s expectations of us are communicated in detail. The terms of our business relationship with the winery have been explicitly put into words and discussed.

Information sharing on important issues has become crucial to maintaining this partnership.

We share a common, specialised IT software system dedicated to facilitate communication with the winery (e.g. Vine Access®).

Grower liaison committees, that communicate my issues and concerns with the large wineries, are effective.

76

Table 4.3: Questionnaire scale items regarding winery feedback

Please indicate by clicking the box that corresponds with your answer

How much negative feedback does this winery provide to you?

How much positive feedback does this winery provide to you?

The next questions in Section 2 detailed the non-coercive communication attempts.

Derived and adapted from Mohr & Nevin (1990) and Mohr et al., (1996) and

scrutinised by wine industry experts, the scale items were as follows:

Table 4.4: Questionnaire scale items: non-coercive communication attempts

In their interaction with you, the winery often tries to influence YOUR attitudes and

behaviours. Please estimate the frequency with which the winery‟s employees (e.g.

winemakers, grower liaison staff, viticultural staff) use the following methods to

influence YOU.

How frequently did the winery‟s employees make a recommendation that by following their suggestions, your business would be more profitable.

How frequently did the winery‟s employees ask you to perform a certain operation, but didn‟t say what penalty may occur if you didn‟t do what they asked.

How frequently did the winery‟s employees say you will be supplying grapes of a certain quality, but didn‟t give you specific information e.g. what crop level they would like, what spray regime they would like or other directions they would like you to take to grow those grapes.

The following division of Section 2 involved questions regarding trust. The scale items

for trust were based on the Kumar et al., (1995) dimensions of trust and were compared

and adapted using scale items from Walter et al., (2003), Bigne & Blesa (2003),

Kingshott & Pecotich (2007), Kwon & Suh (2004), Morgan & Hunt (1994), and

Petersen et al., (2005). The modified scale items were then scrutinised by wine industry

experts to enhance the validity of the items. They are as follows:

77

Table 4.5: Questionnaire scale items regarding trust

When things go bad, we believe that the winery will be ready and willing to offer us assistance and support. When making important decisions, the winery is concerned about our welfare.

When we share our problems with the winery we know that they will respond with understanding.

We can count on the winery to consider how its decisions and actions will affect us.

When it comes to things that are important to us we can depend on the winery‟s support. Even when the winery gives us a rather unlikely explanation, we are confident that they are telling the truth.

The winery has often provided us information that has later proven to be incorrect.

The winery keeps the promises that it makes to our business.

Whenever the winery gives us advice on our business operations, we know that it is sharing its best advice.

Our organisation can count on the winery to be sincere.

The next set of scale items were regarding satisfaction and were based on the scale

items from Kwon & Suh (2004) but were adapted and compared to scale items from

Walter et al., (2003) and Bigne & Blesa (2003), and scrutinised by wine industry

experts. The scale items were as follows:

Table 4.6: Questionnaire scale items regarding satisfaction

Please indicate how strongly you agree with the following statements:

We are very pleased with our working relationship with the winery.

Generally we are very satisfied, with our overall relationship with the winery.

The relationship our business has with the winery has been an unhappy one.

I am happy with the contract I have with the winery for my grapes.

78

The final set of questions in Section 2 of the questionnaire contains scale items

regarding power. The scale items were derived from Wilson & Vlosky (1998) and were

adapted for the wine industry context. They were considered appropriate by wine

industry experts. The scale items are as follows:

Table 4.7: Questionnaire scale items regarding power


We have to follow the winery‟s instructions or they will get their grapes from someone else.

We are expected to follow the winery‟s instructions.

We have influence over the winery‟s actions.

The winery can, if it wanted to, severely penalise us if we are uncooperative.

If we did not want to follow the winery‟s instructions or plans we could sell our grapes to another winery.

4.4 Data preparation and data analysis techniques The questionnaire data was compiled by the online survey web site company. The

survey‟s data was downloaded as a Microsoft Office Excel file and uploaded into the

statistical program SPSS (Statistical Package for Social Sciences). Prior to the Excel

file upload into SPSS, each question in the survey was entered into SPSS, thereby

allowing the uploaded Excel file data to correspond with the questions.

Upon completion of the upload of the Excel file into SPSS, the data was screened for

validity. Analysis was performed using descriptive statistics, such as means and

standard deviations, and graphically illustrated using box plots. Cases that contained

incomplete responses, or responses that were outside ranges or had means or standard

deviations that were not reasonable or believable, were deleted. The purpose of the

screening was not only to remove missing answers or implausible responses but also to

check the pattern of the missing data and why it was missing (Hair et al., 2006). A box

plot analysis showed the missing data to be random and less than 5% of the data points

(Hair et al., 2006).

A total of 444 returned responses were uploaded into SPSS and, following validity

screening (incomplete, blatantly wrong, somewhat suspicious responses), 48 responses

were deleted, leaving 396 valid responses.

79

4.4.1 Univariate Analysis

Univariate analysis was used to determine the frequency, mean and modality of the

descriptive variables in the data set. The descriptive variables were mainly contained in

Sections 1 and 3 of the questionnaire and were based on questions regarding the

demographics of the respondents‟ businesses and the details of the contracting

arrangement between the respondent and the winery, such as contracting dollar

amounts, price per tonne, length of contract, and volume of grapes supplied. The data

was analysed via the univariate statistics to determine the demographics and contracting

arrangements of the grape grower respondents with wineries. The univariate results are

tabled in Chapters 5 and 6..

4.4.2 Multivariate Analysis

Factors analysis, structural equation modelling and cluster analysis were the

multivariate techniques used in the analysis of the research results.

Firstly, factor analysis was performed on the constructs used in the study (e.g. power,

trust, satisfaction, collaborative communication dimension). Factor analysis is a data

reduction technique that investigates the relationships between scaled metric variables

and endeavours to understand the underlying factors (Malhotra et al., 2006; Hair et al.,

2006). Each factor is then extracted and if a dimension extracts on one variable, then

that variable is used as the sole variable in the analysis of that dimension. As such, the

principle component analysis was used to reduce and eliminate variables that did not

contribute to the factor, and confirmatory factor analysis (CFA) was used to validate

the measurement model (Hair et al., 2006). CFA is performed during the structural

equation modelling process via the statistical package SmartPLS to determine if the

scale items correspond to the latent construct. Cronbach‟s alpha reliability coefficient

was utilised to test the internal consistency of the model and composite reliability of the

measurement (Werts et al., 1974). The coefficient describes how well a group of items

focuses on a single construct with an index of 0.7 or higher considered preferable (Hair

et al., 2006). However, it is argued that composite reliability index is more “... reliant in

assessing convergent validity as it takes into account the relative weights of the

indicators of the latent construct while Cronbach Alphas assume equal weight” (Gyau

& Spiller, 2007, pg. 10). Convergent validity refers to whether the construct measures

80

what it is supposed to measure. This is performed by calculating the Average Variance

Extracted (AVE) which assesses whether the construct variance can be explained by the

indicators (Fornell & Larckner, 1981). The recommended smallest value is for each

construct to be at least 0.5, which means that the indicator explains at least 50% of the

variance (Bagozzi & Yi, 1988).

Structural Equation Modelling (SEM) is a statistical technique for testing and

estimating causal relations using quantitative statistical data (Hair et al., 2006). The

SEM process begins with the creation of a model based on the relevant academic theory

and supporting research (Hair et al., 2006). In the case of this study, the relevant

literature and the results of the exploratory research study provided the basis for the

theoretical models which are exhibited in Chapter 3.

The technique used to test the model was Partial Least Squares (PLS) structural

equation modelling. This technique, utilising SmartPLS software 2.01, allowed for the

understanding of the relationship between the latent variables and was considered

appropriate for the study due to the ability of PLS to handle structural equation

modelling of small sample sizes; it uses less strict distributional assumptions than

LISREL or AMOS would use (Chin, 1998; Joreskog & Wold, 1982; Ringle et al., 2005;

Gyau & Spiller, 2007). Effectively it is a prediction-oriented, variance based approach

to SEM (Liljander et al., 2009).

Confirmatory factor analysis (CFA) is also performed by SmartPLS while estimating

the model, thereby allowing a set of quality statistics (such as Cronbach Alphas, mean

and standard deviations) to be obtained. PLS is also a soft modelling form of structural

equation modelling which “…iteratively estimates the parameters of latent variables”

(Gyau & Spiller, 2007, pg. 9).

Under the soft modelling approach, there a two types of variables considered: the

manifest and latent variables. Simply stated, the latent variables were the constructs

identified in the literature, such as the collaborative communication elements (for

example, formality, direct and indirect communication, etc.), trust, satisfaction; the

manifest variables were the questionnaire items (scale items) used to test the latent

variables. In the soft modelling approach, manifest variables that do not make a

significant contribution to their respective latent variables; AVE, Cronbach Alpha, and

81

composite reliability were removed. The analysis is completed until all manifest

variables are significant. A bootstrapping technique was then performed to gain a T-

value for the paths between latent variables which allowed for significance testing of

the paths. A bootstrapping re-sampling of 500 cases was used as per normal with this

type of SEM (Gyau & Spiller, 2007).

The benefit of using PLS over other SEM techniques that use maximum likelihoods

(such as LISREL or AMOS) is that PLS can estimate a model when as little as two

manifest variables are used to measure the latent variable, in addition to the ability to

estimate models with small samples sizes and models that do not have strict

assumptions on residual distributions, such as this study (Dibben & Chin, 2005; Gyau

& Spiller, 2007; Herath & Rao, 2009).

The testing of the SEM was performed by evaluating the inner and outer models. The

outer model is evaluated by examining the individual item reliabilities‟ convergent

validity. Factor loadings of at least 0.4 are considered significant and retained in the

model (Hair et al., 2006; Gyau & Spiller, 2007). The internal consistency of the model

was calculated by appraising the Cronbach Alphas and the composite reliability of the

latent variables (Werts et al., 1974). A loading of greater than 0.7 from the Cronbach

Alphas and 0.5 for the composite reliability is acceptable (Werts et al., 1974, Hair et al.,

2006). The convergent validity of the latent variables is also measured by calculating

the AVE, with a minimum of 0.5 recommended (Bagozzi & Yi, 1988).

The inner model is evaluated via the discriminant validity which details whether each

latent variable is different from the other latent variables. To achieve this, a loading and

cross loading matrix was obtained. The loadings were the Pearson correlation

coefficients to their own latent variables. The loadings must be higher than the cross

loadings (Gyau & Spiller, 2007). Another technique for measuring discriminate

validity is by observing the square root of the AVE, which must be higher than the

correlation between the latent variable and the other latent variables (Chin, 2001).

Bagozzi (1984) suggests that the correlations between the different constructs in the

model must be smaller than 0.8. Table 4.1 illustrates a summary of the statistical

criteria for model estimation using PLS.

82

Table 4.8: Statistical criteria for model estimation via PLS

Statistical Criterion Acceptable Fit Author

Convergent Validity 0.4 or greater Hair et al (2006)

Average Variance

Extracted (AVE

0.5 or greater Bagozzi & Yi (1988)

Cronbach Alpha 0.5 or greater Cronbach, 1970; Gyau &

Spiller, 2007

Composite Reliability 0.5 or greater Werts et al, 1974

Discriminant Validity*

Less than 0.8 Bagozzi, 1994; Chin, 2001

*Correlation between the square root of AVE and correlation between constructs

4.5 Chapter summary This chapter outlined the design of the descriptive and causal research stage and the

methodology employed. Due to the nature of grape growing in the Australian wine

industry, a non-probability sampling technique was employed and quantitative data was

collected from grape grower respondents via an online survey method, with assistance

from regional grape grower associations and private companies that liaise with grape

growers. The questionnaire instrument contained scale items derived from the

marketing literature and these were modified for wine industry standards. Care was

taken to pre-test the survey on grape growers to obtain external validity, and also to use

wine industry experts to give opinions.

The chapter concluded with an outline of the statistical techniques used in the primary

research study, including a detailed discussion on structural equation modelling.

As discussed above, this chapter illustrated the methodology utilised in the primary

research study. The next chapter of the thesis, Chapter 5, discusses the results of the

descriptive research study, which includes the descriptive statistics of the respondents‟

business operations and their trading relationships.

83

Chapter 5: Descriptive statistics of respondents and

trading relationships

5.1 Chapter outline Chapter 5 commences with a discussion of the results obtained from the descriptive

stage of the study, i.e. the quantitative results from the questionnaire instrument. As

detailed in the chapter title, the chapter will exhibit the descriptive statistics of the

business relationship between the grape growers and the wineries, and descriptive,

business related statistics of the grape growers.

The statistics that are presented in this chapter are gleaned from sections 1 and 3 of the

questionnaire. As discussed in Chapter 4, section 1 of the questionnaire contained

items regarding the business relationships growers had with the wineries and included

questions relating to the number of years the grower was contracted to the winery, the

tonnes of grapes supplied and the value and price per tonne of those grapes, the size of

the winery and the type of ownership that the winery had, the region that the winery

was in, and the number of other wineries to which the grower supplied. Section 3 of the

questionnaire contained questions relating to the descriptive statistics of the grower

respondents and contained questions regarding the size of their vineyards, the number

of years the growers had been growing vines, the number of employees in the grape

growing business and the region in which the grape growing business was located. The

statistics from section 2 of the questionnaire relate to the estimation of the conceptual

models and their various hypotheses and are discussed in Chapter 6.

The main purpose of this chapter is to benchmark the respondents‟ responses against

other previous studies utilising the Australian wine industry and Australian wine

industry statistics, to observe if the respondents are representative of the sample

population. It was also of interest to examine if the grape grower respondents of this

study are representative of grape growers in the Australian wine industry.

Firstly, the descriptive results of section 1 of the questionnaire are discussed.

84

5.2 Section 1: Descriptive statistics of grower/winery relations.

5.2.1 Duration of relationship with winery

As discussed in 5.1, section 1 of the questionnaire detailed the business relationships of

the grower respondents. As discussed in Chapter 4, respondents had to focus on the

business relationship that was most important to them. The mean number of years that

growers had the relationship with the winery they were asked to focus on was 8.5 years

and a standard deviation of 8.37 (n=396). The cumulative frequencies, shown in Table

5.1 below, illustrate that most of the relationships (approximately 42%) were less than

or equal to five years. This length of relationship is supported by previous research on

the Australian wine industry that suggest that typical grape supply contracts are

between three and five years in length (Scales et al., (1995); Edmonds, (2000);

Anderson, 2001; Hobley, 2007). It can be concluded that the respondents of this study

are representative of the sample population in terms of the length of contract with

wineries. However, this study observed the “best” winery relationship from the

respondents‟ perspective. It would stand to reason that a “best” relationship would be

ongoing and have a longer length, but the economic turmoil in the industry may be

creating a situation where a “best” relationship is shorter rather than longer. However,

the concept of “best” relationships being longer is speculation, and industry upheaval

potentially may mean that any relationship is “best”. However, this is speculation.

Table 5.1: Years of contractual relationships between respondents and wineries

Years of

contract

Frequency % Cumulative %

5 167 42.2 42.2

6-10 105 26.5 68.7

11-15 71 17.9 86.6

16+ 53 13.4 100

5.2.2 Volume of grapes supplied to winery

A question was posed to respondents to determine the amount (tonnes) of grapes

supplied to the winery. The mean result was 214.4 tonnes with a standard deviation of

85

493.0 (n= 396) which illustrates that the statistics were highly distributed and the high

score for the mean result is manifest in the large number of respondents who supplied

more than 700 tonnes, as opposed to other volume categories such the 300 to 700 tonne

categories. Table 5.2 exhibits these results further and illustrates that the majority of the

respondents supplied less than, or equal to, 100 tonnes of grapes (67.7% of

respondents). This result is in line with that of Hobley (2007) who found that the

majority of grape growers supply less than 100 tonnes of grapes to a winery. Therefore,

it can be stated that the respondents of this study are representative of the sample

population in terms of the volume of grapes supplied to wineries.

Furthermore, it appears that the “best” relationship a respondent has with a winery

involves a smaller rather than larger volume of grapes and perhaps, though this is

speculation, smaller volumes may mean that “best” relationships with wineries are a

result of the production of quality, which results in smaller yields, as opposed to

quantity of grapes. On the other hand, as shown in Table 5.11, 53.3% of respondents‟

vineyard size were less than or equal to 25 acres, which would result in smaller

volumes of grapes being supplied.

Table: 5.2: Volume of grapes supplied to winery by grape grower respondents

Volume of

grapes (tonnes)


100 268 67.7 67.7

101-300 57 11.9 79.6

301-500 24 4.3 83.9

501-700 15 6.3 90.2

701+ 32 9.8 100

5.2.3 Value of grapes supplied to winery by respondents

Section 1 of the questionnaire contained a question asking the respondents to detail the

value of the grapes they supplied to the winery. The statistics showed a mean score of

$138,916 with a standard deviation of $276,133 (n=396). The standard deviation score

86

illustrates a wide distribution of responses. Table 5.3 illustrates the results further and

shows that approximately 43% of all respondents‟ grapes were supplied at a value of

less than or equal to $50,000, and that approximately 75% of all respondents supplied

grapes less than or equal in value to $100,000. The large number of responses in the

$500,000 plus category appears to be elevating the mean. These results cannot be

benchmarked against other similar studies as the value of the produce from a “best”

relationship had not been examined in previous studies. Overall it can be observed that

the “best” relationship a respondent had with a winery involved receiving a relatively

small amount of money (i.e., less than or equal to $50,000).

Table 5.3: Value of grapes supplied to winery by respondents

Value of grapes

($)


50,000 170 42.9 42.9

50,001- 100,000 123 31.6 74.5

100,001- 500,000 85 21 95.5

500,001 + 18 4.5 100

5.2.4 Average price per tonne of grape supplied to winery

Descriptive statistics of the trading relations between the grape grower respondents and

the wineries included data for the average price per tonne of the grapes supplied to the

winery. The descriptive statistics showed a mean score of $1409.4 and a standard

deviation of $916.25 (n= 396). Further analysis of the data, shown in Table 5.4,

illustrates that 32% of all respondents received less than or equal to $1000 per tonne for

their grapes, while 68% of respondents received $1001 and above for their grapes. The

average price per tonne is above the cost of production for grapes, which is between

$250 and $400 per tonne (Davidson, 2010); however, no other studies have observed

the average price per tonne supplied to wineries and thus, benchmarking this statistic is

not possible.

87

Table 5.4: Price per tonne of grapes supplied to the winery by respondents

Price per tonne

($)


500 75 18.9 18.9

501- 1000 52 13.1 32.0

1001- 1500 136 34.3 66.3

1501+ 133 33.7 100

5.2.5 Other wineries supplied and the amount of grapes supplied to those

wineries.

Section 1 of the instrument posed questions mostly involving the business relationship

that was most important to the grape grower‟s business. Therefore, the questions were

specifically asked with respect to that relationship (for example, price per tonne, value

of grapes, and volume of grapes). However, section 1 also posed two questions relating

to the relationships that growers had with other wineries, specifically how many other

business relationships the growers had and the amount of grapes that they supplied to

those other wineries. Respondents‟ results showed that they supplied an average (mean)

of 1.92 other wineries and approximately 22% of their total grape production went to

those other wineries (n=396). Further analysis of the data, illustrated in Tables 5.5 and

5.6, shows that approximately 56% of all the respondents supplied fewer than two other

wineries and approximately 63% of the respondents supplied less than or equal to 25%

of their production to the other wineries. These results are consistent with those of

Hobley (2007), who found that the majority of grape growers have fewer than two

contracts. However, Hobley (2007) only observed relationships as contracts, as opposed

to other types of relationships such as casual relationships not based on contracts or

spot market transactions. Overall, it can be surmised that the respondents were

representative of the sample frame in terms of the number of relationships they had

with wineries.

88

Table 5.5: Number of other wineries to which respondents supplied grapes

Number of other

wineries


Less than 2 219 55.5 55.5

2-4 159 40.2 95.7

More than 4 18 4.3 100

Table 5.6: Percentage of grape production supplied to the other wineries

Percentage of

grape

production (%)


25 248 62.6 62.6

26- 50 122 30.8 93.4

50+ 26 6.6 100

5.2.6 Business details of the winery that was supplied grapes

Much of the discussion of this chapter has been based on the details of trading relations

between the respondent and wineries supplied with grapes, including the price and

dollar amounts that the respondent received and the volumes of grapes supplied. In

section 1 of the questionnaire, the respondents were asked the business details of the

winery to which they were supplying grapes. Specifically, they were asked questions

regarding the size of the winery, the ownership of the winery and the wine region in

which the winery was located.

Respondents indicated that the majority of the wineries that they focused on in the

questionnaire were privately owned, small to medium sized enterprises (illustrated in

Tables 5.7 and 5.8). The wineries were located in all states in Australia, with the

Barossa Valley being the region in which the wineries were mostly located (illustrated

in Table 5.9). While no benchmarking figures are available for the size parameters of

89

wineries, there are 2420 wineries in Australia (Winetitles, 2010) and less than 10% are

publicly owned. However, the largest grape purchasers are the large, privately owned

companies. Seven the top 20 wine companies are privately owned and process 77% of

the grapes in Australia. Furthermore, Constellation Wines Australia and Fosters Wine

Group processed approximately 30% of the grapes from the 2009 vintage (Winetitles,

2010). However, the respondents focused on the “best” relationship and the data in

Table 5.7 shows that the winery was privately owned. Due to the fact that

approximately 10% of wineries in Australia are publicly owned, it can be surmised that

the respondents are representative of the sample frame in terms of the ownership of the

winery that constitutes their best relationship. Most privately owned wineries are small

to medium sized, and therefore the data in Table 5.8 reinforces the notion that the

winery relationships are representative of the sample frame.

Table 5.9 illustrates that the wine region in which the wineries were located was mainly

in the Barossa Valley, Riverland and Riverina (40% of responses). Furthermore, Table

5.11 shows the locations of the wineries by state and the table illustrates that

approximately 80% of all wineries were located in South Australia , NSW and Victoria.

Winetitles (2010) states that 77% of all wineries are located in those three states,

therefore, it can be surmised that the respondents were dealing with wineries that were

representative of wine production in Australia and are indicative of the target sample

population. Interestingly, 7.1% of respondents stated that they did not know in which

region the winery they supplied was located.

Table 5.7: Ownership of the winery to which respondents supplied grapes

Ownership of

the winery


Privately owned 239 60.4 60.4

Publicly owned 126 31.8 92.2

Don‟t know 31 7.8 100

90

Table 5.8: Size of the winery to which respondents supplied grapes

Size of the

winery


Small to medium 243 61.4 61.4

Large 127 32.1 93.5

Don‟t know 26 6.5 100

Table 5.9 Wine region winery was located in

Wine region Frequency % Cumulative %

Barossa Valley 64 16.2 16.2

Riverland 49 12.4 28.5

Riverina 45 11.4 39.9

McLaren Vale 33 8.3 48.2

Don‟t Know 28 7.1 55.3

Hunter Valley 24 6.1 61.4

Yarra Valley 21 5.3 66.7

Mornington Peninsula 11 2.8 69.4

Clare Valley 11 2.8 72.2

Adelaide Hills 11 2.8 75.0

Margaret River 10 2.5 77.5

Coonawarra 9 2.3 79.8

Goulburn Valley 9 2.3 82.1

Swan District 7 1.8 83.8

Granite Belt 7 1.8 85.6

Eden Valley 7 1.8 87.4

Rutherglen 7 1.8 89.1

Great Southern 5 1.3 90.4

Tasmania 5 1.3 91.7

91

Orange 4 1.0 92.7

Mudgee 4 1.0 93.7

Bendigo 3 .8 94.4

Limestone Coast 3 .8 95.2

Heathcote 3 .8 96.0

Geelong 2 .5 96.5

Pyrenees 2 .5 97.0

Pemberton 2 .5 97.5

Langhorne Creek 2 .5 98.0

King Valley 2 .5 98.5

Blackwood 1 .3 98.7

Tumbarumba 1 .3 99.0

Padthaway 1 .3 99.2

Gippsland 1 .3 99.5

Manjimup 1 .3 99.7

Canberra 1 .3 100.0

Table 5.10: State wineries were located in

State Frequency % Cumulative %

South Australia 192 48 48

NSW 87 22 70

Victoria 50 13 83

Western Australia 26 6.7 89.7

Queensland 7 1.9 91.6

Tasmania 5 1.4 93

Don‟t Know 28 7 100

92

5.2.7 Summary of trading relations of grape grower respondents

The respondents in the questionnaire detailed their trading relationships with the winery

they were asked to focus on and the other wineries that they traded with. This section of

the questionnaire (section 1) has highlighted numerous results of interest. A summary

of the results is shown in Table 5.11.

Table 5.11: Summary of the trading relationship of respondents and wineries

Duration of relationship Less than or equal to 5 years

Volume of grapes supplied Less than or equal to 100 tonnes

Value of grapes supplied Less than or equal to $50,000

Price per tonne of grapes supplied $1000-1500

Ownership of winery Privately owned

Size of winery Small to medium

Wine region of winery Barossa Valley

Other wineries supplied Less than 2

Percentage of grapes supplied to other

wineries Less than or equal to 25%

5.3 Section 3: Descriptive statistics of respondents The previous section of this chapter, 5.2, discussed the trading relationship details

between the respondents and the winery. The statistics from this section (5.2) were

derived from section 1 of the questionnaire and contained information relating to price

per tonne, volume of grapes etcetera. Section 3 of the questionnaire posed questions to

the respondents regarding the size and nature of their businesses. This part of the

chapter will exhibit the details of the respondents‟ (grape growers) business,

commencing with a discussion of the size of their vineyards. As previously mentioned,

section 2 of the questionnaire will be discussed in Chapter 6.

93

5.3.1 Size of the respondents‟ vineyards

Respondents were asked to complete a questionnaire item on the size of their vineyards.

The results showed a mean vineyard size of 58.1 acres with a standard deviation of

111.6 (n=396). Analysis of the results, exhibited in Table 5.11, shows that 53.3% of

respondents have vineyards less than or equal to 25 acres. Furthermore, Table 5.11

shows that over 75% of respondents have vineyards less than or equal to 50 acres. This

is similar to the evidence supplied by Hobley (2007) and Phylloxera Board SA (2010)

who commented that the majority of grape growers in Australia, and specifically in

South Australia, had vineyards of less than 50 acres. It can be surmised that the

respondents are representative of the sample frame in terms of the size of their

vineyards.

Table 5.12: Size of respondents vineyards in acres

Size (acres) Frequency % Cumulative %

25 211 53.3 53.3

26-50 87 22 75.3

51- 100 43 10.9 86.2

100+ 55 13.8 100

5.3.2 Number of years respondents operating their viticultural business

Section 3 of the questionnaire posed an item to respondents asking them to detail the

number of years they had been operating their viticultural business. Results showed that

respondents had been running their viticultural business for an average of 19.5 years

with a standard deviation of 13.9 years (n=396). Further analysis of the data, exhibited

in Table 5.12, illustrates that 47.5% of respondents had been operating their business

for less than or equal to 15 years. Table 5.12 also shows that a large number of

respondents had been operating their business for 26 or more years, which accounts for

a higher mean score. The data in Table 5.12 illustrates that 71% of respondents had

operated their businesses for less than 20 years. While no industry data was available to

benchmark this result, a similar study (Hobley, 2007) found that 78% of grape grower

respondents had operated their businesses for less than 20 years. Therefore, it can be

concluded that the respondents are representative of sample population in terms of the

length of business operation. A potential reason for the majority of respondents running

94

their business for less than 15 years may be that a boom in managed investment

schemes in grape production in the late 1990s led to accelerated grape plantings and,

therefore, the establishment of many grape growing businesses (Speedy, 2006).

Table 5.13: Number of years respondents operation of business

Years Frequency % Cumulative %

10 34 8.6 8.6

11-15 154 38.9 47.5

16-20 94 23.7 71.2

21-25 40 10.1 81.3

26+ 74 18.7 100

5.3.3 Number of people employed by respondents‟ businesses

Respondents were asked to detail the number of employees that worked for their

business. Respondents were asked to include all people who were actively working for

the business, including the owner. A mean score of 2.7 people with a standard deviation

of 3.0 (n= 396) was shown in the statistics. Further analysis, exhibited in Table 5.13,

shows that 85.4% of respondents‟ businesses had less than or equal to 3 employees.

While no industry statistics were available to benchmark this result, Hobley (2007), in a

study utilising a similar sample frame, found that 80% of grape growers had fewer than

five employees. With this in mind, it is reasonable to assume that the respondents are

representative of the sample population in terms of the number of employees.

Table 5.14: Number of people employed by respondents‟ businesses

Number of

people


3 338 85.4 85.4

4-6 45 11.4 96.8

7+ 13 3.2 100

95

5.3.4 Wine region location of respondents‟ businesses

Respondents were asked in section 3 of the questionnaire to list in which wine region

their businesses were located. The responses, exhibited in Table 5.14, show that 13.5%

of respondents had their businesses located in the Riverland and that respondents‟

businesses were located in all states of Australia. Furthermore, approximately 25% of

respondents were located in the Murray Valley Irrigation zone which encompasses the

Riverland and Riverina grape growing regions. While no industry data was available on

the number of grape growers in individual regions, investigations found that

approximately 20-30% of all grape growers in Australia are located in the Murray

Valley Irrigation zone (Davidson, 2010). Furthermore, 30% of all grapes harvested in

the 2009 vintage came from these two regions (ABARE, 2010) and 45% of all

respondents resided in South Australia. In addition to this, approximately 50% of all

grape production in the 2009 vintage came from South Australia, which is reflected in

the results, particularly Table 5.16, which shows that 50% of all respondents came from

South Australia (Winetitles, 2010). With these figures in mind, it is reasonable to

assume that the respondents are representative of the sample frame in terms of the

location of their grape growing businesses.

Table 5.15: Wine region location of respondents viticultural businesses

Wine Region Frequency % Cumulative %

Riverland 50 13.5 13.5

Barossa Valley 48 12.9 26.4

Riverina 44 11.9 38.3

McLaren Vale 34 9.2 47.4

Yarra Valley 19 5.1 52.6

Adelaide Hills 17 4.6 57.1

Hunter Valley 16 4.3 61.5

Clare Valley 13 3.5 65.0

Mornington Peninsula 12 3.2 68.2

96

Margaret River 9 2.4 70.6

Swan District 8 2.2 72.8

Coonawarra 8 2.2 74.9

Rutherglen 8 2.2 77.1

Mudgee 7 1.9 79.0

Tasmania 6 1.6 80.6

Goulburn Valley 6 1.6 82.2

Great Southern 6 1.6 83.8

Orange 6 1.6 85.4

Granite Belt 6 1.6 87.1

Heathcote 5 1.3 88.4

Eden Valley 5 1.3 89.8

Langhorne Creek 4 1.1 90.8

Tumbarumba 4 1.1 91.9

King Valley 4 1.1 93.0

Limestone Coast 4 1.1 94.1

Geelong 3 .8 94.9

Wrattonbully 3 .8 95.7

Cowra 3 .8 96.5

Canberra 2 .5 97.0

Pyrenees 2 .5 97.6

Pemberton 2 .5 98.1

Bendigo 2 .5 98.7

Manjimup 2 .5 99.2

Blackwood 1 .3 99.5

Padthaway 1 .3 99.7

Gippsland 1 .3 100.0

97

Table 5.16: State respondents were located in

State Frequency % Cumulative %

South Australia 186 50 50

NSW 89 24 74

Victoria 56 15 89

Western Australia 28 7.6 96.6

Queensland 6 1.7 98.3

Tasmania 6 1.7 100

5.3.5 Technical viticultural qualifications of respondents

Section 3 of the questionnaire asked respondents to list the viticultural qualifications

that they had obtained. Table 5.15 exhibits the qualifications of respondents and shows

that 67.4% of respondents had no formal qualifications (n=396). While no industry

statistics were available regarding the viticultural qualifications of grape growers in

Australia, a similar study found that 65% of grape growers had a technical, bachelor or

postgraduate qualification (Hobley, 2007). However, Hobley (2007) observed whether

growers had these qualifications and not whether these qualifications were viticulturally

based.

Table 5.17: Viticultural qualification of respondents

Qualification Frequency % Cumulative %

None 267 67.4 67.4

TAFE (technical

qualification)

66 16.7 84.1

Bachelor degree 29 7.3 91.4

Postgraduate degree 11 2.8 94.2

Other

(training seminars,

short courses)

23 5.8 100

98

5.3.6 Summary of descriptive statistics of respondents

Section 3 of the questionnaire was designed to highlight the description of the grape

grower respondents. Table 5.16 provides a summary of the descriptive statistics of

these respondents. What can be observed is that, in relation to grape trading terms, the

“best” relationship is a small length contract, for a small volume of grapes that has a

relatively small value but the price per tonne is relatively high. Therefore, it can be

surmised that the “best” relationship is one that provides the highest price per tonne of

the grapes.

Table 5.18: Summary of descriptive statistics of respondents

Size of Vineyards Less than or equal to 25 acres

Years of business operation 11-15

Number of people employed by business Less than or equal to 3

Wine region location of business Riverland

Viticultural qualifications of respondent None

5.4 Chapter Summary Chapter 5 detailed the univariate statistics of the questionnaire. Therefore, the chapter

dealt with sections 1 and 3 of the questionnaire and provided statistics of the

description of the trading relationships between the respondent and the winery and the

descriptive statistics of the respondents and their businesses. Sections 1 and 3 of the

questionnaire have now been discussed. Section 2 of the questionnaire, which relates to

the conceptual models and the examination of hypotheses, is the topic of discussion of

the next chapter, which is Chapter 6.

99

Chapter 6: An integrated model of buyer-seller

relationships in the Australia wine industry

6.1 Chapter outline This chapter presents the main research data collection including the testing of the

hypotheses and conceptual models, as outlined in Chapter 3, providing an integrated

model of buyer-seller relationships in the Australian wine industry.

The previous chapters of this thesis have followed a logical progression to the

estimation of two structural equation models (SEM), which are the integrated model

discussed above, and an alternative model. Previous chapters presented and discussed

literature. Exploratory qualitative research was performed and analysed which allowed

for the creation of the conceptual models with various paths between variables and

hypotheses formulated. Questionnaire scale items to test the conceptual models were

then selected from the literature and modified to be relevant to the context of the

research. This chapter involves statistically estimating the conceptual models, thereby

presenting an integrated model and an alternative model.

This chapter takes a two-step process to the SEM process. In the first step, a

measurement model process is presented whereby the questionnaire variables are

examined via exploratory factor analysis to see if they represent the constructs. The

second step of the process involves examining how each of the constructs are

associated to each other, thereby presenting an integrated model and an alternative

model which were tested using Partial Least Square Regression. Further analysis in the

form of cluster analysis is presented at the end of the chapter.

6.2 Measurement model of constructs In Chapter 3, conceptual models of communication elements between grape growers

and winemakers, their effect on trust and satisfaction (relationship quality), and the

influence of power asymmetry, was formulated. The first step to test the models was to

take the dimensions of communication, trust, satisfaction and power, and subject them

to Exploratory Factor Analysis utilising Principle Component Analysis (PCA). This

100

process was performed to identify the formation of the constructs and to discard items

which did not contribute to the factor (Anderson & Gerbing, 1988; Hobley, 2007). All

variables with factor loadings above 0.5 were retained. This process revealed that all

the dimensions of communication, trust, satisfaction and power were extracting on one

component except the dimension of communication modality. Modality of

communication was hypothesised in Chapter 3 as being either face-to-face or non-face-

to-face communication. Therefore, the hypothesis was dealing with the notion that face-

to-face communication was purely “real time” face-to-face communication between

two actors and seminar communications, and non-face-to-face communication was

concerned with modes that were not face-to-face, as for example email, newsletters and

telephone communications. The PCA results revealed that the construct communication

modality was extracting on more than one component, and that it was extracting not on

the “face-to-face versus non-face-to-face” dimensionality, but on a “direct” or

“indirect” dimensionality. The PCA showed that modalities such as personal direct

email, telephone and face-to-face dimensions were extracting together and as they

involved direct communication from one person to another, as opposed to

communication that is directed to a group of individuals. It is warranted to discuss

those modalities as direct communication. The other modes, such as newsletters, group

written letters, seminars and other modes of communication, could be discussed as

indirect modes as they are not from one actor to one actor, but from one actor to groups

of actors.

The next stage of the analysis involved the use of the Partial Least Squares (PLS)

approach to Structural Equation Modeling (SEM) to test the hypotheses. In the PLS

approach to SEM, the fit of the model is estimated via the inner and outer models.

6.2.1 Evaluation of the outer model

The outer model is evaluated by examining the individual item reliabilities and

convergent validity of the model. The individual item reliabilities were examined

through the factor loadings of the items on their respective constructs. Only items with

factor loadings of at least 0.4 were considered significant and retained in the model

(Hair et al. 2006). Thus, many of the items were not considered significant and were

excluded from the analysis, particularly related to the communication dimension. The

results of the outer model evaluation are exhibited in Table 6.1. As outlined in Chapter

4, the internal consistency of the model was assessed via the Cronbach Alpha

101

(Cronbach, 1970) and the composite reliability of the measurements (Werts et al.,

1974). These indicators rank from 0 (absence of homogeneity) to 1 (maximum

homogeneity), with a usual criteria of both indexes to be greater than 0.7. Table 6.1

illustrates that all composite reliability indices range from 0.784 to 0.944 and the

Cronbach Alphas range from 0.702 to 0.932, thereby satisfying the recommended

thresholds.

Table 6.1: Outer model evaluation of collaborative communication dimensions,

trust, satisfaction and power.

A B C D E F

Variables and indicators Factor

loading

Comp

reliability

Cronbach AVE

Feedback 1.000 1.000 1.000

Commfeed1 How much feedback do you

provide to this winery?

(summate of neg and positive

feedback

1.000

Formality 0.863 0.784 0.551

Commform1 When working with winery,

formal comm vs casual word

or mouth comm

0.567

Commform3 The winery‟s expectations of

us are communicated in

detail.

0.693

Commform4 The terms of our business

relationship with the winery

have been explicitly put into

words and discussed.

0.781

Commform5 Information sharing on

important issues has become

crucial to maintaining this

0.889

102

A B C D E F


loading

Comp

reliability

Cronbach AVE

partnership.

Indirect communication 0.789 0.712 0.667

Commnews Newsletter communication 0.735

Commsemin Seminars communication 0.891

Direct communication 0.853 0.749 0.748

Comcomp Computer: email

communication

0.759

Commface Face-to-face communication 0.960

Non coercive

communication attempts

0.821 0.702 0.699

Commiflu2 How frequently did the

winery‟s employees ask you

to perform a certain

operation, but didn‟t say what

penalty may occur if you

didn‟t do what they asked.

0.881

Comminflu3 How frequently did the

winery‟s employees say you

will be supplying grapes of a

certain quality, but didn‟t

give you specific information

e.g. what crop level they

would like, what spray

regime they would like or

other directions they would

like you to take to grow those

0.786

103

A B C D E F


loading

Comp

reliability

Cronbach AVE

grapes.

Satisfaction 0.900 0.851 0.696

Satisf1 We are very pleased with our

working relationship with the

winery.

0.904

Satisf2 Generally we are very

satisfied, with our overall

relationship with the winery.

0.920

Satisf3 The relationship our business

has with the winery has been

an unhappy one. (RS)

0.753

Satisf 5 I am happy with the contract I

have with the winery for my

grapes.

0.740

Trust 0.944 0.932 0.655

Trust 1 When things go bad, we

believe that the winery will

be ready and willing to offer

us assistance and support.

0.833

Trust 2 When making important

decisions, the winery is

concerned about our welfare.

0.868

Trust 3 When we share our problems

with the winery we know that

they will respond with

understanding.

0.865

104

A B C D E F


loading

Comp

reliability

Cronbach AVE

Trust 4 We can count on the winery

to consider how its decisions

and actions will affect us.

0.816

Trust 5 When it comes to things that

are important to us we can

depend on the winery‟s

support.

0.825

Trust 6 Even when the winery gives

us a rather unlikely

explanation, we are confident

that they are telling the truth.

0.802

Trust 8 The winery keeps the

promises that it makes to our

business.

0.622

Trust 9 Whenever the winery gives

us advice on our business

operations, we know that it is

sharing its best advice.

0.776

Trust 10 The winery offers me a fair

and reasonable price for my

grapes.

0.799

Power 0.784 0.711 0.611

Power 1 We have to follow the

winery‟s instructions or they

will get their grapes from

someone else.

0.798

105

A B C D E F


loading

Comp

reliability

Cronbach AVE

Power 2 We are expected to follow the

winery‟s instructions.

0.690

Power 3 We have influence over the

winery‟s actions.

0.880

6.2.2 Evaluation of the inner model

The first criterion used to measure the inner model was the discriminant validity. As

discussed in Chapter 4, the discriminant validity measures whether every construct is

significantly different from the other measures. To analyse this, loadings and cross

loadings matrices were obtained, whereby the loadings are the Pearson correlation

coefficients to their own constructs (Chin, 2001; Gyau & Spiller, 2007) . All loadings

should be higher than the cross loadings which was the case in this study and is shown

in Table 6.2.

106

Table 6.2: Loadings and cross loadings of indicators and constructs

Direct com Formality Indirect com Noncoercv Power Satisfaction Trust Wineryfeed

Commcomp 0.7256 0.1788 0.1184 0.0026 -0.0641 0.0437 0.0561 -0.0320

Commface 0.9850 0.1006 -0.0246 -0.0616 -0.1148 0.1594 0.2342 0.1228

Commfeed1 0.0984 0.3327 -0.0251 -0.0990 -0.3284 0.5431 0.5603 1.0000

Commform1 -0.0831 0.5673 0.2078 0.0777 0.1973 0.1161 0.0564 0.1522

Commform3 0.1175 0.6934 0.3508 -0.0145 0.2638 0.1552 0.0185 0.1312

Commform4 0.1111 0.7808 0.3144 -0.0146 0.1244 0.2110 0.1310 0.2350

Commform5 0.1325 0.8887 0.2366 -0.1757 -0.0877 0.4489 0.3149 0.3410

Comminflue2 -0.0619 -0.1013 -0.0915 0.8810 0.1674 -0.2662 -0.1024 -0.0791

Comminflue3 -0.0202 -0.0769 0.0088 0.7858 0.1037 -0.1819 -0.1231 -0.0884

Commnews -0.0211 0.2168 0.7350 0.0078 0.1773 0.0172 -0.1019 0.0058

Commsemin 0.0201 0.3158 0.8907 -0.0844 0.1471 0.1062 -0.0933 -0.0387

Power1 -0.1160 -0.0075 0.0209 0.2395 0.7680 -0.3747 -0.3366 -0.2265

Power2 -0.0059 0.1602 0.1721 0.0255 0.6703 -0.1981 -0.2545 -0.1503

Power3 -0.0989 0.0171 0.2248 0.0874 0.7795 -0.3832 -0.5363 -0.3098

Satisf5 0.0701 0.3689 0.1527 -0.1838 -0.2562 0.7399 0.5227 0.3751

107

Direct com Formality Indirect com Noncoercv Power Satisfaction Trust Wineryfeed

Satisf1 0.1659 0.3677 0.0698 -0.2603 -0.4392 0.9074 0.7612 0.5295

Satisf2 0.1438 0.3390 0.0463 -0.2173 -0.4341 0.9201 0.7335 0.5366

Satisf3 0.0895 0.2330 0.0249 -0.2547 -0.3567 0.7529 0.5142 0.3356

Trust1 0.1726 0.1779 -0.0795 -0.0560 -0.4272 0.5862 0.8331 0.4372

Trust10 0.1643 0.2478 -0.0608 -0.1388 -0.4271 0.6645 0.7992 0.4393

Trust2 0.1752 0.1341 -0.1762 -0.0717 -0.5033 0.6197 0.8680 0.4797

Trust3 0.1669 0.2217 -0.1062 -0.1424 -0.4650 0.6394 0.8652 0.5303

Trust4 0.1879 0.1702 -0.0729 -0.0764 -0.4557 0.5623 0.8252 0.4417

Trust5 0.2200 0.2191 -0.1389 -0.0698 -0.4458 0.6353 0.8645 0.4919

Trust6 0.1449 0.1345 -0.2338 -0.0358 -0.4758 0.5882 0.8023 0.4555

Trust8 0.1453 0.2361 0.0314 -0.2408 -0.3169 0.6665 0.6225 0.3353

Trust9 0.1705 0.3194 0.0316 -0.1740 -0.4076 0.6921 0.7756 0.4450

108

Another criterion for measuring the discriminant validity is that the square root of the

AVE which must be greater than the correlation between the construct and the other

constructs in the study (Chin, 2001). This is shown in Table 6.2. The diagonal in the

table displays the AVE square roots instead of the usual values of “1”. This is known as

the Fornel Larcker Test (Fornel & Larcker, 1981; Gyau & Spiller, 2007). Bagozzi

(1994) suggests that the correlations between the coefficients in the model must be

smaller than 0.8. This is the case in Table 6.3.

109

Table 6.3: Correlations of the latent variables and the AVE square roots

Direct comm

Formality Indirectcom Noncoercv Power Satisfaction Trust Wineryfeed

Direct comm

1.0000

Formality 0.1246 1.0000

Indirectcom 0.0043 0.3333 1.0000

Noncoercv -0.0522 -0.1080 -0.0572 1.0000

Power -0.1122 0.0532 0.1923 0.1665 1.0000

Satisfaction 0.1460 0.3937 0.0851 -0.2735 -0.4522 1.0000

Trust 0.2127 0.2525 -0.1168 -0.1326 -0.5420 0.7717 1.0000

Wineryfeed 0.0984 0.3327 -0.0251 -0.0990 -0.3284 0.5431 0.5603 1.0000

110

6.2.3 Results of the structural model

To evaluate the hypotheses that were formulated from the literature and exploratory

research study (and highlighted in Chapter 3) and formed part of the conceptual model,

the R2 and the significance of the paths were used. A graphical representation of the

model is presented below in Figure 6.1.

Figure 6.1 Conceptual model of grape grower perceptions of relationship quality

in the Australian wine industry

The significance of the path‟s coefficients was determined using a bootstrapping

method with 1000 samples. The significance was then determined by using a one tail

Student‟s T distribution test, at a 0.5 significance level. The R2 measured the construct

variance explained by the model. Good fit exists when there is high R2. The R2 for the

two dependent variables in the model was 0.597 for trust and 0.587 for satisfaction

which indicated that the model provided a good fit for the latent constructs for use in

Partial Least Square Regression in this type of study (i.e. non time series study) (Chin,

2001; Gyau & Spiller, 2007). Table 6.4 illustrates the results of the structural model

111

which includes the data for the confirmation (or otherwise) of the hypotheses. Table 6.4

lists the T-Statistics and therefore shows whether the hypotheses were significant or

otherwise.

Table 6.4: Results of the structural model

Hypotheses Constructs Expected Sign Beta coefficients

T-Statistic

H1 Direct Com→Trust + 0.111** 3.033

H2 Direct Com→Sat + 0.035 0.908

H3 Indirect Com→Trust

- -0.101** 1.716

H4 Indirect Com→Sat - 0.063 1.193

H5 Feedback→Trust + 0.358** 7.520

H6 Feedback→Sat + 0.327** 6.085

H7 Noncoerc→Trust - -0.151** 3.551

H8 Noncoerc→Sat - -0.011 0.268

H9 Formality→Trust - 0.169** 3.820

H10 Formality→Sat - 0.261** 5.144

H11 Power→Trust - -0.402** 9.303

H12 Power→Sat - -0.342** 7.403

** Significant at p<0.05,

The results in Table 6.3 show the confirmation of H1, H3, H5, H6, H7, H11 and H12

and the rejection of H2, H4, H8, H9, 10. A graphical representation of the results is

presented in Figure 6.2 below.

112

Figure 6.2 A graphical representation of the of main structural equation model

results


Solid lines represent affirmed hypotheses, dashed lines represent rejected hypotheses.

6.3 Consideration of structural model results The structural model has illuminated numerous results. Of the 12 hypotheses, seven

hypotheses were confirmed and accepted while five were rejected. Of interest were the

five rejected hypotheses and the reasons for their rejection. This was done because the

link between the two constructs was statistically insignificant (i.e. p> 0.05) and the path

was testing the influence that elements of communication, in this case indirect and

direct communication and non-coercive communication attempts, have on satisfaction.

Therefore, it appears that the construct of satisfaction is a central theme to the rejected

hypotheses.

113

The central theme of satisfaction can be put into a wine industry context. As discussed

in Chapter 1 and 2, the Australian wine industry is suffering economic hardship

characterised by (apart from other reasons) an oversupply of grapes. This oversupply is

leading to hardship being felt by growers due to reduced prices per tonne for their grape

products. In many cases, growers were receiving below or close to below cost prices

(Davidson, 2010). This has led to the industry proposing that 20,000 hectares,

(approximately 20%) of grape vines, be removed due to the unsustainably high levels

of grape production (Henry, 2009). It stands to reason that, regardless of the elements

of communication between the two actors, the price that growers receive for their

grapes is so low that they cannot be satisfied in any way by the relationship. This

argument is further validated by the fact that the power asymmetry was having a very

strong negative influence on trust and satisfaction (H11 and H12), evident in high beta

coefficients and T-statistics shown in Table 6.3. As postulated in Chapter 1 and

discussed in Chapter 2, satisfaction was an element of the relationship that was of

interest to observe, particularly in view of the low grape prices received in the industry,

and other industry related issues.

In this study, relationship quality was measured as a multi-dimensional higher order

construct consisting of trust and satisfaction. Authors such as Crosby et al. (1990),

Dorch et al. (1997), Kim & Cha (2002) and Kim et al. (2006)] empirically tested

relationship quality, mostly via SEM and other multi-variate regression techniques,

using trust and satisfaction a separate constructs, and they concluded that higher levels

of trust and satisfaction in the model corresponded with higher levels of relationship

quality. However, Scheer & Stern (1992) and Leuthesser (1997) empirically tested

relationship quality as a uni-dimensional construct whereby the construct of

relationship quality consisted of latent variables of trust and satisfaction. SEM literature

considered whether alternative estimation (also known as two step model estimation)

could be performed in order to observe which model best fits the data concerned

(Joreskog & World, 1982; Anderson, & Gerbing, 1988; McDonald & Ho, 2002). In this

instance, it would be of interest to observe a model which estimated relationship quality

as a uni-dimensional construct as opposed to a multi-dimensional one, thereby

satisfying a theoretical and methodological concern.

As such, the constructs exhibited in Table 3.1 would directly affect relationship quality

in the alternative model as opposed the multi-dimensional affect shown in Table 3.1.

Theoretically, the hypotheses would remain the same, although each independent

114

variable in the model (i.e. power, collaborative communication elements) would affect

the singular dependent variable (i.e. relationship quality). Therefore, the alternative

model hypotheses would be:

H1a. Face-to-face (direct) modes of communication positively influence


H2a. Non-face-to-face (non direct) modes of communication negatively influence


H3a. Uni-directional communication (feedback) from the winery positively

influences relationship quality.

H4a. Non-coercive communication attempts from the winery negatively influence


H5a. Formality of communication from wineries negatively influences relationship

quality.

H6a. A power asymmetry favouring the winery is decreasing grape growers‟

perception of relationship quality.

A graphical representation of the alternative model is presented in Figure 6.3

115

Figure 6.3 Alternative model based on uni-dimensional estimation of relationship

quality

The alternative model estimation is discussed in the next section.

6.4 Alternative structural model estimation As discussed in the previous section, an alternative model that conceptualises

relationship quality as a uni-dimensional construct was estimated. As in the multi-

dimensional construct model, the alternative model was estimated via the inner and

outer model process.

The same methodology was employed as in the alternative model estimation process

and when estimating the inner model items, factor loadings of at least 0.4 were

considered significant and retained in the model (Hair et al. 2006). Thus, many of the

items were not considered significant and were excluded from the analysis, particularly

those related to the communication dimensions. The results of the alternative model‟s

116

outer model evaluation are exhibited in Table 6.4. As previously discussed, the internal

consistency of the model was assessed via the Cronbach Alpha (Cronbach, 1970) and

the composite reliability of the measurements (Werts et al., 1974). Table 6.4 illustrates

that all composite reliability indices range from 0.783 to 1.00 and the Cronbach Alphas

range from 0.714 to 0.943, thereby satisfying the recommended thresholds of a

minimum of 0.7 for both measures (Cronbach, 1970; Werts et al. 1974).

Table 6.4: Outer model evaluation of collaborative communication dimensions,

trust, satisfaction and power of alternative model.

Variables and

indicators

Factor

loading

Comp

reliability

Cronbach AVE

Feedback 1.000 1.000 1.000

Commfeed1 1.000

Formality 0.817 0.784 0.536

Commform1 0.545

Commform3 0.659

Commform4 0.769

Commform5 0.906

Indirect

communication

0.783 0.714 0.651

Commnews 0.937

Commsemin 0.653

Direct

communication

0.853 0.746 0.748

Comcomp 0.725

Commface 0.985

117

Variables and

indicators

Factor

loading

Comp

reliability

Cronbach AVE

Non coercive

communication

attempts

0.823 0.771 0.700

Commiflu2 0.851

Comminflu3 0.822

Relationship

Quality

0.951 0.943 0.602

Satisf1 0.858

Satisf2 0.835

Satisf3 0.622

Trust 1 0.795

Trust 2 0.831

Trust 3 0.836

Trust 4 0.782

Trust 5 0.834

Trust 6 0.773

Trust 8 0.673

Trust 9 0.791

Trust 10 0.798

Power 0.783 0.716 0.547

Power 1 0.757

Power 2 0.668

118

Variables and

indicators

Factor

loading

Comp

reliability

Cronbach AVE

Power 3 0.789

The inner model for the alternative model was evaluated. In this step discriminant

validity was observed and loadings and cross loadings matrices were examined,

whereby the Pearson correlation coefficients were compared to their own constructs

(Chin, 2001; Gyau & Spiller, 2007). All the loadings should be higher than the cross

loadings which was the case. The results of inner model evaluation are shown in Table

6.5.

119

Table 6.5: Loadings and cross loadings of indicators and constructs in the alternative model

Direct comm Formality Indirectcom Noncoercv Power RQ Wineryfeed

Commcomp 0.724857 0.176456 0.104321 0.004319 -0.063268 0.055244 -0.031968

Commface 0.985221 0.101678 -0.037266 -0.059492 -0.114375 0.222187 0.122807

Commfeed1 0.098553 0.338850 -0.009748 -0.099868 -0.329257 0.584158 1.000000

Commform1 -0.083291 0.545459 0.170199 0.071677 0.195595 0.079059 0.152243

Commform3 0.117382 0.659299 0.303213 -0.013195 0.264465 0.064982 0.131180

Commform4 0.110928 0.768881 0.279608 -0.015839 0.124804 0.164319 0.234972

Commform5 0.132427 0.906140 0.210594 -0.172532 -0.086292 0.376137 0.340996

Comminflue2 -0.062003 -0.107520 -0.086712 0.850721 0.164766 -0.162531 -0.079131

Comminflue3 -0.020207 -0.081098 0.064592 0.821803 0.102024 -0.149936 -0.088390

Commnews -0.021266 0.212307 0.936697 0.017765 0.178499 -0.068095 0.005809

Commsemin 0.019965 0.307543 0.652843 -0.083605 0.150698 -0.031477 -0.038685

Power1 -0.115950 -0.021278 0.057949 0.236038 0.757098 -0.367440 -0.226526

Power2 -0.005952 0.145729 0.182647 0.019127 0.667776 -0.249360 -0.150315

Power3 -0.098876 0.006391 0.201411 0.088106 0.788797 -0.513653 -0.309760

120

Direct comm Formality Indirectcom Noncoercv Power RQ Wineryfeed

Satisf1 0.165977 0.375144 0.046365 -0.257533 -0.439535 0.850350 0.529543

Satisf2 0.143900 0.348054 0.023056 -0.214309 -0.433725 0.834888 0.536561

Satisf3 0.089546 0.242099 -0.024616 -0.251529 -0.354832 0.621710 0.335603

Trust1 0.172736 0.187600 -0.076711 -0.057324 -0.428783 0.794774 0.437217

Trust10 0.164394 0.255662 -0.078488 -0.140717 -0.428645 0.798294 0.439307

Trust2 0.175351 0.146014 -0.144919 -0.071040 -0.506581 0.830785 0.479748

Trust3 0.167048 0.232658 -0.112043 -0.142979 -0.466642 0.836326 0.530318

Trust4 0.188016 0.179914 -0.063029 -0.077604 -0.457914 0.781641 0.441711

Trust5 0.220116 0.227720 -0.139035 -0.073719 -0.447452 0.833780 0.491930

Trust6 0.144946 0.141580 -0.206856 -0.037414 -0.479143 0.772802 0.455452

Trust8 0.145447 0.242979 0.013095 -0.240236 -0.316463 0.672681 0.335300

Trust9 0.170548 0.324505 -0.006977 -0.176085 -0.409002 0.791092 0.444972

121

Further discriminant validity tests were performed on the alternative model in the form

of the Fornel Larcker test in which the square root of the AVE (average variance

extracted) must be greater than the correlation between the construct and the other

construct (Fornel & Larker, 1981; Chin, 2001). Table 6.6 illustrates the test and that

the correlations between the coefficients in the model are smaller than 0.8; thereby

further supporting discriminant validity (Bagozzi, 1994).

122

Table 6.6: Correlations of the latent variables and the AVE square roots

Direct comm

Formality Indirectcom Noncoercv Power RQ Wineryfeed

Direct comm 1.0000

Formality 0.1248 1.0000

Indirect com -0.0097 0.2862 1.0000

Noncoercv -0.0501 -0.1133 -0.0168 1.0000

Power -0.1116 0.0374 0.2003 0.1609 1.0000

RQ 0.2024 0.3235 -0.0667 -0.1869 -0.5424 1.0000

Wineryfeed 0.0985 0.3388 -0.0097 -0.0998 -0.3292 0.5841 1.0000

123

The structural model for the alternative model was estimated in the same fashion as the

original method. The significance of the path‟s coefficients was determined using a

bootstrapping method with 1000 samples. The significance was then determined by a

one tail Student‟s T distribution test, at a 0.5 significance level; a T-statistic of a

minimum of 1.65 would create significance at that level (Hair et al. 2006). The R2 of

the model was 0.54 for the dependent variable (relationship quality) which showed a

good fit for the latent construct (Chin, 2001; Gyau & Spiller, 2007). Table 6.7

illustrates the results of the structural model for the alternative model.

Table 6.7: Results of the structural model for the alternative model

Hypotheses Constructs Expected Sign Beta coefficients

T-Statistic

H1a Direct Com→RQ + 0.092** 2.435

H2a Indirect Com→RQ - -0.042** 1.924

H3a Feedback→RQ + 0.366** 8.163

H4a Noncoerc→RQ - -0.058** 1.747

H5a Formality→RQ - 0.209** 5.124

H6a Power→RQ - -0.402** 9.972


The results shown in Table 6.7 for the alternative structural model show the

confirmation of H1a, H2a, H3a, H4a and H6a and the rejection of H5a. A graphical

representation of the results of the alternative structural model is shown in Figure 6.4

124

Figure 6.4 Graphical representation of the alternative structural model results


Solid lines represent affirmed hypotheses, dashed lines represent rejected hypotheses.

To further validate the results of the structural models, exploratory cluster analysis was

performed to observe how satisfaction, power and trust were perceived by the various

groups (clusters) and the demographic and contracting relations (between themselves

and the wineries) of the groups. This is the area of discussion for the next section of

Chapter 6.

6.5 Power, Satisfaction and Trust cluster analysis As discussed previously, an exploratory phase of the study, utilising K- Means cluster

analysis, was performed to observe groups of growers and their perception of

125

satisfaction, power and trust, as the structural model indicated, and industry economic

circumstance postulated, that these dimensions may be diminishing due to rising prices.

In order to perform the cluster analysis the following methodology was employed.

6.5.1 Cluster analysis methodology

The construct of trust was measured on eight items, satisfaction was measured on three

items and power was measured on three items

SPSS statistical program version 17.0 was used for all statistical computations.

Exploratory factor analysis using principal component analysis with a varimax rotation

was applied to the satisfaction, power and trust constructs. In this analysis, all factors

with Eigen values above one were extracted and only factors with loadings above 0.5

were retained. To test for the appropriateness of the factor analysis for the scale, the

Kaiser-Meyer-Olkin Measure of Sampling Adequacy (KMO-MSA) was conducted for

all the scale items with more than one indicator variable. All fell within the accepted

region of greater than 0.5 (Nunnally, 1978). In addition, these measures were purified

using the Cronbach Alpha. The results of the Cronbach Alphas, factor analysis, mean,

medians and standard deviation of the questionnaire items and their results are shown

in Table 6.7.

Table 6.8: Factor analysis and results of Trust, Satisfaction and Power

dimensions

Factors and Items Factor Loadings

Mean Median Standard Deviation

Trust

KMO=.909 Cronbach‟s alpha = .924, Explained variance = 60.16%

When things go bad, we believe that the winery will be ready and willing to offer us assistance and support.

0.835 3.81 4 1.53

When making important decisions, the winery is concerned about our welfare.

0.862 3.86 4 1.63


0.861 3.96 4 1.52

126


0.824 3.66 4 1.51

When it comes to things that are important to us we can depend on the winery‟s support.

0.860 3.93 4 1.44

Even when the winery gives us a rather unlikely explanation, we are confident that they are telling the truth.

0.798 4.13 4 1.57


0.778 4.73 5 1.35


0.798 4.45 5 1.53

Satisfaction

KMO = .678, Cronbach„s alpha = .860, Explained variance =78.16 %


0.907 4.80 5 1.38


0.929 4.90 5 1.43

RS The relationship our business has with the winery has been an unhappy one.

0.811 5.35 6 1.32

Power

KMO = .611 , Cronbach‟s alpha =.668 ., Explained variance = 61.22%


0.856 5.19 6 1.58


0.794 5.85 6 1.11


0.689 5.15 5 1.517

RS= reverse score

Hierarchical cluster analysis was performed using Ward‟s method and the resulting

dendrogram uncovered three distinct clusters The respondents were then clustered into

the three groups using K-Means Cluster analysis and ANOVA and cross-tab analysis

was performed to see how the clusters perceived the active variables of trust,

127

satisfaction and power. The ANOVA and cross-tab analysis was also performed to see

how the cluster perceived the passive variables, which related to contracting conditions,

such as the length of the contract, the price per tonne paid for the grapes, length of time

a respondent had worked as a grape grower, the wine region in which the grower was

located and the size and ownership of the winery to which they were contracted.

F-test and Bonferroni tests were performed to see if there was a statistical difference

between the clusters in terms of active and passive variables. The tests showed that the

differences between some of the active and passive variables in relation to the clusters

were statistically significant and all the variables that were significant within and

between groups were retained (Janssens, 2008). The mean results, by cluster, are

shown in Appendix 2.The cluster analysis illuminated 3 distinct relationship types and

is discussed in the next section with the mean, median and standard deviation scores of

the questionnaire items by cluster shown in Table 6.9. The next section of the chapter

discusses the details of the cluster analysis.

Table 6.9: Questionnaire item mean, median and standard deviation score by

cluster

Questionnaire item Mean Median Standard Deviation

Trust


Cluster 1 (n= 54) 1.83 1 1.40

Cluster 2 (n= 219) 3.66 4 1.21

Cluster 3 (n= 123) 5.86 6 1.11


Cluster 1 (n= 54) 1.50 1 0.77

Cluster 2 (n= 219) 3.75 4 1.25

Cluster 3 (n= 123) 5.07 5 1.28


Cluster 1 (n= 54) 1.67 1 0.93

128


Cluster 2 (n= 219) 3.91 4 1.18

Cluster 3 (n= 123) 5.07 5 1.16


Cluster 1 (n= 54) 1.69 1 1.01

Cluster 2 (n= 219) 3.54 4 1.17

Cluster 3 (n= 123) 4.72 5 1.29


Cluster 1 (n= 54) 1.63 1 0.81

Cluster 2 (n= 219) 3.95 4 1.02

Cluster 3 (n= 123) 4.90 5 1.14


Cluster 1 (n= 54) 1.78 1 1.22

Cluster 2 (n= 219) 4.15 4 1.09

Cluster 3 (n= 123) 5.15 5 1.15


Cluster 1 (n= 54) 2.61 2 1.42

Cluster 2 (n= 219) 4.82 5 0.93

Cluster 3 (n= 123) 5.51 5 0.98


Cluster 1 (n= 54) 2.00 2 1.16

Cluster 2 (n= 219) 4.54 5 1.11

Cluster 3 (n= 123) 5.38 6 1.12

Satisfaction


Cluster 1 (n= 54) 2.48 3 1.29

129


Cluster 2 (n= 219) 4.76 5 0.81

Cluster 3 (n= 123) 5.88 6 0.86

Generally we are very satisfied with our overall relationship with the winery.

Cluster 1 (n= 54) 2.33 2 1.26

Cluster 2 (n= 219) 4.94 5 0.82

Cluster 3 (n= 123) 5.98 6 0.83

RS The relationship our business has with the winery has been an unhappy one.

Cluster 1 (n= 54) 3.69 4 1.97

Cluster 2 (n= 219) 5.59 6 0.88

Cluster 3 (n= 123) 6.30 6 0.66

Power


Cluster 1 (n= 54) 5.91 7 1.75

Cluster 2 (n= 219) 5.82 6 1.02

Cluster 3 (n= 123) 3.76 4 1.39


Cluster 1 (n= 54) 6.35 7 1.33

Cluster 2 (n= 219) 6.18 6 0.76

Cluster 3 (n= 123) 5.04 5 1.11


Cluster 1 (n= 54) 5.87 7 1.74

Cluster 2 (n= 219) 5.42 5 1.14

Cluster 3 (n= 123) 4.33 5 1.66

(RS= reverse scored)

130

6.5.2 Cluster 1: “Unsustainable Relationship”

Cluster 1 contained 54 respondents (14% of respondents). This group experienced

strong negative power asymmetry (i.e. the winery had strong power over them) and had

strong negative satisfaction and strong negative trust (strong distrust of the winery).

The respondents in this group had the longest length of contract with the winery (10.5

years) and received a very low price per tonne for their grapes ($692 per tonne). This

group had also spent the longest period of time as grape growers (26.69 years) and their

businesses were located in warm climate wine regions (70%) such as the Riverland and

Riverina. The “unsustainable relationship” involved a contract with a large, publicly

owned winery (65%).

6.5.3 Cluster 2: “OK relationship”


moderated negative power asymmetry (i.e. the winery had moderated negative power

over them) and experienced low positive satisfaction (i.e. they were slightly satisfied

with the relationship) and low negative trust. This group had the shortest length of

contract (7 years) and received a medium price for their grapes ($1,264). This group

had also spent the shortest period of time growing grapes (17 years) and their

businesses were located in cool to warm wine growing regions such as Coonawarra,

McLaren Vale, Barossa and the Yarra Valley. The “OK relationship” respondents were

contracted to small to medium (70% were SME) wineries which were mostly (65%)

privately owned.

6.5.4 Cluster 3: “Good Relationship”


strong positive power (i.e. they had strong power over the wineries), experienced

moderate satisfaction (i.e. they were moderately satisfied with the relationship), and

moderate positive levels of trust (i.e. they moderately trusted the winery). Their

contract with the winery was for a medium length of time in view of the other clusters

(10 years) and they received the highest price for their grapes of any group ($1,981).

This group had spent a medium length of time in business as grape growers compared

to the other groups (19 years) and their businesses were located in cool wine growing

regions (80% of this group) such as the Adelaide Hills, Barossa Valley, Yarra Valley,

131

Tumbarumba, Eden Valley and Geelong. The “Good relationship” respondents were

contracted to small to medium sized wineries (75% were SME) that were mostly (80%)

privately owned. Table 6.10 provides a summary of the cluster analysis results. Table

6.10 provides a summary of the cluster analysis results.

Table 6.10 Summary of cluster analysis results

Unsustainable Relationship OK relationship Good relationship

54 respondents (14% of

sample)


sample)


sample)

Grape grower experienced

strong power asymmetry

favouring winery

Grape grower

experienced moderate

power asymmetry

favouring winery

Strong power asymmetry

favouring grape grower


strong dissatisfaction and

strong distrust of winery

Grape grower

experienced slight

satisfaction with winery

relationship and low

distrust


moderate satisfaction and

trust with winery

relationship

Grape grower contracted to

winery for 11 years

Grape grower contracted

to winery for 7 years

Grape grower contracted to

winery for 10 years

$692 per tonne contracted $1,264 per tonne

contracted

$1,981 per tonne

contracted

Grape grower in business for

27 years

Grape grower in business

for 17 years

Grape grower in business

for 19 years

Grape grower located in

warm climate wine region


cool to warm climate

wine region


cool climate wine region

Winery mainly a large,

publicly owned business

Winery mainly an SME

(70%), privately owned

business

Winery mainly an SME

(80%), privately owned

business

132

6.6 Chapter conclusion This chapter was concerned with the quantitative phase of the study. In particular it

described the results of section 2 of the questionnaire instrument. Structural equation

modelling illuminated the respondents‟ perceptions of communication and the effect of

power, trust and satisfaction, while the cluster analysis exhibited the types of

relationships that the respondents experienced. Many results have been uncovered and

the discussion and implications of these results are highlighted in the next and final

chapter, Chapter 7.

133

Chapter 7: Discussion, conclusion and implications for

further research

7.1 Chapter outline This is the final chapter of the thesis. The chapter begins with a summary of the study

followed by a discussion of the hypotheses and the cluster analysis which were

performed as part of the exploratory phase of the study. The research questions are

discussed and the chapter ends with a discussion of the conclusion of the study and

recommendations for further research. Firstly, a summary of the study is presented.

7.2 Summary of the research process The study utilised the relationship between grape grower and wineries as the research

context. The justification for using this context was that:

(a) the Australian wine industry is of great economic importance to the economy of

Australia; and that

(b) there is a large volume of interaction between grape growers and wineries,

particularly during the grape growing season, providing a fertile area for B2B

research.

The study relied on the grape grower perspective of the relationship, and justification

for doing so was based on the fact that:

(a) the grape grower forms the most important link in the wine production chain as

the quality of the grapes they produce greatly influences the quality of the wine;

and

(b) the Australian wine industry is moving to focus on promotion of regionality in

wine products, and regionality is grape grower based (i.e. the wine regions

where the grapes are grown determines the region which is displayed on the

wine bottle);

(c) therefore, growing importance is being vested in the grape grower in the wine

industry supply chain.

134

Furthermore, there were numerous grape grower/ winery relationships with

approximately 4500-6500 grape growers and 2420 wineries in the industry, and many

growers had multiple relationships with multiple wineries (ABS, 2009b; Mckenzie

pers. comm., May 2009; Winetitles, 2010). Potentially tens of thousands of grape

grower/ winery relationships exist in the industry. The economic state of the Australian

wine industry also provided the impetus for research as an oversupply of grapes had led

to reduced prices for grape growers and it seemed important to explore how this

phenomenon was impacting on the relationships between grape growers and wineries.

The thesis (based on the research) followed successive stages in the structural equation

modelling (SEM) process. Firstly, an evaluation of the literature uncovered the

dimensionality of the constructs involved in the relationship and the nature of the

concept of relationship quality. Secondly, an exploratory study was performed on grape

growers in South Australia and Victoria to allow for the conceptualisation (in view of

the literature review) of a model and the modification of an alternative model. The

exploratory study allowed for the development of hypotheses and illuminated numerous

issues in the relationship between the two actors. Communication modality was of

importance with face-to-face (direct) communications and non face-to-face (indirect)

communication modes being highlighted by the growers. Feedback from the winery

was also deemed to be important, and the issue of the formality of the communication

between the two actors was of interest.

Linked to the oversupply of grape issues in the wine industry was the issue of power

asymmetry, and the use of power by the wineries over the growers. Overall, the

exploratory study observed that relationship quality (trust and satisfaction) was being

influenced by elements of collaborative communication (as defined by Mohr & Nevin,

1990 and Mohr et al. 1996) and affected by power asymmetry. The qualitative,

exploratory study allowed for the hypotheses to be formed and structural models were

devised.

The structural models were then quantitatively tested (otherwise known as the causal

study) on data gathered from an online questionnaire completed by 396 grape growers

in South Australia, Victoria, Queensland, New South Wales, Western Australia and

Tasmania (all the grape growing states in Australia). The structural models were tested

utilising Partial Least Square Regression (PLS) to test the paths between constructs.

The PLS SEM process utilised confirmatory factor analysis (CFA) to reduce the

dimensionality of the constructs in the inner model of the structural models, and

135

regression was used to estimate the paths between constructs in the outer model of the

structural models. Therefore, the SEM phase of the study tested the paths between the

constructs (i.e. tested the hypotheses) and a discussion of the individual hypotheses is

made in the next section of this chapter.

7.3 Hypothesis discussion The following section of this chapter provides an individual discussion of each of the

hypotheses related to both the main and alternative SEM models. The discussions of the

two models‟ hypotheses have been grouped together for ease of reading: that is to say,

the three hypotheses related to the effect of direct modes of communication on trust,

satisfaction and relationship quality have been grouped.

7.3.1 H1: Direct modes of communication positively influence trust.

The PCA tests showed that the modality of communication was extracting on two

components. One component was extracting on direct modes of communication and the

other on indirect modes of communication. The results of the SEM showed that there

was a positive, statistically significant effect of direct communication on trust. Thus,

the null hypothesis is rejected and H1 is affirmed.

This result seems to affirm the findings of Cannon & Homburg (2001) and Daft &

Lengel (1984); however, it focuses purely on the effectiveness of face-to-face

communication and briefly on “less rich” modes of communication, without indicating

what those less rich modes are. The results of the study show that email communication

(that is direct to the respondent) is considered a direct mode as opposed to a group

email (an email sent to a group of respondents) and has a positive effect on trust. If

“rich” communication is face-to-face communication as discussed by Daft & Lengel

(1984) then, in view of these results, rich communication can be more than face-to-face

and, as a result, direct email communication could be considered “rich”.

7.3.2 H2: Direct modes of communication positively influence satisfaction

Related to the discussion of H1, the PCA results indicate an extraction on two

components and the results of the SEM showed that there is a positive link between

direct communication and satisfaction. However, there was no statistically significant

136

link between direct communication and satisfaction. Therefore, the null hypothesis is

accepted and H2 is rejected.

7.3.2.1 H1a: Direct modes of communication positively influence relationship

quality.

As previously mentioned in Chapter 6, an alternative model based on the uni-

dimensional measurement of relationship quality by Leuthesser (1997) and Scheer and

Stern (1992) was made. The hypothesis measured the effect that direct modes of

communication had on relationship quality. The results of H1a showed a statistically

significant positive effect of direct modes of communication on relationship quality.

Therefore, the null hypothesis is rejected and H1a is affirmed. It appears that the uni-

dimensional measure of relationship quality provides a model which better tests the

effect between the two constructs, as the path between the two is statistically significant

(Hair et al, 2006), as opposed to H2 which was not statistically significant.

7.3.3 H3: Indirect modes of communication negatively influence trust

As discussed in the previous analysis of H1 and H2, the PCA results indicated that

communication was extracting on two components. The “indirect” modes of

communication were seminars and newsletters (i.e. modes that are used to

communicate to a group of respondents, not individuals). The SEM process uncovered

that there was a negative, statistically significant link between indirect modes of

communication and trust. Therefore, the null hypothesis is rejected and H3 is affirmed.

In view of the results of Cannon & Homburg (2001) and Daft & Lengel (1984) in

relation to their notion of “less rich” forms of communication, newsletters and seminars

must be considered “less rich”.

7.3.4 H4: Indirect modes of communication negatively influence satisfaction.

The indirect modes of communication (seminar and newsletter) were hypothesised to

influence satisfaction negatively. The SEM process showed that there was a positive

influence of indirect modes of communication on satisfaction (opposing the

hypothesis); however, the link between the two was statistically insignificant (at the

95% confidence level). Thus, the null hypothesis is accepted and H4 is rejected.

137

Hypotheses 1, 2, 3 and 4 illustrated that there is a link between trust and

communication modality and no link between satisfaction and communication

modality. It can be surmised that no satisfaction was derived from all of the

communication modes; however, there was an effect on trust. This may be the result of

the wine industry economic downturn, whereby the respondents‟ levels of satisfaction

are being affected by other elements of the relationship (other than communication).

This quandary is discussed in later sections of this chapter.

7.3.4.1 H2a: Indirect modes of communication negatively influence relationship

quality.

The results of H3 showed that indirect modes of communication negatively influence

trust; however, the result of H4 showed a statistically insignificant link between the

construct and satisfaction. As such, the results of the main model provide a quandary in

that one link between the construct and one dimension of relationship quality is

significant while the other link is not. As such, the results of the alternative model

whereby relationship quality is uni-dimensional are of interest.

The results of the alternative model showed a statistically significant negative link

between indirect communication and relationship quality based on the Scheer & Stern

(1992) and Leuthesser (1997) uni-dimensional estimation method. As such, the null

hypothesis is rejected and H2a is affirmed. Therefore, the result of H2a echoes those of

H3 and shows that the data provides a better fit for the model when relationship quality

is measured as a uni-dimensional construct as opposed to a multi-dimensional one.


trust

As discussed in Chapter 3 (the exploratory study), the communication in the

relationship was considered uni-directional (it was only coming from the wineries and

not from the growers) and it was posited to be influencing trust. The SEM results

showed that feedback was influencing trust statistically significantly and positively.

Therefore, the null hypothesis is rejected and H5 is affirmed. This result affirms the

results of the exploratory study and is in line with the results of Mohr et al. (1996);

138

however, Mohr et al. (1996) did not test the uni-dimensional nature of communication

and only focused on the bi-directionality of communication which was shown to affect

trust. The uni-dimensional nature of communication would appear, logically, to affect

trust negatively as one actor‟s view is not being heard or acknowledged. However, the

affirmation of H5 disproves this assumption.


satisfaction.

In line with the findings regarding H5, the SEM results showed that uni-directional

communication (feedback) was positively, statistically, significantly influencing

satisfaction, thereby affirming the results of the exploratory study and the works of

Mohr et al. (1996) that feedback positively influences satisfaction. However, as in line

with the discussion of H5, this is a partial fulfilment of the Mohr et al. (1996) results as

that study observed the bi-directionality. Therefore, based on the results of this study,

the null hypothesis is rejected and H6 is affirmed.

7.3.6.1 H3a- Uni-directional communication from the winery positively influences


As in the estimation of H1a and H2a, H3a provided a uni-dimensional estimation of

relationship quality. However, unlike the results of H2 and H4 (when estimated multi-

dimensionally) the results of H5 and H6 were affirmed. In relation, to the alternative

model estimation, the result of H3a affirms the multidimensional estimation in that a

positive, statistical link between uni-directional communication and relationship quality

was found. Therefore, the null hypothesis is rejected and H3 is affirmed. As such, the

alternative model estimation mirrors the results of the main model, thereby adding

weight to both models.


influence trust.

Mohr & Nevin (1990) posit the notion of non-coercive communication attempts and

further stated that it affects the relationship but did not state if this construct directly

influences trust and satisfaction. The SEM results affirm the exploratory study results

139

and showed that there was a statistically significant negative link between non-coercive

communication and trust. Therefore, the null hypothesis is rejected and H7 is affirmed.


influence satisfaction

Connected to the discussion of H7 is the notion that non-coercive communication

attempts negatively influence satisfaction. The SEM results showed that a negative link

between the two constructs exists; however, the linkage was statistically insignificant.

Therefore, the null hypothesis is accepted and H8 rejected.

7.3.8.1 H4a- Non-coercive communication attempts from the winery negatively

influence relationship quality.

Similar to the results of H2 and H4, results of the main model showed a divergence

between H7 and H8 whereby H7 was affirmed and H8 was rejected. Therefore, it was

of interest to observe the estimation of the alternative model whereby relationship

quality was measured uni-dimensionally and, as such, H4a was observed having a uni-

dimensional effect on the non-coercive construct. The results of H4a showed a

statistically significant negative effect of non-coercive communication on relationship

quality. Therefore, the null hypothesis is rejected and H4a is affirmed. As such, the

results of H4a embody the result of H6 and show that, for this construct (non-coercive

communication attempts), the alternative model, based on the uni-dimensional

measurement of relationship quality, provide a better fit for the data.


trust

Mohr & Nevin, (1990) and Mohr et al. (1996) commented that the formality of

communication does have an effect on the relationship; however, they did not test its

effect directly on trust. The exploratory study showed that it negatively influences trust,

but the SEM results showed a statistically significant positive relationship between

formality and trust and therefore, H9 is rejected. This is contrary to the exploratory

study results and may be due to the relatively small sample size of the exploratory study

interview (13). H9 was posited mainly in view of the exploratory study results, as they

140

contradicted the literature. Therefore, the SEM process has affirmed the literature and

contradicted the exploratory study results.


satisfaction.

In line with the literature discussion in 7.3.8, formality of communication positively

influenced satisfaction. The results of the exploratory study were contradictory to this

notion; therefore H10 was posited to negatively influence satisfaction. Subsequently,

the SEM process showed a statistically significant positive effect of formality on

satisfaction. Thus, the results of the SEM affirmed the literature, not the exploratory

study and, as a result, H10 is rejected.

7.3.10.1 H5a- Formality of communication from the winery negatively influences


The results of H9 and H10 were rejected based on the result of the main model that

showed a statistically significant, positive relationship between the multi-dimensional

estimation of relationship quality and the formality of communication. Similar to the

results of the main model, the result of the alternative model in relation to the construct

of communication formality (H5a) showed a statistically significant positive link.

Therefore, the null hypothesis is accepted and H5a is rejected. As such, the main model

and alternative model have shown a statistically significant negative link between the

constructs.

This appears to be a confounding result, and may be because of a fault in the hypothesis

generation stage of the study. The hypothesis was stated negatively based on the result

of the exploratory, qualitative in-depth interviews. However, the formulation of the

hypotheses was done on the basis of comments from one participant of the exploratory

phase of the study which can be considered minimal and as such is listed as a limitation

of the study in 7.7. Mohr et al. (1996) state that formality of communication positively

influences satisfaction. In the hypothesis formulation stage of the study, the researcher

was of two minds as to whether to base the hypotheses on the literature or the results of

the exploratory study. The researcher decided to base the hypothesis on the findings of

the exploratory study, as it would be more contextually accurate as recommended by

the literature (Leedy & Ormrod, 2010). Therefore, the hypothesis would be stated in

141

terms of the context of the study (grape growers‟ opinions), as opposed to the Mohr et

al. (1996) study which was performed on a generic business context.


decreasing growers trust in the winery.

As discussed by Cox et al. (2001), Gaski, (1984) and Seyed-Mohammed & Wilson,

(1990), the power asymmetry in a relationship will have a negative effect on the actor‟s

(not holding the power) perception of the relationship. The discussion from wine

industry literature and the exploratory study showed that there was a power asymmetry

favouring the winery. The subsequent SEM process showed a statistically significant

negative effect of power on growers‟ perceptions of trust in the winery. Therefore, the

null hypothesis is rejected and H11 is affirmed.


decreasing growers‟ satisfaction with the winery.

In line with the literature discussion and exploratory study results highlighted in 7.3.10,

a negative influence of power asymmetry, favouring the winery, was posited. The

subsequent SEM process showed a statistically significant, negative linkage between

power and satisfaction; therefore, the null hypothesis is rejected and H12 is affirmed.

The results relating to H11 and H12 illustrate a high power asymmetry favouring the

winery and this power asymmetry is decreasing relationship quality from the growers‟

perspective.

7.3.12.1 H6a- Power asymmetry in the relationship, favouring the winery is

decreasing grape growers perceptions of relationship quality.

Unlike the results of previous hypotheses, for example H2 and H4, H11 and H12 were

both affirmed. However, it was still of interest to observe the results of the alternative

model which showed a statistically significant effect of power asymmetry decreasing

grape growers‟ perceptions of relationship quality. Therefore, the null hypothesis was

rejected and H6a is affirmed. As such, the results of the alternative model mirror those

of the main model in relation to this construct in that power asymmetry negatively

affects relationship quality.

142

7.4 Cluster analysis results discussion The results of the SEM process showed, apart from other issues related to

communication, that there is a strong power asymmetry favouring the winery. To

observe the effect of this power asymmetry, K-means cluster analysis was performed to

investigate the interaction between active variables such as power, trust and satisfaction

(relationship quality) and passive variables such as the characteristics of the trading

relationship between the two, and the business demographics of the growers and the

wineries. This process was performed to observe if there was any commonality between

the relational dimensions and aspects of the winery and grower business.

Three clusters of relationships were identified.

7.4.1 “Unsustainable Relationship” cluster

Firstly, an “unsustainable relationship” cluster was observed. This cluster was

associated with low relationship quality (low trust and satisfaction) and high power

asymmetry favouring the winery. The elements of the trading relationship of interest

were the very low price per tonne received by the grower, with an average of $692 per

tonne, and a longer than average length of contract, with most of the growers located in

the warm climate areas such as the Riverland and the Riverina area. This price per

tonne can be considered extremely low, and below the cost of production per tonne for

some growers due to increased water and other farming costs (such as fertiliser) (Stone

pers. Comm., May 2010). Furthermore, this price is only slightly above the average

price per tonne for grapes of the 2009 growing season, which was $529 (ABARE,

2010). It must be noted that in the questionnaire, growers were asked to focus on the

most important relationship they had with a winery (noting that growers had

relationships with more than one winery).

In this cluster, the most important relationship was one where the money they received

would barely cover the cost of growing the grapes and, as a consequence, the growers

were making only a small financial profit, or even a loss. Fourteen percent of the

respondents of the whole study were therefore making a small profit or loss from their

grape growing businesses, and consequently the best relationship they had (in terms of

price) was financially inadequate.

143

Of interest was that most of the growers in this cluster were in warm climate wine

regions which are generally producers of lower quality grapes, and as a result wine, and

that these regions are not the focus of recent marketing plans by the major Australian

wine promotion bodies. These plans involve focusing marketing efforts on higher

quality wines (AWBC, 2007) and, therefore, brings into question the overall long term

viability of these regions.

Furthermore, the majority of the relationships in this cluster were with large, publicly

owned wineries, and this may have power asymmetry implications. As highlighted in

Chapter 5, the average respondents‟ businesses had less than three employees and, as

such, have little clout in changing their business procedures; they must adapt to the

wishes of the larger corporation as affirmed by Chwelos et al. 2001 and Kurokawa et al.

2008.

7.4.2 “OK relationship” cluster

Secondly, a cluster of relationships termed the “OK relationship” was identified. The

growers in this cluster experience less power asymmetry (i.e. possessed more power in

the relationship) and higher levels of relationship quality than the “unsustainable

relationship” cluster. Of great interest was the price per tonne received by this cluster,

which was an average of $1264. This price was almost double that of the

“unsustainable relationship” cluster and this cluster spent the shortest length of time in

the relationship compared to the other clusters. This cluster had more relationships with

privately owned, small to medium sized (SME) wineries compared to the

“unsustainable relationship” cluster, thereby suggesting the alliance nature of two SME

actors (as most respondents‟ businesses had fewer than three employees), where

bonding behaviour and social bonds are built between the two parties, and each become

loyal to each other (Achrol & Gundlach, 1998; Duncan & Moriarty, 1998). This is in

opposition to the cultural and social distance that exists when SME and large

corporations interact which can lead to a decrease in bonding behaviour (Andersen et

al. 2009).

Therefore, in view of this cluster analysis, 55% of the respondents‟ best relationship

was part of the “OK relationship” cluster.

144

7.4.3 “Good Relationship” cluster

Thirdly, a cluster of relationships termed the “Good relationship” was identified. This

cluster had the highest levels of relationship quality compared to the other two clusters

and, correspondingly, the power asymmetry in this cluster was lower than was the case

with the other two clusters. Therefore, respondents in this cluster possessed the most

power in relationships compared to the other clusters. This cluster received the highest

average price per tonne of any of the clusters ($1981) and the members of this cluster

were primarily located in cool growing regions. The members of this cluster had

relationships with more SME wineries than any of the other clusters.

7.4.4 Questionnaire item results discussion, by cluster

The results of the questionnaire item, particularly the mean scores of the items, were of

interest in relation to the scores given by each cluster. There were a number of great

discrepancies between the mean results, and the most notable are discussed.

The questionnaire item “Trust 1” had a mean score of 1.83 for the “Unsustainable

Relationship” cluster, and 3.66 and 5.89 for the “OK Relationship” and “Good

Relationship” clusters, respectively. The questionnaire item read that “when things go

bad, we believe the winery will offer us support and assistance”. Therefore, the

“Unsustainable Relationship” cluster believed that they would be offered minimal

support when things went bad, with the other two clusters receiving more support, and

the “Good Relationship” cluster receiving the most support. This result shows that the

relationships that have higher levels of relationship quality receive more assistance and

support during difficult times.

Questionnaire item “Trust 2” had a mean score of 1.50 for the “Unsustainable


Relationship” clusters, respectively. The questionnaire item read that “when making

important decisions, the winery is concerned about the respondent‟s welfare”.

Therefore, the “Unsustainable Relationship” cluster believed that the winery had little

concern for their welfare when making decisions, with the “OK Relationship” and

“Good Relationship” cluster believing that the winery was more concerned with their

welfare, with the “Good Relationship” cluster receiving the most care. The result was

that this questionnaire item showed that respondents who experienced a higher level of

145

relationship quality had relationships with wineries that were concerned about their

welfare when making decisions.

The questionnaire item “Trust 6” had a mean score of 1.78 for the “Unsustainable


Relationship” clusters, respectively. The questionnaire item read that when wineries

offered an unlikely explanation, the respondents believed that the winery was telling the

truth. In view of the results, the “Unsustainable Relationship” cluster believed that the

winery was lying (i.e. not telling the truth) when giving an explanation, with the “OK

Relationship” and “Good Relationship” clusters believing that the winery was telling

more of the truth, and the “Good Relationship” cluster experiencing the most truthful

responses when given an explanation. Therefore, respondents that had relationships

higher in relationship quality involved wineries that told more truth when giving

explanations.

Questionnaire item “Power 1” had a mean score of 5.91 for the “Unsustainable


Relationship” clusters, respectively. The questionnaire item read that the respondents

had to follow the winery‟s instructions or they would get their grapes from somewhere

else, with a higher mean score illustrating the respondent had to follow the winery‟s

instructions or be discarded, and vice versa. Therefore, the “Unsustainable

Relationship” had to follow the winery‟s instructions or be discarded; this was less of a

requirement for the “OK Relationship” and was the least for the „Good Relationship”

cluster. Thus, respondents who had relationships that contained higher levels of

relationship quality had greater power in the relationship and consequently feared less

the possibility of being discarded if they did not follow the winery‟s directions

explicitly. It appeared that respondents who did experience higher levels of relationship

quality could set their own agendas, to a certain degree, and possibly be more able to

use their own initiative when growing their grapes.

Questionnaire item “Power 4” had a mean score of 5.82 for the “Unsustainable


Relationship” clusters, respectively. The item read that if the winery wanted to, it could

severely punish the respondent if he/she was uncooperative. Therefore, it appeared that

the “Unsustainable Relationship” cluster could be highly punished. The “OK

Relationship” and “Good Relationship” cluster were likely to be punished to a lesser

degree, with the “Good Relationship” cluster likely to be punished least if they were

146

uncooperative. The results showed that respondents who had relationships with a higher

level of relationship quality would be punished the least if they were uncooperative.

7.4.5 Cluster results summary

The cluster analysis illuminated some interesting features regarding the relationships

between grape growers and wineries and the effect these relationships have on

relationship quality and the influence of power asymmetry. The influence of growing

region, price per tonne and the size of the winery on relationship quality and power

asymmetry was also of interest. In view of the cluster analysis findings, for a grape

grower to attain higher levels of relationship quality and power in the relationships they

must:

1. be located in a cooler climate wine region;

2. have short relationships with wineries; and

3. deal with SME wineries.

In view of this study‟s results, if grape growers achieved these three objectives, they

would potentially attain higher prices for their grapes.

Furthermore, taking into account the mean score of questionnaire items by cluster that

showed high discrepancies between means, respondents who experienced relationships

that contained a higher level of relationship quality:

1. received more support from the winery during difficult times;

2. had wineries who were more concerned for the respondents‟ welfare when

making decisions;

3. were involved wineries that told more of the truth;

4. were allowed to work more without having to follow explicit instructions;

5. had a reduced fear of retribution for not following instructions; and

6. experienced less punishment if they were uncooperative.

.

7.5 Research Question Summary Three research questions were devised for this study. The following sections of this

chapter discuss the research questions, incorporating how the questions were answered.

147

7.5.1 Question 1: Which relational constructs constitute relationship quality?

The concept of relationship quality has been widely discussed in marketing literature.

However, there was much conjecture as to the antecedents of relationship quality and

no specific consensus has been reached by the various authors. A detailed analysis of

the literature in Chapter 2 revealed that relationships that are high in trust and

satisfaction are high in relationship quality. Therefore, this research question was

answered by stating that the relational constructs of trust and satisfaction led to

relationship quality. The study also examined two perspectives of relationship quality,

namely a multi-dimensional and uni-dimensional perspective. Both perspectives

conceptualise that relationship quality is comprised of trust and satisfaction; however,

in the multi-dimensional conceptualisation, proposed by Crosby (1990), Dorch (1998)

and Kim et al. (2006), relationship quality is perceived as two separate constructs (trust

and satisfaction ). If a relationship is high in those two constructs, it is high in

relationship quality. This perspective of relationship quality was examined in the main

SEM model. The study also examined the uni-dimensional nature of relationship

quality, proposed by Stern & Scheer (1992) and Leuthesser (1997), and is tested in the

alternative SEM model, by which relationship quality is measured as a single construct

with latent variables consisting of trust and satisfaction.

7.5.2 Question 2: Which elements of the grape grower/ winemaker relationship

affect grape growers‟ perceptions of relationship quality?

The exploratory study, highlighted in Chapter 3, involved qualitative interviews with

grape growers. The purpose of those interviews was to identify the factors in the

relationship that affected relationship quality, and to provide weight to the construction

of the conceptual model. The interviews uncovered that communication was important,

particularly the dimensionality of communication, and had an effect on relationship

quality. Furthermore, the construct of power also affected the relationship. The

quantitative phase of the study tested the relationship between these elements and

observed the effect between them. In combining the effect that the communication

elements and power had on trust and satisfaction (in the SEM process), the following

observations were made.

1. Direct communication positively influences relationship quality.

2. Indirect communication negatively influences relationship quality.

3. Uni-directional communication positively influences relationship quality.

148

4. Non-coercive communication attempts negatively influence relationship quality.

5. Formality of communication positively influences relationship quality.

6. Power asymmetry negatively influences relationship quality

The notion of power asymmetry favouring the winery is particularly evident in the

study and can be shown in the SEM models presented in Chapter 6. In both the main

and alternative models, the beta coefficient of the regression analysis (the affect

dimension) between the construct of power asymmetry and relationship quality has the

strongest effect in the model. In the main model, the beta coefficient for the paths

between power asymmetry and trust and satisfaction are the strongest in the main

model with -0.402 and -0.341, while the same is true in the alternative model with a

beta coefficient of -0.402 between power asymmetry and relationship quality. As such,

in both models the effect of power asymmetry is the strongest of all affects. This result

quantitatively shows that power asymmetry is a major factor in the relationships

between grape growers and wineries in the Australian wine industry.

7.5.3 Question 3: Are there any commonalities between wine grape growers in

their perceptions of relationship quality?

K- Means cluster analysis was performed in the quantitative phase of the study to

observe if there were any commonalities between grape growers in their perception of

relationship quality. Commonalities were found, mainly based on the nature of the

trading relationships with wineries, the regions in which the grape growers were

located, and the size and ownership of the wineries that growers were trading with. The

main commonalities were that:

1. Grape growers in cooler climate wine regions experienced higher levels of


2. Grape growers who traded with smaller sized wineries experienced higher

levels of relationship quality.

3. Grape growers who traded with privately owned wineries, as opposed to

publicly owned wineries, experienced higher levels of relationship quality.

4. Grape growers who received a higher price per tonne for their grapes

experienced higher levels of relationship quality.

5. Grape growers who had shorter contracts with wineries experienced higher

levels of relationship quality.

149

7.6 Conclusion The wine industry is important to the economy of Australia. Not only does the industry

generate vast sums of income from domestic and export sales, but it employs tens of

thousands of people directly and indirectly through allied industries that support the

wine industry.

The industry is currently suffering economic hardship, mainly due to an oversupply of

grapes, and relationships between grape growers and winemakers have become

adversarial, resulting in a break down in the relationship between the two actors. The

industry may return to a sustainable level of grape supply and, if this occurs, the

relationship between the two actors must become fruitful in order to produce wine fit

for the market.

Furthermore, the wine industry is restructuring its grape production abilities to

concentrate on the production of higher quality wine, and to emphasise the regionality

of wine products in marketing efforts, particularly in export markets. Quality of wine

and regionality of wine products is grape grower derived; therefore, the grape grower

plays a vital role in the future prosperity of the wine industry. As such, the grape

grower perspective of the relationship between the two actors is of importance.

This study attempted to ascertain the relational factors that are of importance to the

grape grower and that affect relationship quality .The exploratory study highlighted that

the dimensionality of communication and the power asymmetry in the relationship,

favouring the winery, was influencing relationship quality. The results of the causal

study highlighted that face-to-face and direct email communication positively affected

relationship quality, while non-direct modes (such as seminars and newsletters)

negatively affected relationship quality and that the power asymmetry was leading to

decreased grape prices and lower relationship quality. It appeared that the price of the

grape growers‟ produce (grape) had a direct correlation with relationship quality,

whereby the higher the price they received, the higher the level of relationship quality

they experienced. Further analysis as part of the causal study highlighted the effect that

the size and ownership of the winery had on relationship quality, with growers dealing

with larger, publicly owned wineries experiencing lower levels of relationship quality.

Grape growers that had their businesses in warmer climate regions, as opposed to

cooler climate regions, also experienced lower levels of relationship quality.

150

If the Australian wine industry is to find prosperity in the future, it must invest in the

relationships between grape growers and wineries by focusing on the needs and wants

of the actor that has the greatest impact on the core quality of the end product, so that

the end product is fit for market. This study attempted to uncover these needs and wants

via a multistage research process, which is illustrated in pages of this dissertation.

7.7 Study Limitations This study has various limitations associated with the sample of respondents and the

scope of the study. As previously discussed in Chapter 3, the exploratory, qualitative

study contained 13 respondents for the in-depth interviews. This number is small and

the respondents were only from two states of Australia (South Australia and Victoria),

and it can be posited that they are not representative of grape growers Australia wide.

Furthermore, respondents in the qualitative phase of the study were not located in the

two largest respondent localities in the quantitative phase of the study. These two areas

were the Riverland and Riverina wine regions. As a result, data was not obtained in the

qualitative research phase from these regions and is, therefore, a limitation of the study.

The causal, quantitative study involved the responses of 396 grape growers from all

states of Australia; however, not all wine regions in Australia had respondents

contained in the 396 and, therefore, that presents a limitation. The respondents were

asked to answer the questionnaire items in view of their most important relationship. A

limitation of the study is that not all of the relationships that respondents had with

wineries were recorded. Respondents had an average of two other relationships with

wineries, as shown in Chapter 5; however, limitations in regard to time and length of

response related to respondent fatigue, made this so. In addition, the formulation H, in

regards to formality of communication, was based on the comments of two IDI

participants and it was later found in the causal stages of the research that an opposite

effect was observed (negative as opposed to positive effect on relationship quality).

The formulation of the hypothesis was flawed due to the small number of responses (2)

and as such is a limitation of the study. In comparison, the formulation of H7 and H8

was based on one comment from a participant; however, the hypotheses were affirmed.

In any case, developing the hypotheses on one response is a limitation of the study.

The questionnaire was distributed via an online mode, as it was deemed cost effective

in administration and could be accessed by respondents, as expert opinions suggested

151

that respondents had access to internet services. However, the rural nature of the

respondents does suggest that some potential respondents would not be able to

complete the survey due to sporadic internet connections, or having no internet

connection at all. Therefore, a different mode of survey administration, such as paper-

based, mail administration or face to face administration, may have led to more valid

responses.

The SEM process of this study utilised Partial Least Squares Regression to estimate the

paths between constructs. This method was used due to the small number of indicator

variables for certain constructs (i.e. less than three for certain constructs). However, this

method was subject to some criticisms for not being as robust as other methods such as

those employed by AMOS or LISREL, which use maximum differences to test paths

(Chin, 1998; Hair et al. 2006; Ringle et al. 2005), and as such can be considered a

limitation of this study. Also, the latent variables for certain constructs, particularly

“satisfaction”, are small in number (fewer than four for a dependent variable), and as

such may be providing an AVE (average variance extracted) that is small and this may

account for the statistically insignificant results shown in H2, H4 and H8. The

Cronbach Alpha for each latent variable is passable (Cronbach, 1970); however, the

small number of variables may have led to the insignificant results and as such is a

limitation of the study (Hair et al, 2006; Ringle, 2005). If the dependent variable of

“satisfaction” contained more reliable latent variables in terms of higher AVE‟s and

Cronbach alphas, the results of H2, H4 and H8 may have been significant.

Cluster analysis was performed in the final phase of the study to uncover the types of

relationships that grape growers had with wineries. The analysis showed types of

relationships; however, the number of responses that made up the cluster (396

responses) is small for cluster analysis purposes. Even though cluster analysis is an

exploratory research method (Janssens et al. 2008), it is usually performed on larger

sample sizes in order to gain robust results. However, ANOVA results performed on

the passive variables in the cluster analysis showed statistically significant differences

between the clusters and therefore the clusters were sound. Nevertheless, the smaller

sample size utilised in the cluster analysis is a limitation of the study.

152

7.8 Recommendations for further research The results of this study have highlighted a number of areas for further research. This

study primarily focused on the grape grower perspective of communication

dimensionality, power and relationship quality in regards to the relationship between

themselves and their buyers (the wineries). While justification was given for this, the

winery‟s perspective is also important to view, and further research could test the

constructs highlighted in this study on wineries. A model incorporating both

perspectives of the relationship could then be attained.

The respondents of this study were asked to answer questionnaire items in relation to

the most important relationship that they had with wineries. Further research could be

performed investigating all the relationships that growers had with wineries and SEM

could be performed examining the differences between those relationships and the way

that the constructs examined in this study differ between relationships. The results of

the study have shown an interaction between price, power and satisfaction, whereby

price of product (in this study, grapes) moderates the effect that power asymmetry has

on satisfaction. This moderating effect could be examined in further research.

The study uncovered the effects that the size and the ownership of the winery had on

relationship quality. It was shown that, when dealing with small to medium sized,

privately owned wineries, growers experienced less power asymmetry (they had more

power in the relationship) and relationship quality was higher. It appeared that buyer

size and ownership moderated the effect that power had on relationship quality, and this

concept requires further investigation.

The interaction of elements of the relationship between the two (for example, price per

tonne of grapes, length of the relationship, size and ownership of the winery, wine

region of the grower) and relational constructs (e.g. satisfaction, power, trust) was

observed in this study. However, further research could be performed, investigating

how grower-specific characteristics affect the relational constructs such as the type of

grape (red or white) or variety of grape (for example, Chardonnay, Riesling, Shiraz)

that is produced by the grape growers or how the yield per acre of the grapes produced

has an effect or mediates an effect between relational variables. This investigation

would be of interest as it is known that certain varieties are renowned in certain wine

regions (for example, Coonawarra is known for Cabernet Sauvignon production), and if

growers produce grapes that are renowned in a region, do they experience less power

asymmetry and higher levels of relationship quality (Domine, 2000)?

153

The study focused on the interaction between price per tonne and power asymmetry;

however, it did not consider the interaction between yield of grapes per acre and power

asymmetry. It is known that lower yield per acre creates a higher quality of grapes

(Domine, 2000) but if grape growers produce lower yield (higher quality grapes) as

opposed to higher priced grapes, do they experience a higher level of relationship

quality and lower power asymmetry? Therefore, the concept that if yield per acre

increases, power asymmetry increases, requires further investigation.

Further analysis of questionnaire items illuminated a number of areas for future

research, particularly how demonstrative aspects of trust and power asymmetry affect

relationship quality. For example, future research could focus on the effect that a

partner‟s assistance during difficult economic times has on the other partner‟s

perceptions of relationship quality. Furthermore, a business‟ concern for the welfare of

their partner when making decisions, and its effect on relationship quality perceptions

of the partner, could be observed. The effect of dishonesty on partner relationship

quality perception could also be observed. In relation to power asymmetry, the effect of

partner initiative on power asymmetry, observed in addition to the effect that power

asymmetry has on the level of punishment used by the business, could be examined.

7.9 Study contribution This study has provided a contribution to both academic marketing and Australian wine

industry literature. Firstly, the study has provided an insight into communication in

B2B relationships and shows that communication can be direct or indirect and focused

on specific modes. The study further examined Mohr & Nevin‟s (1990) concept of

collaborative communication, and provided an extension of this theory, but focussing

on modality and uni-directionality. The study also highlighted the effect of power

asymmetry on relationship quality. This study also aided the Australian wine industry

by highlighting the state of various grape grower and winery relationships, by providing

clusters of relationships, and by providing various industry statistics such as grape

grower and winery business details (e.g., place of business, size). As mentioned in 1.5,

this study attempted to extend Hobley‟s (2007) work by further investigating the effect

that communication (and its elements) and power asymmetry had on the relationship

between the two actors. The study differed from Hobley (2007) by focusing on

individual communication elements (whereas Hobley, 2007 focused on communication

as a single construct), and power asymmetry and its overall effect on relationship

154

quality. Finally, a major contribution of the study was to show the effect that lowering

grape prices is having on grape growers‟ perceptions of relationship quality.

This study has also added to the knowledge of communication in an agribusiness

context by applying the Mohr & Nevin (1990) and Mohr et al, 1996 collaborative

communication framework to the context, and has as such extended parts of Storer‟s

(2005) inter-organisational information management systems (IOIMS) by examining

different elements of communication and their effects on agribusiness relationships

such as computer based modes and the formality and non-coercive abilities of

communication. As such, the study has added to Storer‟s (2005) work by examining

communication between business actors in an agribusiness context.

This study differs from other studies that have observed the relationship between grape

growers and wineries (such as Scales et al. (1995), Anderson, 2001 and Hobley, 2007),

by observing the concept of relationship quality, in particular, the effect that the price

of grapes and communication elements have on relationship quality. The study has also

contributed to knowledge by quantitatively observing the relationships in the wine

industry, which is different from studies by Benson-Rea (2005) and Rampersad (2008)

which were mainly qualitative in nature. Furthermore, the study satisfied a need for

research into communication between grape growers and wineries as proposed by

Spawton & Walters (2003), Chong (2007) and Brown (2008), and showed that

communication from the grape grower to the winery is limited and uni-directional, and

that direct forms of communication such as face-to-face and personal email

communication have better effects than indirect forms such as seminars and

newsletters. These results appear to highlight what is discussed in the academic

literature, such as that by Daft & Lengel (1984), which shows that personal or rich

forms of communication have better outcomes, and add to their work by highlighting

the effect of electronic communication forms, such as emails or communications on the

internet, which were not examined by Daft & Lengel (1984) due to those forms of

communication not being present in the past.

This study also stands out from these other studies by quantitatively testing the effect of

power asymmetry on the relationship between the two actors. However, the other

studies were performed when grape oversupply was not as prevalent and therefore not

justified. The study has shown that the Australian wine industry grape grower and

winery relationship is different in terms of power asymmetry from other wine industries

where the power asymmetry is favouring the grape grower. This is evident in the

155

Champagne region where, due to a limited number of grape growers, they have a power

asymmetry over the Champagne houses (wineries) (Charters & Menival, 2010). This

study also illustrated three types of grape grower relationships with wineries based on

relationship quality and power asymmetry, a contribution that has not been provided

before.

7.10 Study implications for the Australian wine industry This study has provided a number of implications for Australian winery and grape

grower interactions. The main proposition to come from this study is that direct

communication (for example face-to-face and direct email modes) should be used by

wineries when interacting with grape growers instead of indirect communication

(seminars and newsletter modes). Lowered grape prices are also affecting grape

growers; however, this can be linked to issues outside of the control of the wineries and

may not be able to be changed by wineries. Nevertheless, the study has shown that

wineries can control their behaviour and need to be more truthful, give more support to

grape growers during these difficult times, and show concern for their grape growers

when making decisions. By recognising that these issues exist, wineries will improve

their relationships with grape growers, and therefore the sustainability of the Australian

wine industry. The results of the study could also aid policy makers in the Australian

wine industry. The study highlighted three types of relationships, of which one was

unsustainable. By taking note of the business characteristics of the unsustainable

relationship group, policy makers could target these businesses and give them the

necessary support to encourage them to leave the industry.

156

Appendix 1: Questionnaire

In regards to your most important winery business relationship

How many years have you been contracted to the winery (number of years) Approximately how many tonnes of grapes do you supply to the winery (number

of tonnes) Approximately what is the dollar amount of those grapes ($) What is your average price per tonne of grapes you supply to the winery ($) Do you sell your grapes to other wineries?

If so, how many other wineries (number of wineries) What proportion (%) of your grapes do you sell to the other wineries(%)

157

For each of the following methods, over the 2009 Vintage growing season (August 08- May 09), please estimate the frequency (the number of times) with which the winery communicates with you via these various methods. Please type in the "number of times" as a number, e.g. "4" rather than "four". If you did not communicate via a certain method, please put "0"

Face to face interaction with winery people (number of times) (Required) Telephone interaction (telephone calls) with winery people (number of times)

(Required) Written letters and all written correspondence (non-electronic e.g. no email)

(number of times) (Required) Direct Email, from a wine representative to you(number of times) (Required) Seminars [e.g. Grower Days (winery - growers meetings)] (number of times)

(Required) General newsletters from the winery (number of times) (Required) Other (number of times)

158

The next few questions relate to the formality of communication between you and the winery. When you liaise with the winery, there are formal and informal methods of communication. For example, if communication is formal it is done on a regular basis and is written down, whereas informal communication is generally verbal (in words) and not done on a regular basis.

Please indicate how strongly you agree on the following statements

Strongly Disagree

1

2 3 4 5 6 Strongly

Agree 7

When working with this winery, formal communication channels are followed (i.e. communication is formal, regular and structured) versus casual informal, word –of-mouth modes).

The terms of our business contract with the winery have been written down in detail.

The winery‟s expectations of us are communicated in detail. The terms of our business relationship with the winery have been explicitly put into words and discussed.

Information sharing on important issues has become crucial to maintaining this partnership.

We share a common, specialised IT software system dedicated to facilitate communication with the winery (e.g. Vine Access®).

Grower liaison committees, that communicate my issues and concerns with the large wineries, are effective.

159

The next few questions are regarding the feedback that the winery provides to you and vice versa.

Please indicate by clicking the box that corresponds with your answer

None

1 2 3 4 5 6 A lot

7 How much positive feedback do you provide to this winery? How much negative feedback do you provide to this winery? How much negative feedback does this winery provide to you? How much positive feedback does this winery provide to you?

In their interaction with you, the winery often tries to influence YOUR attitudes and behaviours. Please estimate the frequency with which the winery‟s employees (e.g. winemakers, grower liaison staff, viticultural staff) use the following methods to influence YOU.

Very

infrequently 1 2 3 4 5 6

Very frequently

7 How frequently did the winery‟s employees make a recommendation that by following their suggestions, your business would be more profitable.

How frequently did the winery‟s employees ask you to perform a certain operation, but didn‟t say what penalty may occur if you didn‟t do what they asked.

How frequently did the winery‟s employees say you will be supplying grapes of a certain quality, but didn‟t give you specific information e.g. what crop level they would like, what spray regime they would like or other directions they would like you to take to

160

grow those grapes.

The following question are about trust in your business relationship with the winery. Please indicate how strongly you agree with the following statements:

Strongly Disagree

1 2 3 4 5 6 Strongly

Agree 7





161

Strongly Disagree


Agree 7



The winery has often provided us information that has later proven to be incorrect.

The winery keeps the promises that it makes to our business. Whenever the winery gives us advice on our business operations, we know that it is sharing its best advice.


The following questions are about how satisfied you are with the business relationship you have with the winery.


Strongly Disagree


Agree 7



The relationship our business has with the winery has been an unhappy one.

162

The following questions are about power in the relationship. By power we mean the ability to influence another person‟s actions.


Strongly Disagree


Agree 7


We are expected to follow the winery‟s instructions. We have influence over the winery‟s actions. The winery can, if it wanted to, severely penalise us if we are uncooperative.

If we did not want to follow the winery‟s instructions or plans we could sell our grapes to another winery.

163

We would like to find out some details regarding you, your business and the winery. Please rest assured that all information is kept strictly confidential and is not passed onto any organisation. No one is named in person or identified in anyway If you are asked to give a number e.g. 5 acres, type in "5" rather than "five" Please list the size in acres of your vineyards (number of acres) (Required) How many years have you been growing grape vines (number of years)

(Required) How many people work for your grape growing business (number of people)

(Required) Please indicate which wine region you are located in (Required) Wine Region Do you have any formal grape growing qualifications? If so list

164

In what wine region is the winery that you supply grapes to (i.e. the winery you have focussed on and

discussed in the survey) (Required) Wine Region Is the winery (that you have focussed on and discussed in this survey)Click on the box (Required) In your opinion is the size of the winery (Click on box) (Required)

Thank you for completing this survey! To be in the running to win $2000 worth of Viticultural services from Davidson Viticulture, please enter your name (including business name), address and phone details below. Please take note that all of your details are kept in strictest confidence. Your details will not be handed on to anyone

165

Appendix 2: Cluster Analysis Results

166

Descriptive statistics of active variables, by cluster

N Mean Std. Deviation Std. Error

95% Confidence Interval for Mean

Minimum Maximum

Between-

Component

Variance Lower Bound Upper Bound

Trust Unsustainable Relationship 54 -1.7757443 .65463284 .08908425 -1.9544247 -1.5970639 -2.59249 .29294

OK Relationship 219 -.0261752 .56347996 .03807645 -.1012203 .0488699 -1.90935 1.27316

Good Relationship 123 .8261997 .63169695 .05695820 .7134453 .9389542 -.65120 2.40560

Total 396 .0000000 1.00000000 .05025189 -.0987946 .0987946 -2.59249 2.40560

Model Fixed Effects .59815865 .03005860 -.0590958 .0590958

Random Effects .68302098 -2.9388021 2.9388021 1.10619863

Satisfaction Unsustainable Relationship 54 -1.8519137 .85781989 .11673450 -2.0860534 -1.6177739 -3.29015 -.15743 OK Relationship 219 .0080012 .51075600 .03451369 -.0600220 .0760245 -1.38496 1.58266 Good Relationship 123 .7987891 .53894408 .04859495 .7025906 .8949877 -.85374 1.58266 Total 396 .0000000 1.00000000 .05025189 -.0987946 .0987946 -3.29015 1.58266 Model Fixed Effects .57802539 .02904687 -.0571067 .0571067

Random Effects .69543275 -2.9922056 2.9922056 1.14698810

Power Unsustainable Relationship 54 .5858628 1.26153005 .17167250 .2415314 .9301943 -4.14183 1.43853

OK Relationship 219 .3817124 .58615600 .03960875 .3036473 .4597775 -1.51639 1.43853

Good Relationship 123 -.9368423 .79627588 .07179778 -1.0789732 -.7947114 -4.14183 .60245

Total 396 .0000000 1.00000000 .05025189 -.0987946 .0987946 -4.14183 1.43853

Model Fixed Effects .77591317 .03899110 -.0766572 .0766572

Random Effects .53856240 -2.3172470 2.3172470 .68548296

167

ANOVA results of active variables, by cluster

Sum of Squares df Mean Square F Sig.

Trust Between Groups (Combined) 254.387 2 127.194 355.494 .000

Linear Term Unweighted 254.051 1 254.051 710.050 .000

Weighted 236.465 1 236.465 660.896 .000

Deviation 17.922 1 17.922 50.091 .000

Within Groups 140.613 393 .358

Total 395.000 395 Satisfaction Between Groups (Combined) 263.693 2 131.847 394.617 .000


Weighted 238.244 1 238.244 713.063 .000

Deviation 25.450 1 25.450 76.171 .000

Within Groups 131.307 393 .334 Total 395.000 395

Power Between Groups (Combined) 158.398 2 79.199 131.551 .000


Weighted 130.747 1 130.747 217.173 .000

Deviation 27.651 1 27.651 45.929 .000

Within Groups 236.602 393 .602

Total 395.000 395

168

Bonferroni Test on active variables, by cluster

Dependent Variable (I) Power Sat Trust

Cluster 3

(J) Power Sat Trust

Cluster 3 Mean

Difference (I-J) Std. Error Sig.

95% Confidence Interval

Lower Bound Upper Bound

dimension1

Trust

dimension2

Cluster 1 dimension3

Cluster 2 -1.74956909* .09088219 .000 -1.9680746 -1.5310635

Cluster3 -2.60194404* .09764580 .000 -2.8367112 -2.3671769


Cluster 1 1.74956909* .09088219 .000 1.5310635 1.9680746

Cluster3 -.85237495* .06739921 .000 -1.0144210 -.6903289

Cluster3 dimension3

Cluster 1 2.60194404* .09764580 .000 2.3671769 2.8367112

Cluster 2 .85237495* .06739921 .000 .6903289 1.0144210

Satisfaction

dimension2


Cluster 2 -1.85991490* .08782321 .000 -2.0710658 -1.6487640

Cluster3 -2.65070280* .09435917 .000 -2.8775680 -2.4238376


Cluster 1 1.85991490* .08782321 .000 1.6487640 2.0710658

Cluster3 -.79078791* .06513064 .000 -.9473797 -.6341961

Cluster3 dimension3

Cluster 1 2.65070280* .09435917 .000 2.4238376 2.8775680

Cluster 2 .79078791* .06513064 .000 .6341961 .9473797

Power

dimension2


Cluster 2 .20415048 .11788960 .012 -.0792883 .4875892

Cluster3 1.52270512* .12666316 .000 1.2181724 1.8272379


Cluster 1 -.20415048 .11788960 .012 -.4875892 .0792883

Cluster3 1.31855464* .08742821 .000 1.1083534 1.5287559

Cluster3 dimension3

Cluster 1 -1.52270512* .12666316 .000 -1.8272379 -1.2181724

Cluster 2 -1.31855464* .08742821 .000 -1.5287559 -1.1083534

Cluster 1= Unsustainable Relationship, Cluster 2= OK Relationship, Cluster 3= Good Relationship

169

Descriptive statistics of passive variables, by cluster



Minimum Maximum

Between-

Component


How many

years

contracted

Cluster 1 54 10.50 10.062 1.369 7.75 13.25 0 56

Cluster 2 219 7.18 6.830 .461 6.27 8.09 0 70

Cluster3 123 9.98 9.604 .866 8.26 11.69 0 60

Total 396 8.50 8.367 .420 7.68 9.33 0 70

Model Fixed Effects 8.256 .415 7.69 9.32

Random

Effects 1.228 3.22 13.79 3.173

Average price

per tonne

Cluster 1 54 $692.69 $453.003 $61.646 $569.04 $816.33 $100 $2,000

Cluster 2 219 $1,264.91 $681.833 $46.074 $1,174.11 $1,355.72 $0 $4,000

Cluster3 123 $1,981.32 $1,097.062 $98.919 $1,785.50 $2,177.14 $250 $7,000

Total 396 $1,409.40 $916.237 $46.043 $1,318.88 $1,499.92 $0 $7,000

Model Fixed Effects $811.896 $40.799 $1,329.19 $1,489.61

Random

Effects $363.864 $-156.18 $2,974.98 $310,591.4

92

Years have

you been

growing grape

vines

Cluster 1 54 26.69 23.946 3.259 20.15 33.22 4 168

Cluster 2 219 17.81 11.549 .780 16.27 19.35 3 100

Cluster3 123 19.37 10.303 .929 17.53 21.21 3 50

Total 396 19.51 13.856 .696 18.14 20.88 3 168


Random

Effects 2.460 8.92 30.09 13.277

170

Descriptive statistics of passive variables, by cluster



Minimum Maximum

Between-

Component


How many

years

contracted

Cluster 1 54 10.50 10.062 1.369 7.75 13.25 0 56

Cluster 2 219 7.18 6.830 .461 6.27 8.09 0 70

Cluster3 123 9.98 9.604 .866 8.26 11.69 0 60

Total 396 8.50 8.367 .420 7.68 9.33 0 70


Random

Effects 1.228 3.22 13.79 3.173

Average price

per tonne

Cluster 1 54 $692.69 $453.003 $61.646 $569.04 $816.33 $100 $2,000

Cluster 2 219 $1,264.91 $681.833 $46.074 $1,174.11 $1,355.72 $0 $4,000

Cluster3 123 $1,981.32 $1,097.062 $98.919 $1,785.50 $2,177.14 $250 $7,000

Total 396 $1,409.40 $916.237 $46.043 $1,318.88 $1,499.92 $0 $7,000

Model Fixed Effects $811.896 $40.799 $1,329.19 $1,489.61

Random

Effects $363.864 $-156.18 $2,974.98 $310,591.4

92

Years have

you been

growing grape

vines

Cluster 1 54 26.69 23.946 3.259 20.15 33.22 4 168

Cluster 2 219 17.81 11.549 .780 16.27 19.35 3 100

Cluster3 123 19.37 10.303 .929 17.53 21.21 3 50

Total 396 19.51 13.856 .696 18.14 20.88 3 168


Random

Effects 2.460 8.92 30.09 13.277

Cluster 1= Unsustainable Relationship, Cluster 2= OK Relationship, Cluster 3= Good Relationship

171

ANOVA results of passive variables

Sum of Squares df Mean Square F Sig.

How many years contracted Between Groups 863.877 2 431.938 6.337 .002

Within Groups 26786.621 393 68.159

Total 27650.497 395 Average price per tonne Between Groups 7.254E7 2 3.627E7 55.025 .000

Within Groups 2.591E8 393 659175.592 Total 3.316E8 395

Years have you been

growing grape vines

Between Groups 3413.208 2 1706.604 9.261 .000

Within Groups 72419.769 393 184.274

Total 75832.977 395

172

Bonferroni Test on passive variables, by cluster

Bonferroni

Dependent Variable (I) Power Sat Trust

Cluster 3

(J) Power Sat Trust

Cluster 3 Mean

Difference (I-J) Std. Error Sig.

95% Confidence Interval

Lower Bound Upper Bound

How many years contracted

dimension2


Cluster 2 3.317* 1.254 .026 .30 6.33

Cluster3 .524 1.348 .032 -2.72 3.76


Cluster 1 -3.317* 1.254 .026 -6.33 -.30

Cluster3 -2.793* .930 .009 -5.03 -.56

Cluster3 dimension3

Cluster 1 -.524 1.348 .032 -3.76 2.72

Cluster 2 2.793* .930 .009 .56 5.03

Average price per tonne

dimension2


Cluster 2 $-572.228* $123.357 .000 $-868.81 $-275.64

Cluster3 $-1,288.632* $132.537 .000 $-1,607.29 $-969.98


Cluster 1 $572.228* $123.357 .000 $275.64 $868.81

Cluster3 $-716.404* $91.483 .000 $-936.35 $-496.45

Cluster3 dimension3

Cluster 1 $1,288.632* $132.537 .000 $969.98 $1,607.29

Cluster 2 $716.404* $91.483 .000 $496.45 $936.35

Years have you been

growing grape vines

dimension2


Cluster 2 8.872* 2.063 .000 3.91 13.83

Cluster3 7.311* 2.216 .003 1.98 12.64


Cluster 1 -8.872* 2.063 .000 -13.83 -3.91

Cluster3 -1.561 1.530 .024 -5.24 2.12

Cluster3 dimension3

Cluster 1 -7.311* 2.216 .003 -12.64 -1.98

Cluster 2 1.561 1.530 .024 -2.12 5.24

*. The mean difference is significant at the 0.05 level.

173

Climate of Growers‟ wine region, by cluster

Cool

NSW

Cool

QLD

Cool

SA

Cool

Tas

Cool

Vic

Cool

WA

Don’t

know

Warm

NSW

Warm

SA

Warm

Vic

Warm

WA TOTAL

Cluster 1 1 0 17 0 1 2 0 13 18 0 2 54

Cluster 2 13 5 64 4 31 10 24 31 27 6 4 219

Cluster3 8 1 56 2 22 8 1 16 5 2 2 123

Total 22 6 137 6 54 20 25 60 50 8 8 396 Wine Region of Grower, by cluster

Cluster 1= Unsustainable Relationship, Cluster 2= OK Relationship, Cluster 3= Good Relationship.

174

Ownership of winery, by cluster

Chi-Square Test of winery, by cluster test

Value df

Asymp. Sig. (2-

sided)

Pearson Chi-Square 27.511a 2 .000

Likelihood Ratio 26.895 2 .000

Linear-by-Linear

Association

24.920 1 .000

N of Valid Cases 365

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is

17.26.

C

r

o

s

s

t

a

b

C

o

u

n

t

175

Appendix 3: IDI discussion questions 1. Is communication important in your relationship with

wineries?

PROMPT: Are there any types or modes of communication that you

like?

PROMPT: What is the feedback from the winery like?

PROMPT: Do you have any other comments about the communication

you have with wineries?

2. How has the economic down turn in the industry

affected your business?

PROMPT: Has it affected your relationship with wineries?

PROMPT: How?

3. Any other issues you would like to discuss about your

winery relationships?

4. How is the vintage going?

PROMPT: Any water issues?

PROMPT: Any frost problems?

PROMPT: How is the drought effecting growing?

176

Bibliography

ABARE. (2009), Australian wine grape production projections to 2010- 2011,

Canberra, Australia.

ABARE. (2010), Australian wine grape production projections to 2011- 2013,

Canberra, Australia.

Achrol, R.S., (1997), Changes in the theory of inter-organisational relations in

marketing: Toward a network paradigm. Journal of the Academy of Marketing Science,

vol 25, no 1, pp. 56-71.

Achrol, R.S., & Gundlach, G.T., (1999), Legal and social safeguards against

opportunism in exchange, Journal of Retailing, vol 75, no 1, pp 107-124.

Anderson, E. & Weitz, B.A., (1989), Determinants of continuity in conventional

industrial channel dyads, Marketing Science, vol. 8, no. 3, pp. 310-323.

Anderson, J.C., (1995), Relationships in business markets: exchange episodes, value

creation and their empirical assessment', Journal of the Academy of Marketing Science,

vol. 23, no. 4, pp. 346-350.

Anderson, J.C., Håkansson, H. & Johanson, J., (1994), Dyadic business relationships

within a business network context, Journal of Marketing, vol. 58, no. 3, pp. 1-15.

Anderson, J.C, & Gerbing, D.W., (1988), Structural equation modelling in practice: a

review and recommended two-step approach', Psychological bulletin, vol. 103, no. 3,

pp. 411-423.

177

Anderson, J.C. & Narus, J.A., (1990), A model of distributor firm and manufacturing

firm working relationships, Journal of Marketing, vol. 54, no. 1, pp. 42-58.

Andersen, P.H., Christensen, P.R. & Damgaard, T., (2009), Diverging expectations in

buyer-seller relationships: Institutional contexts and relationship norms, Industrial

Marketing Management, vol 38, pp 814-824.

Anderson, K., (2001), Where in the world is the wine industry going' in Annual

Conference of the Australian Agricultural and Resource Economics Society, Centre for

International Economic Studies, Adelaide.

Australian Bureau of Statistics (ABS) (2005), Australian wine and grape industry, Cat

No. 1329.

Australian Bureau of Statistics (ABS) (2009a), Australian wine and grape industry, Cat

No: 1329.0, Canberra, Australia.

Australian Bureau of Statistics (ABS). (2009b), 2009 Vineyard Estimates, Cat No:

13290.55.002, Canberra, Australia.

AWBC, 2007, Wine Australia: Directions to 2025, An Industry Strategy for Sustainable

Success, Kent Town, Adelaide.

Axelsson, B. & Easton, G., (1992), Industrial Networks: A New View of Reality,

Routledge, London.

Babbie, E, (2004), Paradigms, theory and social research, the practice of social

research, Wadsworth/ Thomson Learning, Belmont

178

Bagozzi, R., (1994). Structural equation model in marketing research: principles of

marketing research, Blackwell Publishers, Oxford.

Bagozzi, R., & Yi, Y., (1988). On the evaluation of structural equation models. Journal

of the Academy of Marketing Science, vol 13, no 3, pp 989-1006.

Barnhill, J.A. & Lawson, W.M., (1980), Toward a theory of modern markets, European

Journal of Marketing, vol. 14, no. 1, pp. 50-60.

Batt, P., (2003), Building trust between growers and market agents. Supply Chain

Management: An International Journal, vol 8, no1, pp. 65-78.

Baumgartner, H., Homburg, C., (1996), Application of structural equation modeling in

marketing and consumer research: a review, International Journal of Research in

Marketing, vol. 13 no.2, pp.139-61.

Benson- Rea, M., (2005), Network strategy in the New Zealand wine industry: how

firms in an industry understand and use their business relationships, PhD dissertation,

The University of Auckland.

Bendapudi, N. & Leone, R.P., (2002), Managing business-to-business customer

relationships: following key contact employee turnover in a vendor firm, Journal of

Marketing, vol. 66 no.2, pp.83-101.

Bigne, E., & Blesa, A., (2003), Market orientation, trust and satisfaction in dyadic

relationships: a manufacturer-retailer analysis, International Journal of Retail &

Distribution Management, vol. 31, no 11/12, pp. 574-90.

179

Blankenburg -Holm, D,. Kent, E. & Johanson, J., (1996), Business networks and

cooperation in international business relationships, Journal of International Business

Studies, vol. 27, no. 5, pp. 1033-53.

Blau, P.M., (1964). Exchange and Power in Social Life. New York: John Wiley &

Sons.

Boyce, C. & Neale, P., (2006), Conducting in-depth interviews: a guide for designing

and conducting in-depth interviews for evaluation input, Pathfinder International,

Watertown, USA. Online accessed 3/5/09 URL:

http://www.vision2020.info.tt/pdf/Guidelines_Tools_Techniques/Tools/m_e_tool_serie

s_indept h_interviews.pdf

Bradley, M.F., (1977), Buying behaviour in Ireland's public sector, Industrial Marketing

Management, vol. 6, no. 2, pp. 251-258.

Brown, R., (2008), Whole of supply chain approach leads to business success,

Australian & New Zealand Wine Industry Journal, Vol 23, no 2, pp 4-5.

Brown, J. R., Dev, C.S & Lee, D.J., (2000). Managing Marketing Channel

Opportunism: The Efficacy of Alternative Governance Mechanisms, Journal of

Marketing, vol 64, no 2, pp. 51-65.

Brusco, S., (1986), Small firms and industrial districts: the experience of Italy. New

firms and regional development in Europe, pp. 184-202.

Cannon, J.P., & Perreault, W.D., (1999), Buyer-seller relationships in business markets,

Journal of Marketing Research, vol. 36, no 4, pp. 439-460.

http://www.vision2020.info.tt/pdf/Guidelines_Tools_Techniques/Tools/m_e_tool_series_indept%20h_interviews.pdf

http://www.vision2020.info.tt/pdf/Guidelines_Tools_Techniques/Tools/m_e_tool_series_indept%20h_interviews.pdf

180

Cannon, J.P., Achrol, R.S. & Gundlach, G.T., (2000), Contracts, norms and plural

forms of governance, Journal of the Academy of Marketing Science, vol. 28, no. 2, pp.

180-194.

Cannon, J. P., & Homburg, C., (2001), Buyer-supplier relationships and customer firm

costs,” Journal of Marketing, vol 65, pp. 29-43.

Carson, D., & Coviello, N. (1996), Qualitative research issues at the

marketing/entrepreneurship interface, Journal of Marketing Intelligence and Planning,

vol 14; no 6, pp. 51-58.

Cavana, R.Y., Delahaye, B.L. & Sekaran, U., (2001), Applied Business Research:

Qualitative and Quantitative Methods, John Wiley & Sons, Australia, Ltd, Sydney.

Ceceres, R.C., & Paparoidamis, N.G., (2007), Service quality, relationship satisfaction,

trust commitment and business-to-business loyalty. European Journal of Marketing, vol

41, no 7/8, pp 836-867.

Centre for International Economics (CIE). 2004, A National Wine-grape Growers

Association: a discussion paper, Department of Agriculture, Canberra

Charters, S., & Menival., D., (2010), The impact of the geographical reputation on the

value created by small producers in Champagne, Proceedings of The 5th International

Conference of the Academy of Wine Business Research, Auckland, New Zealand,

February 2010.

Chelwos, P., Benbasat, I & Dexter, A.S., (2001), Research report: empirical test of an

EDI adoption model, Information Systems Research, vol 12, pp. 304- 321.

181

Chin, W.W., (1998). The partial least squares approach for structural equation

modelling. modern methods for business research. Lawrence Erbaum Associates.

Mahwah, NJ.

Chong, S., (2007), Business process management for SME‟s: An exploratory study of

implementation factors for the Australian wine industry, Journal of Information

Systems and Small Business, Vol 1, no 1-2, pp 41-58.

Clancy, P. (2005), 'Time to bridge the great divide', The Australian and New Zealand

Wine Industry Journal, vol. 20, no. 2, p. 4.

Claycomb, C., & Frankwick, G. L., (2004), A Contingency Perspective of

Communication, Conflict Resolution and Buyer Search Effort in Buyer-Supplier

Relationships. Journal of Supply Chain Management, vol 40, pp 18-34.

Cohen, L., & Maldonado, A., (2007). Research Methods In Education. British Journal

of Educational Studies, vol 55, no 4, pp. 469 – 470.

Corey , E.R. (1991), Industrial Marketing, Cases and Concepts, Prentice Hall,

Englewood Cliffs.

Cox, A., Sanderson, J. & Watson, G., (2001), Supply chains and power regimes:

towards an analytic framework for managing extended networks of buyer and supplier

relationships”, Journal of Supply Chain Management, Spring pp. 28-35.

Cronbach, I.J., (1970). Essentials of Psychological testing, 3rd Edition, Harper and Row,

New York,

182

Crosby, L.A., Evans, K.R. & Cowles, D., (1990). Relationship quality in services

selling. An interpersonal influence perspective. Journal of Marketing, vol 54, no 3, pp.

68-81.

Cultural Portal. (2008), Australia‟s wine industry, culture and recreation, online

accessed 30/8/09: URL: http://www.recreation.gov.au/articles/wine

Daft, R.L., & Lengel, R.H., (1984), Information richness: a new approach to managerial

behavior and organizational design. In: Cummings, L.L. & Staw, B.M. (Eds.), Research

in organizational behavior 6, (191-233). Homewood, IL: JAI Press.

Davidson, D., (2010), Grape prices in Murray valley wine zone, Murray Valley

Winegrowers Association, Mildura, Australia.

Deloitte (2009), Annual financial benchmarking survey for Australian wine industry-

Vintage 2008, Deloitte Touche Tohmatsu, Sydney.

Deloitte & WFA ,2006, Annual financial benchmarking survey for Australian wine

industry- Vintage 2005, Deloitte Touche Tohmatsu, Sydney.

Department of foreign affairs and trade (DFAT). (2009), The Australian wine industry.

Online accessed 3/2/10, URL: http://www.dfat.gov.au/facts/wine.html

Dibben, J., & Chin, W.W., (2005). Multi- group comparison: Testing a PLS model on

the sourcing of application software services across Germany and the USA using

permutation based algorithm. In: Bliemel and Eggert and Fassott and Henseler

(eds):Handbuch PLS-Pfadmodelierung, pp 135-160, Schäffer-Poeschel Verlag,

Stuttgart

http://www.recreation.gov.au/articles/wine

http://www.dfat.gov.au/facts/wine.html

183

Domine, A., (2000), Wine, Konemann VerlagsgesellschaftmbH, Cologne.

Donaldson, B., & O‟ Toole, T., (2000) Classifying relationship structures: relationship

strength in industrial markets, Journal of Business & Industrial Marketing, vol. 15 no 7,

pp 491 – 506.

Doney, P.M., & Cannon, J.P., (1997), An examination of the nature of trust in buyer-

seller relationships, Journal of Marketing, vol. 61, no. 2, pp. 35-51.

Dorch, M.J., Swanson, S.R. & Kelly, S.W., (1998). The role of relationship quality in

the stratification of vendors as perceived by customers. Journal of the Academy of

Marketing. Science. vol 26, no 2, pp. 128-142.

Duffy, R.J., (1999), Trail Blazing, Purchasing Today, April, pp. 45-52.

Duncan, T., & Moriarty, S.E., (1998), A communication based marketing model for

managing relationships, Journal of Marketing, vol 62, pp 1-13.

Dwyer, R.F., Schurr, P.H. & Oh, S., (1987), Developing buyer-seller relationships,


Edmonds, M., (2000), 'Meeting productivity and price requirements', in Proceedings

ASVO Viticulture Seminar - Modern Viticulture Meeting Market Specifications, eds. C

Davies, C Dundon & R Hamilton, Australian Society of Viticulture and Oenology,

Adelaide, pp. 24-27.

Emerson, R.M., (1962), „Power-dependence relations‟, American Sociological Review,

vol. 27, pp. 31-41.

184

Everitt, B.S. (1996) Making Sense of Statistics in Psychology, Oxford University Press,

Oxford.

Farmers protests in Kosovo turn violent (2010), Radio Free Europe: Radio Liberty,

Online Accessed, 5/2/11, URL:

http://www.rferl.org/content/Farmers_Protest_In_Kosovo_Turns_Violent/2162972.html

Ford, D., (1984), Buyer/seller relationships in international industrial markets,

Industrial Marketing Management, vol. 13, p. 101-112.

Ford, D., (1990), Understanding Business Markets: Interaction, Relationships and

Networks, Academic Press, London.

Ford, D., Berthon, P., Brown, S., Gadde, L.E., Hakansson, H., Naude, P., Ritter, T &

Snehota, I. (2002), The Business Marketing Course: Managing in Complex Networks,

John Wiley & Sons Ltd, New York.

Fornell, C., (1992), A national customer service barometer: the Swedish experience,

Journal of Marketing, vol 55, no 1, pp 1-21.

Fornell, C., & Larcker, D.F., (1981). Evaluating structural equations with unobservable

variables and measurement error. Journal of marketing Research, vol 18, no 1, pp 39-

50.

Frazier, G.L., (1983), Inter-organisational exchange behaviour in marketing channels: a

broadened perspective, Journal of Marketing, vol. 47, no. 2, pp. 68-78.

Gall, M. D., Gall, J. P., & Borg, W. R. (2003). Educational research: An introduction

(7th ed.), Allyn & Bacon. Boston.

http://www.rferl.org/content/Farmers_Protest_In_Kosovo_Turns_Violent/2162972.html

185

Ganesan, S., (1994), Determinants of long-term orientation in buyer-seller relationships,


Gaski, J., (1984), The theory of power and conflict in channels of distribution”, Journal

of Marketing, vol. 48, no. 2, pp. 9-29.

Geyskens, I, Steenkamp, J.E.M. & Kumar, N., (1999), A meta-analysis of satisfaction

in marketing channel relationships, Journal of Marketing Research, vol. 36, no. 2, pp.

223-238.

Geersbro, J. & Ritter, T. (2010), External performance barriers in business networks:

uncertainty, ambiguity, and conflict, Journal of Business & Industrial Marketing, vol

25, no.3, pp. 196–201.

Guba, E.G & Lincoln, Y.S. (2005), Paradigmatic controversies, contradictions and

emerging confluences, in Denzin, N.K & Lincoln, Y.S (eds). The Sage Handbook of

Qualitative Research, 3rd Edition, pp. 191- 216, Sage Publications, London.

Gummeson, E, (1987), "The New Marketing: Developing Long-Term Interactive

Relationships", Long Range Planning, vol. 20, no. 4, pp. 10-20.

Gundlach, G.T., Achrol, R.S. & Mentzer, J.T., (1995), The structure of commitment in

exchange, Journal of Marketing, vol. 59, no. 1. pp. 78-92.

Gyau, A., & Spiller, A., (2007), Determinants of trust in the international fresh produce

business between Ghana and Europe. International Business Management, vol 1, no 4,

pp 104-111.

186

Hair, J.F., Black, W.C., Babin, B.J., Anderson, R. & Tatham, R.L., (2006), Multivariate

Data Analysis, 6th Edition, Pearson Prentice Hall, Upper Saddle River, New Jersey.

Håkansson, H., (1982), International Marketing and Purchasing of Industrial Goods. An

Interaction Approach., John Wiley & Sons Ltd., Chichester, England.

Håkansson, H. & Ford, D, (2002). How should firms interact in business networks?,

Journal of Business Research, vol 55, no.2, pp. 133–139.

Håkansson, H., Johanson, J. & Wootz, B., (1977), Influence tactics in buyer-seller

processes, Industrial Marketing Management, vol. 5, pp. 319-332.

Hall, C.M., (2004), Small firms and wine and food tourism in New Zealand: issues of

collaboration, clusters and lifestyles, in Small Firms in Tourism: International

Perspectives, ed. R. Thomas, pp.167-181, Elsevier, Oxford.

Hayman, P.T., Soar, C. J., Sadras V.O. & McCarthy, M.G., (2007). Can we identify

dangerous climate change for Australian viticulture? Australian Wine Industry

Technical Conference, Adelaide.

Heide, J.B., (1994). Interorganizational governance in Marketing channels Journal of

Marketing, vol 58, no 2, pp 71-85.

Heide, J.B., & John, G., (1992), Do norms matter in marketing relationships? Journal of

Marketing, vol. 56, no. 2, pp. 32-44.

Henry, P (2009), Australian wine industry: State of Play, Presentation to Adelaide Hills

Grape and Wine Association.

187

Herath, T., & Rao, H., (2009), Encouraging information security behaviours in

organisations: role of penalty pressures and peculated effectiveness, Decision Support

Systems, vol 47, pp 154-165.

Hesse-Biber, S., Dupuis, P. & Kinder, T., (1991), HYPEResearch: A computer program

analysis of qualitative data with an emphasis on hypothesis testing and multimedia

analysis. Qualitative Sociology, vol 14, no 4, pp. 289- 306.

Hines P., (1993), Integrated materials management: the value chain redefined,

International Journal of Logistics Management, vol. 4, no. 1, pp. 13-22.

Hines, P., Lamming, R., Jones, D., Cousins, P. & Rich, N., (2000), Value Stream

Management: Strategy and Excellence in the Supply Chain, Prentice Hall, England.

Hobley, L., (2007), The value of trading relationships between buyers and sellers of

wine grapes in Australia, PhD dissertation, Curtin University, Western Australia.

Hobley, L.E. & Batt, P.J., (2005), Value creation in relationships between Australian

wineries and their wine-grape suppliers, Curtin University of Technology, Perth.

Holden, M.T. & O'Toole, T., (2004), A quantitative exploration of communication's

role in determining the governance of manufacturer-retailer relationships, Industrial

Marketing Management, vol. 33, no 6, pp. 539-548.

Hutt, M.D. & Speh, T.W. (2010). Business Marketing Management: A Strategic View

of Industrial and Organisational Markets, 10th Edition, The Dryden Press, Orlando.

188

IAC (1995), Wine grape and wine industry in Australia, Industry Assistance

Commission, Canberra.

Ivens, B. S., (2004). How relevant are different forms of relational behavior? An

empirical test based on Macneil's exchange framework". The Journal of Business &

Industrial Marketing, vol 19, no 5, pp. 300-309.

Janssens, M. A., (2008). Evolution of cooperation in a one-shot Prisoner's Dilemma

based on recognition of trustworthy and untrustworthy agents. Journal of Economic

Behavior & Organization, vol 65, pp 458-471.

Janssens, W., Wijnen, K., De Pelsmacker, P. & Van Kenhove, P, (2008), Marketing

Research with SPSS, Pearson Education, Esssex.

Johnston, W.J. & Lewin, J.E. (1994), A review and integration of research on

organisational buying behaviour, Marketing Science Institute, pp 94-111.

Joreskog, K.G. & Wold, H., (1982). The ML and PLS techniques for modelling with

latent variables: Historical and competitive aspects. In K.G Joreskog and H. Wold

(eds),Systems under indirect observation, part 1, North –Holland, Amsterdam, pp 263-

270.

Kaufmann, P. J., (1987). Commercial exchange relationships and the “negotiator's

dilemma?. Negotiation Journal, vol 3, no 1, pp. 73-80.

Kaufmann, P.J. & Stern L.W., (1988), Relational exchange norms, perceptions of

unfairness, and retained hostility in commercial litigation". The Journal of Conflict

Resolution, vol 32, no 3, pp. 534.

189

Kelly, P., (1974), Functions performed in industrial purchases decisions with

implications for marketing strategy, Journal of Business Research, vol. 2, no. 3, pp.

421-33.

Kim, W.G., & Cha, Y., (2002). Antecedents and consequences of relationship quality in

hotel industry, International Journal of Hospitality Management. vol 21, no 4. pp

321-338.

Kim, W.G., Lee, Y. K. & Yoo, Y.J., (2006). Predictors of relationship quality and

relationship outcome in luxury restaurants. Journal of Hospitality and Tourism

Research, vol 30, no 2. pp. 143-169.

Kinnear, T.C, Taylor, J.R, Johnson & Armstrong, S., (1993), Australian marketing

research, McGraw-Hill, Sydney.

Kingshott, R. P. J., & Pecotich, A., (2007), The impact of psychological contracts on

trust and commitment in supplier-distributor relationships. European Journal of

Marketing, vol 41, 1053-1072.

Kotler, P., Brown, L., Burton, S., Deans, K. & Armstrong, G. (2010). Marketing 8th

Edition, Frenchs Forest: Pearson Australia.

Kumar, N., Scheer, L.K., & Steenkamp, J.E.M., (1995), The effects of perceived

interdependence on dealer attributes, Journal of Marketing Research, vol. 32, no 3, pp.

348-356.

Kurokawa, S., Manabe, S. & Rassameethes, B., (2008), Determinants of EDI

(electronic data interchange) adoption and integration in the US and Japanese

190

automobile suppliers, Journal of Organizational Computing and Electronic Commerce,

vol 18, no 1, pp 1-33.

Kwon, I.W.G., & Suh, T., (2005), Trust, commitment and relationships in supply chain

management: A path analysis. Supply Chain Management, vol 10, pp 26-33.

Lagace, R., Dahlstrom, R. & Gassenheimer, J.B. (1991). The relevance of ethical

salesperson behaviour on relationship quality: the pharmaceutical industry. Journal of

Personal Selling and Sales Management, vol 4, pp. 39-47.

Lambe, C.J., Wittman, M.C., & Spekman, R.E., (2001). Social Exchange Theory and

Research on Business-to-Business Relational Exchange. Journal of Business-to-

Business Marketing, vol 3, no 1, pp. 1-36.

Leedy, P.D. & Ormrod, J.E., (2010), Practical Research: Planning and Design, 9th

Edition, Merrill Prentice Hall, Ohio.

Leenders, M.R. & Fearne, H.E., (1997), Purchasing and Supply Chain Management,

Irwin, Chicago.

Leonidou, L., (2004), Industrial manufacturer-customer relationships: the

discriminating role of the buying situation, Industrial Marketing Management, vol. 33,

no 8, pp. 731-742.

Leonidou, L.C., Barnes, B.R. & Talias, M.A. (2006), Exporter-importer relationship

quality: the inhibiting role of uncertainty, distance and conflict”. Industrial Marketing

Management, vol 35, no 5, pp. 576-588.

191

Leonidou, L.C., Katsikeas, C.S. & Hadjimarcou, J., (2002), Building successful export

business relationships: a behavioural perspective. Journal of International Marketing,

vol 3, pp. 96-115.

Liljander, V., Polson, P. & van Riel, A., (2009), Modelling consumer responses to an

apparel store brand. Store image as a risk reducer, Journal of Retailing and Consumer

Services, vol 16, pp 281-290

Leuthesser, L., (1997). Supplier relational behavior: An empirical assessment.

Industrial. Marketing Management. vol 26, no 3, pp 245-524.

Lusch, R., & Brown, J., (1996), Interdependency, Contracting, and Relational Behavior

in Marketing Channels. Journal of Marketing, vol 60, no 4. pp 19-38.

Lysons, K. & Gillingham, M., (2003), Purchasing and Supply Chain Management, 6th

Edition, Pearson Education Ltd, England.

Mackenzie, H., & Hardy, K., (1996), Manage your offering or manage your

relationships, Journal of Business and Industrial Marketing, vol 11, no 6, pp. 20-37.

Malhotra, N., Hall, J., Shaw, M. & Oppenheim, P., (2006), Marketing Research: An

Applied Orientation, 2nd Edition, Prentice Hall, Frenchs Forest.

Mandjak, T. & Simon, J., (2004), An integrated concept on the value of business

relationships: how could it be useful? in 20th Annual IMP Conference, IMP,

Copenhagen, Denmark.

192

Mariampolski, H., (2001), Qualitative Market Research: A Comprehensive Guide, Sage

Publications Inc., London.

Medlin, C. J. (2001), Relational norms and relationship classes: from independent

actors to dyadic interdependence, PhD dissertation, University of Adelaide.

Mohr, J., Fisher, R. and Nevin, J., (1996), “Collaborative communication in interfirm

relationships: moderating effects of integration and control.” Journal of Marketing, vol

60 no 3. Pp 103-115.

Mohr, J., & Nevin, J., (1990), Communication strategies in marketing channels: A

theoretical perspective. Journal of Marketing, vol 54, no 4. pp 36-51.

Mohr, J.J. & Spekman, R., (1994), Characteristics of partnership success: partnership

attributes, communication behaviour and conflict resolution techniques, Strategic

Management Journal, vol. 15, no 2, pp. 135-152.

Morgan, R.M. & Hunt, S.D., (1994), The commitment-trust theory of relationship

marketing, Journal of Marketing, vol. 58, no. 3, pp. 20-38.

Moriarty, R.T. (1983), Industrial Buying Behaviour: Concepts, Issues and Applications,

Lexington Books, Toronto.

Morris, M., (2005), The influence of national culture on buyer- seller trust and

commitment, PhD dissertation, University of Maryland, Baltimore.

Naude, P. & Buttle, F., (2000). Assessing relationship quality. Industrial Marketing

Management. vol 29, no 4. pp ,351-361.

193

Noordewier, T.G., John, G. & Nevin, J.R. (1990). Performance outcomes of purchasing

arrangements in indutrial buyer-vendor relationships. Journal of Marketing, vol 54, no

4, pp. 80-93.

Nunnally, J.C., (1978), Psychometric Theory, 2nd Edition , McGraw-Hill Book

Company, New York.

O'Neal, C., (1993), Concurrent Engineering with Early Supplier Involvement: A Cross

Functional Challenge. Journal of Supply Chain Management, vol 29, pp 2-9.

Oliver, R., (1980), A cognitive model of the antecedents and consequences of

satisfaction decisions, Journal of Marketing Research, vol 17, no 4, pp. 460- 469.

Osmond, R. & Anderson., (1998), Trends & Cycles in the Australian wine industry,

1850- 2000, Centre for International Economic Studies, University of Adelaide.

Ozanne, U.B. & Churchill, G.A., (1971), Five dimensions of the industrial

adoption process, Journal of Marketing Research, vol. 8, no. 2, pp. 322-328.

Parsons, A., (2002), What determines buyer-seller relationship quality? An

investigation from the buyer‟s perspective. Journal of Supply Chain Management,

Spring, pp 4-12.

Petersen, K. J., Ragatz, G. L. & Monczka, R.M., (2005), An Examination of

Collaborative Planning Effectiveness and Supply Chain Performance. Journal of Supply

Chain Management, vol 41, 14-25.

Phillips, R., (2000), A short history of wine, Harper Collins, New York.

194

Phylloxera & Grape Industry Board, (2010), “2010 South Australian Winegrape

Utilisation and Pricing Survey, Phylloxera & Grape Industry Board, Adelaide.

Plewa, C., (2005), Key drivers of university-industry relationships and the impact of

organisational culture differences : a dyadic study, PhD dissertation, University of

Adelaide.

Porter, M., (1985), Competitive Advantage: Creating and Sustaining Superior

Performance, Free Press, New York.

Powell, W., (1996). Inter-organizational collaboration in the biotechnology industry.

Journal of Institutional and Theoretical Economics, vol 152. pp. 197-225.

Prahinski, C. & Fan, Y., (2007), Supplier evaluations: The role of communication

quality. Journal of Supply Chain Management, vol 43, pp 16-28.

Primary Industries and Resources South Australia (PIRSA) 2006, The Oversupply of

Cool Climate Wine Grapes, South Australian Wine Industry Council.

Quinn, R., (2008), French winemakers turn to terrior, Newser, Online Accessed 2/2/11,

URL: http://www.newser.com/story/33976/french-winemakers-turn-to-terror.html

Rampersad, G., (2008), Management of innovation networks in technology transfer,

PhD dissertation, University of Adelaide.

Ravald, A. & Gronroos, C., (1996), The value concept and relationship marketing,

European Journal of Marketing, vol. 30, no. 2, pp. 19-31.

http://www.newser.com/story/33976/french-winemakers-turn-to-terror.html

195

Redondo, Y.P. & Fierro, J.J.C., (2005), Moderating effect of type of product exchanged

in long-term orientation of firm-supplier relationships: an empirical study, Journal of

Product and Brand Management, vol 14, no 5, pp. 424-437.

Redondo, Y.P. & Fierro, J.J.C., (2007), Assessment and reassessment of supply chain

relationships: a case study in the Spanish wine industry, International Journal of

Entrepreneurial Behaviour & Research, vol. 12, no. 2, pp. 82- 106.

Reve, T., & Stern, L., (1979), Interorganizational relations in marketing channels,

Academy of Management Review, vol. 4, July, pp 405- 416.

Ringle, C. M., Wende, S. & Will, A., (2005). SmartPLS 2.0. Hamburg: Online

Accessed 5/6/10, URL: http://www.smartpls.de

Ritter, S., & Sue, V., (2007) Conducting Online Surveys, Sage publications, London.

Robinson, P.J., Faris, C.W. & Wind, Y., (1967), Industrial Buying and Creative

Marketing, Allyn and Bacon, Boston.

Rolland, C., & Prakash, N., (2005), Online questionnaire design, establishing guidelines

and evaluation existing support, in Managing Modern Organisations through

Information Technology, Proceedings of the 2005 Information Resources Management

Association Conference, pp 411-414.

Rosenbloom, B. (2004). Marketing Channels: A management perspective, 7th Edition,

Thomson, Ohio.

http://www.smartpls.de/

196

Sako, M., (1997). Does trust improve business performance? Christel Lane and

Reihnand Beckman edition. Trust between and within organizations. Oxford University

Press, Oxford, United Kingdom.

Scales, W., Croser, B & Freebairn, J (1995). Winegrape and wine industry in Australia:

A report by the committee of the inquiry in to the wine and grape industry, Australian

government, Canberra, Australia.

Scheer, L.K & Stern, L.W (1992), The effect of influence type and performance

outcomes on attitude toward the influencer, Journal of Marketing Research, vol 29, no

1, pp. 128-42.

Schiffman, L., Bednall, D., Cowley, E., O‟Cass, A., Watson, J. & Kanuk, L. (2001),

Consumer Behaviour, 2nd Edition, Prentice Hall, Frenchs Forest.

Seyed-Mohamed, N. & Wilson, D.T., (1990), Exploring the Adaptation Process”,

Institute of the Study of Business Markets, Report 16-1990, University Park,

Pennsylvania.

Simon, J., Mandjak, T. & Szalkai, Z. (2003), Analysis of Business Interactions and

Values from the Buyer‟s Side in the Hungarian Hospital Market, Proceedings or the

19th Annual IMP Conference, September 4-6, Lugano, Switzerland.

Smitka, M. J., (1991), Competitive ties: Subcontracting in the Japanese automotive

industry”. New York: Columbia University Press.

Spawton, T., & Walters, D., (2003), Supply chain management in the wine sector,

Bulletin O.I.V, Vol 76, no 867-868, pp 398- 424.

197

Speedy, B., 2006, The hangover: It‟s sour grapes for wineries, The Australian. Online

accessed, 24 July, 2007, URL:

http://www.theaustralian.news.com.au/story/0,20867,19423116-643,00.html.

Stanford. (2007). Wine industry overview, Australian Wine Industry Technical

Conference ,Adelaide, South Australia.

Steenkamp, J. E.M. & Baumgartner, H., (2000), On the Use of Structural Equation

Models in Marketing Modelling, International Journal of Research in Marketing, vol 17

no 3, pp 195-202.

Storer, C.S., (2005), Inter-organisational information management systems and

relationships in agribusiness food chains of organisations, PhD thesis, Curtin

University.

Thibaut, J.W. & Kelley, H.E., (1959). The Social Psychology of Groups.New

Brunswick: Transaction Books.

Ticehurst, G.W., & Veal, A.J., (2000), Business Research Methods: A Managerial

Approach, Pearson Education Pty Ltd, Frenchs Forest.

Trent, R.J. & Monczka, R.M., (1998), Purchasing and supply management: trends and

changes throughout the 1990s, International Journal of Purchasing and Materials

Management, vol. 34, no. 4, pp.2.

Turnbull, P.W. & Wilson, D.T., (1989), Developing and protecting profitable customer

relationships, Industrial Marketing Management, vol. 18, no.3, pp.233-238.

http://www.theaustralian.news.com.au/story/0,20867,19423116-643,00.html

198

Ulaga, W., (2001), Customer value in business markets: an agenda for inquiry,

Industrial Marketing Management, vol. 30, no. 4, pp. 1-7.

Vadarajan, P.R., & Cunningham, M.H., (1995), Strategic alliances: a synthesis of

conceptual foundations. Journal of the Academy of Marketing Sciences. vol 23, no 4,

pp 305-320.

van Weele, A., (2000), Purchasing and Supply Chain Management: Analysis, Planning

and Practice, 2nd Edition, Chapman and Hall, London.

van Weele, A.J., (1994) Purchasing Management: Analysis, Planning and Practice

Chapman & Hall, London.

Wacquant, L., (1992). Positivism. In Bottomore, Tom and William Outhwaite, ed., The

Blackwell Dictionary of Twentieth-Century Social Thought, Blackwell Synergy,

London.

Walsh. (1979), The Australian wine industry, 1789-1979, Australian National

University, Canberra.

Walter, A., Muller, T. A., Helfert, G. & Ritter, T., (2003), Functions of industrial

supplier relationships and their impact on relationship quality. Industrial Marketing

Management, vol 32, 159-169.

Webster, F.E. & Wind, Y., (1972), A general model for understanding organizational

buying behaviour, Journal of Marketing, vol. 36, no. 1, pp. 12-19.

199

Werts, C.E., Linn, R.L. & Jöreskog, K.G., (1974). Interclass reliability estimates:

testing structural assumptions. Educational and Psychological Measurements, vol 34 ,

25-33

Wilson, D.T. (1995), An integrated model of buyer-seller relationships, Journal of

Academy of Marketing Science, vol 23, no 4, pp 335-345.

Wilson, D.T. & Vlosky, R., (1998), Interorganizational information system technology

and buyer-seller relationships. Journal of Business & Industrial Marketing, vo1. 13, no

2/3, pp. 215- 232.

Winetitles, (2010), 2010 Australian and New Zealand Wine Industry Directory,

Winetitles, Adelaide.

Wine Australia. (2009), History of the Australian wine industry, online accessed

28/8/09: URL: http://www.wineaustralia.com.au/australia/default.aspx?tabid=5189

Wong, A. & Sohal, A., (2002). An examination of the relationship trust, commitment

and relationship quality. Journal of. Retail and Distribution Management, vol 30, no 1,

pp. 34-50.

Woo, K. & Ennew, C.T. (2004), Business to business relationship quality: An IMP

interaction-based conceptualization and measurement. European Journal of Marketing,

vol 38 no 9/10, pp. 1252-1271.

Wray B., Palmer A. A. & Bejou D., (1994). Using neural network analysis to evaluate

buyer-seller relationships. European. Journal of Marketing. vol 28, no 10, pp. 32-48.

http://www.wineaustralia.com.au/australia/default.aspx?tabid=5189

200

Wright, K.B., (2005), Researching internet based populations: Advantages and

disadvantages of online survey research, authoring software packages and web services,

Journal of Computer-Mediated Communication, vol 10, no 3, Article 11.

Young, L.C., & Wilkinson, I.F., (1997), “The space between: towards a typology of

inter-firm relations”, Journal of Business-to-Business Marketing, vol 4, no 2, pp. 53-

97.

Zaheer, A., McEvily B., & Vincenzo P., (1998). The strategic value of buyer-seller

relationships. International Journal of Purchasing and Materials Management,

(Summer): pp 20-26.

Zikmund, W., & Babin, B., (2007). Exploring Marketing Research, 9th Edition,

Thomson South- Western, Mason.

an integrated model of buyer-seller relationships in the

Documents