acs73_30

8/6/2019 acs73_30

1/7

189PRELIMINARY COMMUNICATION

Summary

Tis paper investigated the determinants o loan repayment perormance amongwomen sel-help groups in Bayelsa State, Nigeria. A structured questionnairewas used to gather inormation rom one hundred and twelve (112) womenarmers rom eight communities within the States, using the multistagesampling technique. Te study revealed that credit was available or agriculturalproduction, processing and petty trading among women armers. Te meanamount o loan disbursed to armers was N18,147.3 while the mean amount oN15,000 was repaid. Loan repayment percentage was determined to be 83.73%while percentage deault was 17.27%. Te estimated regression model indicatedthat women as household heads (X2), interest rate (X3) and household size (X4),negatively and signifcantly aected the loan repayment perormance o womenarmers while price stability o arm proceeds (X5) and commitment to sel help

groups (X6), positively and signifcantly aected the loan repayment o womenarmers in sel help groups in the area. It was thereore recommended thatmore sel help credit/saving groups be established in the state and that fnancialinstitutions within the state should orm a linkage with this groups or e cientloan scheme.

Key words

determinants; loan repayment; perormance; women sel-help

Agriculturae Conspectus Scientifcus | Vol. 73 (2008) No. 3 (189-195)

Determinants of Loan RepaymentPerformance Among Women Self Help

Groups in Bayelsa State, Nigeria

George M. M. UGBOMEH 1

Felix Odemero ACHOJA 2

Victor IDEH 2

Albert Ukaro OFUOKU 2( )

1 Agricultural Science Unit Department o Vocational Education, Delta State University,Abraka, Nigeria2 Department o Agricultural Economics & Extension, Delta State University, AsabaCampus. Nigeria

e-mail: [email protected]

Received: January 24, 2007 | Accepted: June 18, 2008

8/6/2019 acs73_30

2/7

Agric. conspec. sci. Vol. 73 (2008) No. 3

190 George M. M. UGBOMEH, Felix Odemero ACHOJA, Victor IDEH, Albert Ukaro OFUOKU

Introduction

Te active participation o women in agriculture hascalled or the current paradigm o rural economic de-velopment via women empowerment. Nweze (1995) ob-served that rural women: (i) are too poor to save, (ii) lackability to organize fnancial sel-help activities and (iii)

need cheap credit to expand production and income intheir arms and non-arm activities. Nwajiuba (1999) ac-centuates the centrality o credit, especially or womenarmers to increase their investment in the absence o ad-equate savings. Credit is a critical input because it can beused to overcome other obstacles, such as lack o labour.More than 80% o Nigerian rural women armers, whoproduce most o the ood and agricultural raw materials,make use o many kinds o capital, such as seeds, erti-lizers, animal eeds, agro-chemicals, fshing equipmentor which they need fnancial support in the orm o mi-cro-credit. However, the women armers are perpetually

marginalized in the institutionalized credit programme.Hence, the micro-credits o the rural women armers mustbe satisfed largely outside the organized fnancial marketssuch as indigenous sel help group (SHG) or the purposeo pooling savings and credit mobilization. Micro-creditis the ultimate economic vehicle, through which resource-poor-women armers can be empowered economically toovercome poverty.

According to Oluwalana et al. (2004), micro-credit inthe rural area ranges between N20,000-N100,000. Arisingrom the above, women armers as the socially and econom-ically weak and disadvantaged, oen resort to the option o

sel-help saving\credit associations as alternative sources oinormal credit (Upton, 1996). Te relevant social aspectso people rom Arica, according to Ijere (1992) are thoseaspects that deal with their attitudes to lie and their modeo behaviour and relationship with one another as well astheir customs. He hypothesized that these issues shouldbe typifed by such norms as honesty, airness, equity, de-mocracy and mutual ellow eelings.

Kazeem and Nwizu (1998) stated that mobilization in-volves the pooling o all available resources or an eec-tive operation. It is a process o ensuring collective actionbased on social solidarity, commitment and conviction.

Group credit programme has its origin in social cohesionbetween individuals that are held together by a commontie or the purpose o assisting each other with loans tostart o or expand their businesses.

Group credit mobilization has the advantage o in-creased coverage o arming community, it allows orparticipation o more people o similar status and sex,time economy with respect to group credit extension pro-grammes.

Upton (1996) postulated that fnancial mobilizationunder group cohesion have some advantages. For instance,since the group takes collective responsibility or repay-ment, there is less risk o deault. Akubuilo et al. (1998)emphasized that credit obtained must be repaid accord-ing to agreed terms.

Credit repayment perormance could be inuenced bya myriad o actors such as interest rate, unstable priceso agricultural commodities, and the social relations andresponsibilities o the borrower. Te main programme oSHG includes saving mobilization, credit delivery and re-covery. Te ormation o sel-help group originated romthe ailure o many fnancial institutions to deliver sus-tainable fnancial services to the poor due to actors suchas high transaction costs, low repayment rates and declin-ing unding, rom government and international donors(Mohammed, 2005).

A sel help group (SHG) is a voluntary association

o people at the grass roots level to meet the challenges oeconomic and business activities in the rural cash econ-omy (Mohammed, 2005). Te members are held togetherby common objective usually ocused on addressing eco-nomic challenges aecting their well being. Group co-hesiveness is one o the actors that aect group structure.Group cohesiveness reers to the extent to which memberso a group want to continue as members o the group. Tiscan be expressed by the level o commitment o the mem-bers to the success o the group. Social cohesion has theadvantages o promoting savings and loan repayment be-haviour among members.

According to Otu (2003), the commonest and the mostwidespread orm o SHGs are the Rotating Savings andCredit Association known in the fnancial parlance asROSCA. It is sub-set o the inormal fnancial institutionsthat cover a wide spectrum o economic activities includingarming, fshing, petty trading and local processing andrelated micro enterprises that serve as sources o livelihoodo millions o Nigeria in both rural and urban areas.

While there is the general support or credit mobili-zation by social groups and fnancial institutions, Hulmeand Mosley (1996) maintained that the fnancial sel-sustainability o all micro-credit groups need to be thor-

oughly examined.

Te broad objective o the study is to investigate theDeterminants o Loan Repayment Perormance amongWomen Sel Help groups in Bayelsa State o Nigeria. Tespecifc objectives o the study are:

i. to describe the operations and characteristics o theservices rendered by women sel-help fnancial groupsin the study area;

ii. to determine the volume o loan procured and repaidby women armers in the study area;

8/6/2019 acs73_30

3/7


191Determinants o Loan Repayment Perormance Among Women Sel Help Groups in Bayelsa State, Nigeria

iii. to evaluate loan repayment perormance o respond-ents;

iv. to identiy and analyze the socio-economic determi-nants o loan repayment perormance o women arm-ers in the study area.

Research methodology

Te Study Area, Sampling echniques and DataCollection echniques

Te study was carried out in Bayelsa State o Nigeria.It is one o the states in the Niger-Delta region o Nigeria.Tis area was selected or the study due to the act thatthe women armers are at the oreront o agriculturalproduction in the area but the socio-cultural setting thatenvelops them tends to inhibit their accessibility to agri-cultural resources, including fnance.

A total o 112 (one hundred and twelve) respondent

women armers who are members o women only saving/credit association in the study area were randomlyselected using the multi-state techniques, rom our localgovernment areas o Bayelsa State, Nigeria, and studied.Te distribution is shown in the able 1.

niques o Banqual (1993), adopted students t-test to test thesignifcant dierence between the mean volume o micro-credit borrowed and the mean amount repaid. Tis wasused as an index o credit repayment perormance o therural women armers in the study ollowing Achoja andIdeh (2005). Te ormula or calculating the students t-test

according to Koutsoyiannis (2001) is given as ollows:

t = 1 2

1 2

1 2

X X

SD SD

n n

At n1 + n2 2 degree o reedom

where:

1X = Mean Volume o loan borrowed by women arm-

ers rom credit association

2X = Mean Volume o loan repaid by women armers

to credit association

SD1 = Standard deviation o loan borrowed by womenarmers

SD2 = Standard deviation o loan repaid by women

armers

n1

= Number o women armers that borrowed loan

n2 = Number o women armers that repaid loan

Model specifcation

Multiple regression analysis was conducted to deter-mine the eect o the predictor variables on the loan re-payment perormance o the women armers. Te modelis specifed explicitly as ollows:

Yr = a + b0 + b1X1 + b2X2 + b3X3 + b4X4 ++b5X5 +b6X6, +

where :

YR = Volume o loan repaid (N)

X1 = Cost o loan recover (y)

X2 = Women as household heads (dummy)

X3 = Interest rate (%)

X4 = Household size (Number)

X5 = Price stability o arm proceeds

X5 = Commitment

= Stochastic disturbance term

Results and discussion

Socio Economic Characteristics o Women

Farmers in the Study Area

Te results showed that majority (51.3%) o the re-spondents were married. his explains the degree osocial responsibilities o the women armers. Te distri-bution o educational attainment showed that 52.5% hadno ormal education, 30.0% had primary education, 7.0%had secondary education and 0.5% had tertiary education.

Table 1. Distribution o respondents

State L.G.A Communities No. ofRespondents

1. Alagbem 1. Aghoro2. Legugbene

1414

2. Ekeremor 1. Ekeremor

2. Ayamassa

14

14Bayelsa 3. Oporomor 1. Ndoro

2. Peretonigbere1414

4. L.G.A 1. Ofoni2. Sagbama

1414

4. L.G.A 8 Communities 112

(Source: 2003 Survey Data)

Structured questionarie was the instrument used ordata collection. Five enumerators were selected and trainedto assist in the feld survey. Tey were however supervisedby the research o cers. Out o the one hundred and twentycopies o questionnaire administered, one hundred and

twelve (112) copies were retrieved rom the represents orthe purpose o data analysis.

Data Analytical echniques

Tree methods o analysis were utilized viz: descrip-tive analysis, parametric correlation, t-test and ordi-nary least square (OLS) o multiple regression analysis(Koutsoyiannis, 2001).

Banqual (1993) used the percentage o total loan repaidas a measure o repayment perormance o the borrower.But in this study, the researchers in addition to the tech-

8/6/2019 acs73_30

4/7



Te ew that had tertiary education oen assume leader-ship position o the groups. Te results urther revealedthat 41.88% o the women armers were classifed as lowincome earners. Tis implies that their low income willlead to low savings, low investment and low arm output.Tis justifes their need or credit acilities. About 35.62%

o them are medium income earners while 22.50% othem were in the high income group. Further more, ma-

jority o the women armers (48.8%) had household size osix and above. About 40.6% had household size betweenour and fve, while 10.6% had household size o one tothree. Te act that majority o the women armers hadairly large household size indicated that they need to beeconomically empowered through eective savings andcredit schemes.

Te Operations and Characteristics o WomenSel-Help Savings/Credit Groups

Te results o the study revealed that the majority (80%)

o the women sel-help savings/credit groups in the studyarea convened meetings on monthly basis during whichapplications or loans are made orally by members. Teseare documented and orwarded to awarding committee orapproval. Te monthly subscriptions so collected ormedthe pool o fnancial resources o the group rom whichloans can be disbursed to qualifed members. 80% o therespondent women armers that participated in the survey,agreed that they enjoyed access to credit in their groupswith litt le or no di culty.

Stated below are the characteristics o services thatmeet women armers fnancial needs in women sel-help

savings/credit groups in the study area:

i. Credits were available or agricultural production,processing and petty trading.

ii. Collateral was not required because substitutes such assolidarity to the groups, character reerences and per-sonal eects such as jewellery, were acceptable .

iii. Credits were available in small amounts or short-termworking capital.

iv. Credit repayment schedules ft women armers busi-ness cycles.

v. Credit/loan sizes may be increased upon satisactory

repayment o the frst-time loans.vi. Micro-enterprises with ew employees owned and op-erated by women armers were eligible.

vii. Signature o male relative or spouses were not requiredbeore credits applications were approved.

viii. Literacy was not a requirement or credit acquisi-tion.

ix. Loan applications were easily and quickly processed.

x. Since the women sel-help savings/credit associationswere organized on intra-community basis, location was

convenient and sae or women armers in the studyarea.

xi. Te hours o operation o the groups were compatiblewith the women armers economic activities and do-mestic obligations.

About 80% o the respondents agreed that loans wereairly distributed to eligible members without bias. Alsothe study revealed that interest generated rom loanswere shared as dividends to members at end o the fnan-cial year.

Te Volume o Loan Borrowed and Repaid by

Women Farmers in the Study Area

Within the period o the survey (2005) a total amounto N2,032,500 was disbursed with a mean amount oN18,147.30. Te minimum amount o loan disbursed towomen armers was N9,500 while the maximum amounto loan disbursed was N45,000. Te very low standard

deviation o the amount o loan disbursed (1.4) indicatedthat in women sel-help savings/credit associations, loanswere airly and uniormly distributed to eligible members.Te mean value o loan (N18,147.30) disbursed to womenarmers in the study area, within the period under review,was relatively substantial to start and manage micro-enter-prises, whether on-arm or o-arm. I well managed, thiscould grow over time to expand the capital base o suchmicro-enterprises, thereby alleviating the poverty o suchwomen armers. Tis result is at variance with the asser-tion o Mahaja (1999) that micro credit is not an answer topoverty eradication. Micro credit would serve as catalystin the growth o micro enterprises in Nigeria.

Within the period o the survey study, a total N1,681,430was repaid by the women armers to sel-help savings/creditgroups in the study area. Te mean amount o loan repaidby women armers was N15,000. Te minimum amounto loan repaid was N2,375, while the maximum amounto credit repaid was N45,000. Tese fndings agreed withthe earlier report o Akubuilo et al. (1988) that credit ob-tained must be repaid according to agreed terms. Tis isto encourage credit recycling in sel help groups. Te rel-atively low standard deviation o amount o loan repaid(1.3), is indicative o the act that members o women sel-help credit associations were relatively homogeneous intheir ability to repay loans. Women armers o relativelythe same economic status would be more comortable topool their fnancial resources together under the umbrellao savings\credit associations.

Loan Repayment Perormance o Women

Farmers

H0; Tere is no signifcant dierence between the meanamount borrowed and the mean amount repaid.

8/6/2019 acs73_30

5/7



From the results o the study, loan repayment per-centage was determined to be 82.73% while percentagedeault was determined to be 17.27%. Te statistical di-erence between the mean value o loan obtained (X1) andthe mean amount o loan repaid (X2) by women armersis presented in able 2.

Further analysis using students t-test at 5% level osignifcance, revealed that there is no signifcant dier-ence between the mean amount borrowed and the meanamount repaid by women armers in sel-help savings/credit groups in the study area.

Te implication o these results is that women armersin sel-help savings/credit groups in Bayelsa State, Nigeria,exhibited high loan repayment perormance within theperiod o the survey.

Model o Loan Repayment Perormance o

Women Farmers

We assumed in the study that loan repayment perorm-ance o the women armers may be inuenced by cost oloan supervision, domestic responsibility o women arm-ers, interest rate charged, house hold size and the priceso arm proceeds.

Te results o the multiple regression o the actors thatinuence loan repayment perormance o women arm-ers in the study area are presented below. An evaluationo the model or loan repayment perormance or ruralwomen armers showed that R2 value was 0.854 (85%)while the R2 (adj) value was 0.8230 (82%). Tis indicatesthat approximately 85% o the variation in the dependentvariable (volume o loan repaid), was due to the joint e-ects o the selected predictor variables that were capturedin the model.

est o Signifcance o Predictor Variables in the

Loan Repayment ModelHypothesis

H02: Te individual predictor variables have no signif-cant eect on the loan repayment perormance o womenarmers in the study area.

i.e H0: bs = 0

H1: bs 0

In the estimated regression model, an eort was madeto identiy which o the coe cients o the selected predic-tor variables provides a statistically signifcant contribu-

tion to the model. Te signifcance o the variables in themodel was evaluated at 1% and 5% levels o signifcance.Te linear unction was chosen as the lead equation aertrying the semi- log and double- log unctions or the loanrepayment model. Te model is presented below:

Yr = 13820 + 0.3350 X1 - 0.226X2

(1.85)** (1.34)Ns (-1.14)**

-346.9X3 - 581.4X4 + 613.8X5 + 0.361X6(1.70)** (-0.63)** (0.80)* (0.23)**

R2 = 85%, R2 (adj) = 82%, F = 69.166, DW = 1.51

= 1.728

where:

YR = Volume o loan repaid [Naira]

X1 = Cost o loan recovers [Naira]

X2 = Women as household heads (dummy)

X3 = Interest rate (%)

X4 = Household size (Number)

X5 = Price o arm proceed [Naira]

X6 = Commitment [Dummy]

= Stochastic disturbance term

Note: the fgures in parenthesis just beneath the regres-sion coe cient in the lead equation are the correspond-ing t-ratios

* = Signifcant at 1%** = Signifcant at 5%

All the variables entered the model with the expect-ed signs thus conorming with apriori expectations. Alsofve explanatory variables were signifcant in loan repay-ment model. Tese parameters relate to: women house-hold headship (X2), interest rate charged on loan (X3),household size (X4), the price o arm proceeds (X5) andcommitment (X6)

Determinants o Loan Repayment Perormanceo Women Farmers in Sel Help Group

Cost o loan recovery (X1) was positive in the modelbut was no a signifcant determinant o loan repaymentperormance o women armers in the study area. Tis im-plies that with little or no recovery eort, women armersin sel-help groups have the tendency to repay borrowedunds without being coerced by the awarding association.Tey believed that the credit association is their own andthey are committed to it since they are stake holders in the

Table 2. Statistical Dierence Between Volume o Loan Borrowed and Amount o Loan Repaid

Variables Means X SD Df t. cal t. tab Remark

Amount borrowed (X1)Amount Repaid (X2)

N 18,147.30N 15,000.00

1.41.3

222 0.49 1.64 NotSignificant

(Source: 2005 Survey Data)

8/6/2019 acs73_30

6/7



association. Tey indeed demonstrate the spirit o socialcohesion. Tey tend to encourage one another to pay uptheir outstanding loans. Tis is the only way micro-creditmobilization can be sustained among women in sel-helpcredit groups in Bayelsa State o Nigeria.

Women as Household Heads (X2)

Tis variable was negative in the model. Tis was inline with apriori expectation that women who assumedheadship position in their amily compete with the armenterprises or the scarce fnancial resources o the armer.Hence a woman armer in traditional Arican setting whocaters or domestic chores such as house rent, eeding, chil-dren education and health care o amily members, mightdivert agricultural credits to any or all o the householdssocial responsibilities. Tese elements o household re-sponsibilities could act as sources leakages o arm credits.Te consequence is that such women armers may perormpoorly in loan repayment.

Interest Rate (X3)

rue to a priori expectation, the results o the studyrevealed that interest rate charged had a negative eecton loan repayment perormance o women armers in thestudy area. Te higher the interest rate charged by thris/credit associations, the more the di culties women arm-ers encountered in repaying borrowed und and vice versa.As the lending behaviour o fnancial institutions is inu-enced by interest rates so also credit repayment o the bor-rower could be inuenced by interest rates. Interest rate, itoo high, becomes a burden to the borrower, as it increases

the cost o production. Hence interest rate subsidies, insel-help groups would ordinarily stimulate borrowingand the repayment perormance o the women armersin Bayelsa, Nigeria.

Household size (X4)

Household size o the respondents entered the modelwith a negative sign which indicates that household sizeimpacts negatively on the loan repayment perormance owomen armers in the study area. A emale armer whohas a large amily size to cater or may divert some o theborrowed und to unintended purposes or the upkeepo her amily. Tese may be regarded as sources o credit

leakage and they tend to negatively aect her loan repay-ment perormance in the rural area.

Price o arm proceeds (X5)

Te price o arm products turned out to be positiveand signifcant in the loan repayment perormance modelo rural women armers in Bayelsa State. Tis result showsthat when the prices o arm produce go up, women armerswould be empowered economically. Tus they would havehigh tendency to repay borrowed und, both the principal

and the interest charged. Contrarily, when women armersare victims o the risk o crop ailure and price all, theirloan repayment perormance could be poor.

Commitment (X6)

Tis variable captures the eect o social cohesion in

the model. It is positive and signifcant in the model. Tisshows that the more the members are committed to thegoals o sel help groups, the higher is the tendency or loanrepayment. Tis result is in agreement that social cohe-sion can improve the attitude o members towards groupgoals. Upton (1996) has earlier postulated that since thegroup takes collective responsibility or repayment, thereis less risk o deault. Since members are committed tothe group, women armers tend to be more comortablein pooling their fnancial resources under the umbrella osavings and credit scheme. Tis can enhance sustainablecredit recycling system.

Conclusion and recommendation

Te determinants o loan repayment perormanceamong women sel help groups was investigated in BayelsaState o Nigeria. Te results o the study showed that ma-

jority o the women armers surveyed were married, un-educated, belonged to low income group and had largeamily size. Tat not withstanding, they recorded high loanrepayment perormance within the period o the survey.Te social cohesion o women armers in sel-help creditgroups must have positively inuenced their ability to repaytheir loans. Tese women armers repaid their loans withlittle or no supervision.

It was recommended that more sel-help credit\savingsgroups should be established in the study area. Furthermore,commercial banks and other ormal credit institutions inBayelsa State can eectively operate their credit policiesby orming linkages with these groups since they are e -cient in loan repayment. Te sole aim is to reduce povertyand promote a high standard o living. When the reversehappens, then theory o poverty as postulated by Lewis(1967) would be defnitely entrenched.

ReerencesAchoja F.O. and Ideh, V (2005) Perormance o Rural Women

Farmers in Micro Credit Repayment: Evidence romthe Coastal Area o Delta State, Nigeria in Ogisi D.O.Okuneye B.P. and Oyaide W.J. (eds) Economic Reormsand Management o Nigerian Agriculture. Proceding o theFarm Management Association FAMAN at the 19th AnnualConerence held at Delta State University, Asaba Campus. PP125 133.

Adamu, B. (2005) Small and Medium Industries EquityInvestment Schemes (SMIEIS): Pro or Anti Industrialization.Te Bullion, Vol. 28 (4) 41 50

8/6/2019 acs73_30

7/7



Akubuilo, C.J.C.E.E Umebali and M.O. Ude (1998) ComparativeAnalysis o Formal and inormal services in journal o creditor Small Scale Agricultural Ventures in Udi Local Areao Enugu State in Journal o Agriculture, echnology andEducation Vol. 3 (1): 64-70.

Banqual K.A.M and B.L. Meyer (1993) Te Political Economy oRural Loan Recovery: Evidence rom Bangladesh: Savings andDevelopment Vol. XVII, pp23-38.

Hulme D. and P. Mosley (1996). Finance Against Poverty: London:Roultedge.

Ijere, M. O. (1992) Leading Issues in Rural Development: Enugu-Auna Publications.

Kazeem K and Nwizu, D. C. (1998) Mobilizing the Education Eliteor the each one teach one (EOO) Delivery System in NigeriaCited in Oriao S.O. and Olubor R. O. (eds) Advances in feldo Education. Te Nigeria Experience: Benin UNIBEN. Pp.18-28.

Koutsyiannis. A.(2001) Correlation Teory in Teory oEconometrics. New York Replika Press pp31-43.

Lewis, o. (1967) La Vida (A Puerto Rican Family in the Culture oPoverty San juan and new york) London. Wartin Sector andWarberg Ltd.

Mahaja, V. (1999) Is Micro the answer Poverty Eradicationcited in Adult Education and Development. Vol. 53, GermanyInternational Cooperation o German Adult Education andAssociation (112/-DVV): 40-56.

Mohammed, S. F (2005) Te Sel Help Groups SHGs LinkageBanking Programme: concept and practice in Nigeria. TeBull ion, Vol29 (4): 15-18.

Nwajiuba, C.U. (1999) Adaptive Welarism: A Paradigm o PolicyDriven Development. Owerri Jockel Options Business Bureau.

Nweze N.I. (1995) Te Role o Inormal Finance in Rural Areas oNigeria Cited in Eboh E.C. et al (eds). Rural Development inNigeria: Concepts Process and Prospect. Enugu Auto CenturyPublication Co. pp. 10 17

Oluwalana E.O A Okuneye, P.A and Sokoyo G.O. (2004) MicroCredit or Agricultural Development. A Study o WomenGroups in Ogun State, Nigeria.

Okuneye P.B. and Evboumwa G.O. (eds) Agribusiness in AricanCentury. Proceeding o the Arican Farm ManagementAssociation (AFMA) at the 7th Biennial Congress Held atthe Central Bank o Nigeria (CBN) Corporate Headquarters,Abuja, Nigeria. Pp 25 35

Out M.F. (2003) Inormal Credit Market and MonetaryManagement in Nigeria Occasional Paper Central Bank oNigeria Research Department No. 29 Pp. 5 10

Upton M. (1996): Te Economics o ropical FarmingSystem; Cambridge University Press. Cambridge.

acs73_30

acs73_30

Documents