software exports development in costa rica: potential for policy reforms

Software Exports Development in Costa Rica:Potential for Policy Reforms

Brian NicholsonManchester Business School, Manchester M15 6PB, UK.E-mail: [email protected]

Sundeep SahayDepartment of Informatics, University of Oslo, Blindern N-0316 Oslo, Norway.E-mail: [email protected]

ABSTRACT

Software industry development is acknowledged as an important engine of economic growth formany developing countries. The role of national policy has been identified as a catalyst to softwareindustry and software exports development. Academic and practitioner frameworks have emerged toguide policymakers to the factors that are critical to success. This article adopts an institutional theoryperspective on the strategy-making and implementation process and examines how historically rooted,taken-for-granted assumptions limit the possible courses of action and may also simultaneously openup new opportunities. Certain complexities and contradictions are highlighted through a longitudinalcase study of Costa Rica where there is an ongoing strategic planning effort to increase softwareexports. Key findings from the research are related to improving our understanding of how softwareexports policy is influenced by local, emergent institutions that have been shaped by particularhistorical circumstances. C© 2009Wiley Periodicals, Inc.

Keywords: software exports policy; institutional theory; action research

1. INTRODUCTION

Software industry development is an important enabler of economic growth (Al-Jaghoub,2004; Heeks & Nicholson, 2004; Kambhampati, 2002), and thus policymakers and otherstakeholders in many emerging countries, such as Costa Rica, Jordan, and Vietnam, havebecome actively interested in developing their country’s software industry sector, particu-larly exports (Al-Jaghoub; Carmel, 2003; Duong, 2004; Paus, 2005). The success of Indiaand Ireland have contributed to this intense growth in interest in boosting exports, especiallyIndia, which has experienced a spectacular growth over the last two decades. According toIndia’s National Association of Software and Service Companies (NASSCOM), IT servicesexports were US$18 billion in 2007 and are predicted to increase 28% to US$23 billion in2008, continuing a pattern of yearly growth experienced since the mid nineties. The sectoris a major employer and is expected to occupy 865,000 people in 2008. The majority of

Leiser Silva was the accepting Guest Editor for this article.

Information Technology for Development, Vol. 15 (1) 4–16 (2009) C© 2009 Wiley Periodicals, Inc.Published online 9 January 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/itdj.20107

4

SOFTWARE EXPORTS DEVELOPMENT IN COSTA RICA 5

Indian IT services exports are comprised of custom application development. Indian soft-ware companies have tended to steer clear of developing software products due to the needfor greater investment, higher risk, and greater complexity and need for domain expertise(NASSCOM, 2008). Ireland’s software sector has experienced a similar pattern of growth,including a doubling of numbers of software firms (291 to 690) during the nineties, withsignificant implications for employment generation. The growth in the software sector inIndia has contributed to productivity spillovers in other service companies and beneficialdemonstration effects to other sectors (Arora, Arunachalam, Asundi, & Fernandes, 2001).

Although many new entrant countries are attempting to become the “next India” in soft-ware and services exports, this must be considered along with other development priorities(such as water, sanitation etc.) and the potential of alternative clusters as contributors toeconomic development. India’s software export story is not all good news, for instance,an export orientation in India has led to an underserved domestic market. Also, the globalsoftware and services market is highly dynamic and increasingly competitive with manynew and emerging players in cheap labour countries where vendors are hungry for workand quickly developing technical and language capabilities. Volatility in the global marketfor software and services was spectacularly demonstrated in 2001 when around 100,000Indian programmers were “benched” largely as a result of a USA economic downturn.

When examining those countries that already engage in software exports, the role of na-tional policy has been identified as an important driver of the software industry and exportsdevelopment both in developing, e.g., India (Kambhampati, 2002) and in developed, e.g.,Finland (Ein Dor, Myers, & Raman, 2001) country contexts. Duong (2004) describes howthe many favourable policies, such as tax exemptions adopted by the Vietnamese govern-ment, have had beneficial implications for the software industry in areas such as humanresource development, telecommunication infrastructure, IT investment, and intellectualproperty rights (IPR) protection.

Policy formulation is concerned with developing and strengthening an interconnectednetwork of organizations, people, infrastructure, legislation, markets, and finance. Policyframeworks have contributed to shaping industry trajectories, for example, the dominantfocus on products in Israel, on foreign direct investments (FDI) in Ireland, and on softwareservices in India (Heeks & Nicholson, 2004). While previous studies have focused primarilyon a post-hoc analysis of the impacts of policy measures (e.g., Al-Jaghoub, 2004), there hasbeen limited analysis of ongoing processes through which policy measures are constructedand implemented.

The aim of this article is thus to improve our understanding of the institutional processes:their formulation and influences on national software exports strategy formulation and im-plementation in the context of developing countries. The central research question guidingour study is, how do institutional influences create or impede the potential for policy reformin the industry? An empirical study of the Costa Rican national software exports strategyformulation and implementation forms the setting for this inquiry.

The article is organized as follows: We begin with a theoretical discussion drawn frominstitutional theory, followed by a description of the methodology, analysis, and conclusions.

2. THEORETICAL FRAMEWORK

The theoretical perspective drawn upon in our analysis of national software exports strategyformulation and implementation is derived from institutional theory. Institutional analysis

Information Technology for Development DOI: 10.1002/itdj

6 NICHOLSON AND SAHAY

has wide acceptance as a valuable theoretical lens in relation to information and communi-cation technology (ICT) and development (e.g., Avgerou, 2003; Silva & Figueroa, 2002). Agood example is shown in Silva and Figueroa’s paper that draws on an institutional lens inrelation to ICT policy in Chile to examine why some policies achieve their objectives whileothers may not. A key challenge confronting researchers considering using institutional the-ory in their empirical work is the apparent lack of a central frame of theoretical referenceand set of concepts. Scott (2004) divides institutional theorists into regulative, normative,and cultural-cognitive. Scott writes that most economists and rational choice theorists stressregulative elements; early sociologists favoured normative elements, and more recent orga-nizational sociologists and cultural anthropologists emphasize cultural-cognitive elements(p. 465). In this article, we have chosen the Nobel Prize winning economist DouglassNorth’s (1990) seminal work in institutional theory to illuminate our field data because itprovides clear concepts from which to leverage understanding of the institutional processeswithin the Costa Rican software exports policymaking process. In the sections to follow,we narrow down on some central tenets of North’s conception of institutions that form thebasis of the conceptual tools.

For North (1990), institutions represent the rules and norms that individuals follow intheir daily lives, “the rules of the game in a society. . . .the humanly derived constraintsthat shape interaction” (p. 3). Institutions can take the form of formal rules or informalnorms, for example, the formal constitution of a country or the socially accepted normsof behaviour towards the elderly in a particular society. Institutions only affect behaviorwhen they are enforced, formally when rules are infringed such as a fine for speeding,or informally, such as a sneer towards someone who is disrespectful to an elderly person.Where there is limited or no overlap between formal and informal rules, informal rules tendto take priority in shaping behavior.

There are costs of conducting formal enforcements that increase as the overlap betweenformal and informal rules shrinks. Thus, the smaller the overlap, the more difficult it willbe for policymakers to influence individual behaviour, with higher enforcement costs. Aschematic describing this logic is presented in Figure 1 below.

For North (1990), organizations are created to take advantage of the opportunities insociety, which, together with the standard constraints of economic theory, are determinedby institutions. As organizations evolve, they alter the institutions (p. 7). Sautet (2005) drawson North’s institutional economics to point out that formal and informal institutions changeand evolve over time when amended or evolve spontaneously based on their usefulness.Informal norms are more difficult to influence than formal rules and ignoring the informalnorms can make enforcement of formal rules problematic (p. 6).

Figure 1 Interaction between formal and informal rules (Sautet, 2005).



Thus, the institutional approach drawn upon in this article emphasizes that reform of theformal institutions in a national software exports strategy and implementation process willrequire knowledge of both the formal rules and the informal norms (institutions) that existin a given society. Whether individuals and organizations follow the reformed formal ruleslegislated by the policy makers will depend on the alignment between formal rules andinformal norms of conduct.

This conceptualization helps us to explore and analyse the challenges presented by insti-tutions during processes of development oriented planning efforts. Software exports policyformulation involves a multiplicity of actors, organizations, and their associated institu-tions (formal and informal). The contradictions between formal and informal institutionsencountered during complex interactions between individuals and organizations provideboth the potential for reform or the persistence of institutions. This guiding logic forms thebasis of our analysis of the software exports strategy formulation process in Costa Rica.

3. METHODOLOGY: AN ACTION RESEARCH APPROACH

The authors were invited in 2002 by representatives of Costa Rica’s Inter American Devel-opment Bank funded software industry development organization Prosoftware. We were toparticipate as consultants along with Cegesti, a San Jose based not-for-profit consulting or-ganization. Together, we were tasked with developing the Costa Rican strategy for enablingsoftware exports. Our participation was within an action research framework (Baskerville& Wood Harper, 1996; Checkland, 1991) as we were expected to help in the formulation ofthe national strategy and its subsequent evaluation. The cycle of intervention and reflectionis represented in the Figure 2 below.

The methodology (M) was divided into phases of problem definition, analysis, andaction. The problem definition and analysis phase involved the development of a plan forthe empirical work after conducting a literature search of the Costa Rican software industryto identify the current state, the key stakeholders, and the particular technological andgeographical focus of the industry. The authors visited San Jose for 3 weeks in August

Figure 2 Cycle of action research (Checkland, 1991).



TABLE 1. Summary of Data Collection Sources

Data collection mechanism 2003 2005 2006 Total

Interviews (co present) 18 628Interviews (telephonic) 4

Focus groups 2 2Workshops 4 4

2003 and returned for a week in September 2003. Interviews with stakeholders rangingfrom the Minister of Science and Technology to university researchers enabled us todiagrammatically “map” the stakeholders and their formal and informal influences on theformulation and implementation of a national software exports strategy using a rich picturetechnique (Checkland, 1991).

Interviews were semi-structured and focused on the respondent’s background, interests,and influences with respect to software exports, the issues they considered relevant indeveloping a strategy, and to gauge their level of commitment to the development andimplementation of a national strategy. Two focus group sessions were conducted: one withrepresentatives from the private sector and the other with university staff. In these focusgroups, discussions were held on issues involving multi-stakeholder relationships such asuniversity-private sector linkages, and problematic relations between small- and medium-sized firms with financial institutions. A large national-level workshop was conducted onthe topic of global trends in the software sector, bringing together various stakeholders inorder to provide a broader awareness of the multiple inter-organizational issues, gain their“buy in” to the strategy formulation process, and elicit suggestions on specific initiativesto be taken. Three smaller workshops were held on themes identified through our problemidentification process, focusing particularly on policy formulation and implementation.The meetings, workshops, and interviews were mostly conducted in English, and when therespondents preferred to discuss the issues in Spanish, the services of a local interpreterwas drawn upon. During the national-level workshop and one of the smaller workshops,professional simultaneous translation was provided. In Table 1, a summary of the varioussources of data collection over time is presented.

In addition to these primary sources of data collection, various secondary data sources,such as industry reports, organization brochures, Web sites, industry statistics, etc., wereanalyzed to gain an understanding of other related institutional aspects.

The action research orientation focused on the need to create formal and informal struc-tures to enable the analysis, design, and implementation of the software exports strategy.The first step of the analysis process was to conduct a stakeholder analysis during and fol-lowing the various methods of data collection described above. This process of analysis wasinspired by Peter Checkland’s (1990) Soft Systems Methodology, specifically the creationof a “rich picture” that helped to trace the interlinkages between the various stakeholders.Various problematic themes were identified through this process, and we produced our ini-tial analysis of the strengths, weaknesses, opportunities, and threats facing the ICT exportsector.

In the action phase, four task groups,1 jointly composed of members from differentstakeholder organizations, were brought together and given the responsibility for designing

1The four task groups university-private sector linkages; marketing, finance; and the role of CAPROSOFT(The national software association).



solutions and shaping their implementation. For example, a task force group was establishedconsisting of members from the university and private sector and were charged with aspecific responsibility of how best to strengthen university and private sector linkages insupport of software export processes. These task force groups were provided with supportfrom the consultants, objectives, and resources such as best practices from other countries.They were asked to develop recommendations on actions (short and medium term) requiredto meet the specified objectives outlined in the strategy plan and also suggest indicatorsto measure progress. For example, the task group on university-private sector linkageswas comprised of members from both the universities and the private sector firms, whowere expected to jointly discuss the challenges to building these linkages and developjoint recommendations on concrete practical actions to address the problems. Identificationof the need for reform of university formal mandates to allow industry participation inresearch led to lobbying activity of senior university management. A workshop involvedprivate sector software companies and university researchers where private sector firmsdescribed possible projects for joint participation with researchers. In this way, the taskgroup contributed to the identification of the problems, assessment, and design of thesolutions, and they took responsibility and ownership of the action plan. The working ofthe task group was facilitated by the Consultants Team (comprised of Cegesti and ourselves),providing resources in the form of the experiences and best practices from other countries,supervising joint meetings, and monitoring progress.

The above approach built upon and also emphasized three issues. First, strategy formu-lation is seen as an ongoing, continuous, and iterative process. Second, the process is notlinear with a clear start and end, but is cyclical in nature where the inputs provided by thetask groups feed into making refinements to the strategy and actions emanating from it.Third, the strategy-making process has to be primarily owned by the people who have adefinite stake in the process.

The initial framework of ideas (F) comprised of the Heeks and Nicholson (2004) SoftwareExport Success Model, which guided the early phases of data collection and analysis andwas helpful in mapping out the relevant stakeholders. This model emphasizes the importanceof creating synergies between a complex range of interconnected actors and issues such asnational policy, education, infrastructure, and international demand.

The data analysis was a process involving extensive discussions between the variousstakeholders and us, whereby we shared our opinions and findings and gained their feed-back and further comments. A report was presented in 2003 to Prosoftware, the organizationthat commissioned the study, which provided the bedrock for the articulation of the na-tional strategy. The report was widely distributed, and we received feedback from variousstakeholders, which provided further useful input to the analysis and framing of this article.Reflection on “F, M,” as indicated in Figure 2, has taken place over time and has been mademore concrete in academic papers based on our findings (Nicholson & Sahay, 2005). Thefeedback we received from reviewers and conference attendees contributed to our contin-uing academic analysis of the themes of the case. Our continued reading of the academicliterature (especially related to Institutional Theory) and discussions we’ve had with col-leagues have led to the current framing of the article. Thus, our approach is comprised ofboth action and reflection, leading to research themes analysed using appropriate analyticalframeworks.

These cycles of reflection on the F, M, and A have continued over time. In mid 2005, nearly2 years after our initial actions, four audio conferencing interviews were conducted. Theseinterviews were facilitated by Cegesti and involved some key members of the software



industry organization. The discussion was focused on the challenges of implementingthe specific policy directions that we had recommended in 2003. Furthermore, one ofthe authors returned to Costa Rica in March 2006 to make a presentation at a softwareindustry conference; part of the presentation reflected on the progress of the softwareexport promotion strategy. During this time, six further interviews were undertaken withkey stakeholders, which brings us up-to-date with the events and the effectiveness of thestrategy, and to identify future measures. The findings described below are a result of thiscontinuing reflection and analysis of the themes from the study.

4. FINDINGS

We first describe some background on the Costa Rican software industry. Subsequently,we describe the relevant stakeholders, analyse the institutions (formal rules and informalnorms) and contradictions.

4.1 The Costa Rican Software Industry

The Costa Rican software sector gained significant global publicity when the IT giant Intelestablished a development unit in San Jose, enabled through the direct involvement of theCosta Rican President (Ketelhoehn & Porter, 2002). According to Camara Costarricensede Tecnologıa de Informacion y Comunicacion (CAMTIC; 2006) market research, in2005 there were 150 software development companies, 91% of these are micro, smalland medium sized. Applications “for and about the Internet” and “administration andengineering” prevail in software development. The revenues for 2005 for the softwaresector were US$173 million and the software development sector employed 4,800 staff.in 2004–2005, there was a reduction in production and banking software and an increasein customized application development. It is claimed that 30% of production is exported:software exports products accounted for $71 million and exports to Central America havedwindled (from 60% in 2004 to 40% in 2005), whereas exports to the U.S. have grown,rising from 13.3% in 2004 to 22% in 2005.

4.2 Analysis

The institutional approach drawn on in this article emphasizes that reform of the formalinstitutions in a national software exports strategy and implementation process will requireknowledge of both the formal rules and the informal norms. Our role was in support ofreform of appropriate institutions towards increased software exports, and we were thusattempting to collaborate with concerned stakeholders to align the formal rules legislatedby policymakers and informal norms of conduct. The stakeholders involved in the policy-making process included a range of actors from various organizations as follows:

• Minister of Science and Technology• Relevant university researchers and faculty• Consultants from the Inter-American Development Bank (IADB) sponsored organi-

zation Prosoftware• Managers from private sector firms including corporate financiers (e.g., Intel)• Members and officials from the software industry association CAPROSOFT• Staff from IDB, banks, venture capitalists, and angel investors



TABLE 2. Formal, Informal Institutions and Contradictions in the Costa Rica Case

Nature of influences

Formal rules Informal norms Contradiction

CAPROSOFT—registry ofassociation is legally defined.Bylaws of this registry.

• “Wine drinking club”• Informal influence of

longer term members tonew sectors joining

• Maintenance of status quo

• “Plateau of comfortableexistence”

• Status quo• “Costa Ricans will not

offend each other”CINDE—Mandate of bringing in

foreign direct investment (FDI)across sectors. Responsible toprivate shareholders, not to theindustry or CAPROSOFT.

Lack of participation orcoordination with thesoftware industry inappropriateness of FDI.

FDI killing the industrythrough the competitiveforces it is generating.

• Officials from foreign direct investment promotion organization la Coalicion Costar-ricense de Iniciativas de Desarrollo CINDE

• Officials from the Costa Rican trade promotion organization Promotora ComericioExterior de Costa Rica PROCOMER

• Officials from the Instituto Costarricense de Electricidad (ICE), the Governmentowned organization responsible for Telecommunications

• Cegesti (researchers/consultants) and ourselves—assigned the task of developing andimplementing the software export strategy making process

Table 3 summarizes the institutional analysis in two selected areas of the policymakingprocess, relating to, first, CAPROSOFT and, second, CINDE, the organization responsiblefor FDI in Costa Rica. The first two columns show our interpretation of the relevant formaland informal rules. The final column shows the contradiction between the attempts of thestrategy process to influence reform of institutionalized practices, formal or informal, andthe associated potential for and challenges to reform.

In the sections to follow, we first discuss the unwillingness to change of the CAPROSOFTactors. CAPROSOFT members comprise the entrepreneurs who were relied upon the mostto reform institutions. Second, we discuss some contradictions of the modus operandi ofCINDE, which, while fulfilling its official mandate of attracting foreign direct investment,did not consider adequately its impact on the local industry.

4.2.1 Reluctance to Change Designated Change Agents. The IADB loan forstrategic planning process in Costa Rica designated CAPROSOFT as the agent for policyreform. A key recommendation of the strategy task force group on the role of CAPROSOFTwas to create a new organization with a wider formal constitution to include both the localand the foreign activities of the software industry, as opposed to its current primarily localmandate. Second, the existing CAPROSOFT was regarded by the consultants (Cegestiand us) and relevant task force members as having a narrow focus in which businessprocess outsourcing and call center firms were excluded from formal membership to theassociation. Exemplars from successful countries such as India suggested that a nationalsoftware association (NASSCOM) should have a wider and all-inclusive mandate to developholistic synergies and have a powerful lobbying mandate with the national government.

We were told by a senior Costa Rican-based academic—who had previously been taskedto write reports for the Costa Rican software industry—that CAPROSOFT was far from



the designated dynamic catalyst. This perception was held by other stakeholders we met,including some CAPROSOFT members who regarded the organization as a “wine-drinkingclub” in which the majority of members had limited motivation or recognition for the needfor reform and preferred the status quo. This is due to the following three reasons: first,the larger and older established software firms perceive themselves as “big fish in a smallpond” where they reached revenues in a stable market to maintain a comfortable lifestyle.International expansion into a strongly competitive market was not seen as critical totheir survival. Such moves were perceived as a threat to their comfortable status quo andmay undermine their existence. A software industry venture capital financier told us in aninterview:

Here, people will be happy to earn a good salary, have a nice car, a nice house, have their childrenin a good school and owning 100% of their companies. And that’s it. So there is a limited level ofpreference of risk taking or ambition.

Second, older, affluent, and established CAPROSOFT members informally influencedthe process by opposing the new sectors (such as call centers and other IT-enabled servicesfirms) from becoming members of the association. Third, we experienced an informallyexpressed cynicism from several of the respondents concerning the origin of the financialaid for the strategy. IADB aid was perceived by some CAPROSOFT members as part of theU.S.’s historical attempts to control several of the Central and Latin American countries andthe violent and negative outcomes that resulted. Apparently, this contributed to some of theCAPROSOFT members negative view of the attempt to reform institutional structures as itrepresented another subversive attempt to perceived “American imperialism.” It is worthyof note that Costa Rica was the last country to join the free trade agreement among theCentral American countries and the U.S., joining only after a very close referendum. As aresult of these informal institutions, IADB-funded initiatives, such as the cost free ISO9001training and accreditation, had a very limited take up among the local industry.

In the midst of these informal institutions, the task force faced a difficult situation increating software export policy. We were also told of an informal social institution amongCosta Ricans that involves avoiding offending each other: Because of the small size of thecountry, there is always the distinct possibility of meeting someone who may request areturn for a favor. This seemed to have the effect of preventing clear and decisive action inthe task force groups for fear of causing offence resulting in negative action in the future.A senior CAMTIC member told us:

Costa Rica is so small. It’s so small that nobody can take the luxury of confronting anybody, becausein no time you will meet again in other circumstances. Most likely the situation will be the other wayaround. So, when you come to live here, you want no enemies. Confrontation is just not worth it. Sopeople prefer to manage relationships.

Costa Rica has the smallest population in Central America (approximately 4 million) andSan Jose, its capital where government and businesses have their headquarters, is a relativelysmall city. San Jose has a population of around 350,000—the size of a neighbourhood inDelhi, Sao Paulo, or Guatemala City.

Due to recommendations from the task force, CAPROSOFT’s formal rules and consti-tution were eventually rewritten to include call center and back office processing firmsand was renamed Camara Costarricense de Tecnologıa de Informacion y Comunicacion



(CAMTIC) in 2005. However, changing the formal and informal institutions came onlyafter a number of senior members had resigned from CAPROSOFT in opposition to theformation of CAMTIC.

4.2.2 Foreign Direct Investment (FDI) Growth Suppressing Local Industry. Akey organizational actor in development of the software industry and exports was CINDE,whose formal mandate was to promote FDI into Costa Rica. Promoting the growth ofthe software industry with targeted FDI was shown to be important in the case of Ireland(O’Riain, 1997). In India too, which attracted Texas Instruments and Motorola, the first firmto reach Capability Maturity Model level 5, and had demonstrative effects in encouragingfurther foreign firms to set up similar operations. These processes spawned a whole newgeneration of entrepreneurs in India and contributed to new institutional structures thatwere previously dominated by family owned businesses.

However, in Costa Rica, the process of encouraging FDI presented difficulties as it wasuncoordinated and FDI entry tended to undermine the growth of smaller indigenous firms.Formally, CINDE was responsible to private shareholders and not subject to any directgovernment controls. An early recommendation of one of the task force groups was torewrite the formal institutions of CINDE so that their FDI promotion efforts would alsorespect the growth needs of the local firms, such as inviting their representatives to be partof decisions in formalizing FDI. These negative perceptions that some respondents had ofthe unilateral role of CINDE was based on the experience of Intel, which set up in CostaRica. Based on higher salaries and the glamour of the multinational corporation image,Intel was able to recruit trained software staff from other local firms. The relatively smallsize of the Costa Rican national pool of software staff contributed to magnify the negativeeffects of Intel’s entry.

Despite considerable pressure exerted by the task force groups, the formal CINDEmandate was never reformed. Although CINDE representatives attended the task forcegroup meetings, there was a definite absence of any real participation or attempts to fostercoordination with the software industry in deciding the appropriateness of candidates forFDI. The consequence of this in 2006 was that several new firms had set up in Costa Ricaand drained the human resource from the smaller firms. There was a definite increasedresistance to CINDE and their FDI attempts, expressed loudly by delegates at the industryconference in 2006. A senior CAMTIC member told us in an interview in 2006:

The companies come in here, set up shop, and hired people and that doesn’t mean that any localcompanies are part of it. That’s my point. Yesterday, we had news that Fujitsu is setting up a hightech call center with 500 people. And we hear similar stories every month.

The historical interests of the private shareholders of CINDE were to support their busi-ness interests, and they thus did not support a reform of the constitution, aims, and objectivesof their organization. Some multinational firms such as Hewlett Packard increased theircapacity by employing local staff and set up offshore outsourcing operations from the U.S.A bidding war between major firms such as Sykes and Proctor and Gamble to poach stafffrom each other has led to increased salaries, which could not be matched by the localindustry. CINDE recognized the problem:

Sykes, IBM, or Hewlett-Packard, or Intel, they have a different type of operation by definition, butthey are demanding the same human resources. (CINDE official)



Higher education and other training colleges have not been able to provide the necessaryhuman resources to meet the fast-growing demand for skilled staff. Several software firmswere unable to afford the increased salary and in 2006, were reported to be on the brink ofclosure. A senior CAMTIC member explained:

The only way for these local companies to compete is to be able to make similar offers like these.And the only way to do that is if they do work for clients that also pay them well. And that won’thappen with local companies working with local customers.

Costa Rican-based firms were unable to recruit from neighbouring Central Americancountries due to formal restrictions on visas and work permits.

5. DISCUSSION AND CONCLUSIONS

The central research question guiding our study is, how do institutional influences createor impede the potential for policy reform in the industry?

It is important to emphasize the complexity associated with trying to introduce policy-level reforms in a politically charged sector like software exports. Given this acknowledge-ment, we were quite modest of what we may be able to achieve within the short time wehad at our disposal. However, our approach was to “sow some seeds of change” in termsof trying to identify some contours of the proposed software exports policy (such as theneed to expand the mandate of the software association) and create some locally basedorganizations (such as the task force groups) by which these contours could be constructed

Institutional analysis of the situation facing software exports policymakers in Costa Ricapresents a “messy problem situation” (Checkland, 1990). From a practical standpoint, theconceptual tools of institutional theory allow a disciplined examination of the impact ofcontext by examining the constraints and enabling influences to policy reform. As shownthrough the case, CINDE would not reform its formal or informal practices to accommodatethe growth of the software exports sector due to other conflicting priorities. The older, moreestablished CAPROSOFT members resisted reform of the formal constitution due to vestedinterests. Institutional theory thus allows an understanding of why certain institutions prevailand others may change, thus offering some potential to debate routes for reform.

There are several theoretical and practical implications of this analysis. First, “factors”models of the process of software exports development have been presented by researchersto show best practices of other countries, some of which can be emulated by followernations (Carmel, 2003; Heeks & Nicholson, 2004). However, our analysis indicates thatthese models do not shed much light on the “why” and “how” of these policy reform efforts.

The “why” of whether these reforms takes place are shaped by the historical rootednessof the formal and informal institutions that may shape policymaking efforts. Escobar (1995)points out the need to consider structures of power, along with local discourses of progressand development. Vested interests, attitudes towards progress, and development, powerstructures, and various other antecedent conditions, as apparent in the Costa Rican case,are masked in decontextualised models (e.g., Heeks & Nicholson, 2004; Carmel, 2003).

The “how” of creating these policy reforms, as attempted in this case, involves movingaway from the assumption that there is a unitary group (typically the national-level policy-makers) who will formulate the vision, which can then logically and rationally diffuse it tobe accepted and implemented by the other stakeholders. The Costa Rican case shows that a



linear, top-down methodology is quite misleading and inappropriate. Our approach was totry and create inter-sectoral task-force groups to approach the process of policymaking inan incremental manner and to provide required resources to support these small steps. Theinter-sectoral membership of the task groups tried to increase the sensitivity to the interlock-ing and dislocating issues and constraints between organizations that influence softwareexports success. Thus, one area of contribution of this article is to show the importanceof the multiplicity of institutional influences on a seemingly “rational” strategy-makingprocess and to critique the view that it is likely to progress harmoniously. The applicationof institutional theory to software exports development provides important insight. Infor-mal institutions are often poorly understood and rarely written down. Practitioners in thisdomain need to acquaint themselves with local formal and informal institutions. In our ownaction research, we were partnering with a local San Jose-based consultancy firm staffedmainly by Costa Ricans but with some consultants from neighbouring Central Americancountries. This alliance helped us to forge our understanding, to a certain degree, of thelocal institutions and events. Much of this was gathered in informal situations and webelieve that research in such scenarios requires ethnographic techniques to attempt to graspthe informal attitudes that comprise institutions.

Our message is not all gloomy and there is potential in Costa Rica, for example, thetask force groups in Costa Rica had some success. As described above, a new industryassociation, CAMTIC, was formed and embraced a wider membership, although therewere some casualties from the old established CAPROSOFT membership. In 2006, themarketing task force successfully bid for further IADB funding (known as ‘Link’) tosupport small firms providing capital funds for their growth together with an incubator forsmall firm development. These initiatives provide an optimism for the future.

ACKNOWLEDGEMENTS

This project was financed by the InterAmerican Development Bank Prosoftware Pro-gramme. We are grateful to our interviewees and collaborators at Cegesti, San Jose, CostaRica.

REFERENCES

Al-Jaghoub, S. (2004). Building a knowledge based economy using ICT for development and the roleof the nation state: A case study of Jordan. Unpublished doctoral thesis, University of Manchester,UK.

Avgerou, C. (2003). Information systems and global diversity. Oxford: Oxford University Press.Arora, A., Arunachalam, V., Asundi, J., & Fernandes, R. (2001). The Indian software services industry.

Research Policy, 30, pp. 1267–1287.Baskerville, R., & Wood Harper, T. (1996). A critical perspective on action research as a method for

information systems research. Journal of Information Technology, 11, pp. 235–246.Camara Costarricense de Tecnologıa de Informacion y Comunicacion. (2006). Costa Rica ICT Sector

3rd International Information and Communication Technologies Summit March 8–9, San JoseCosta Rica. Proceedings published by Procomer (http://www.procomer.com).

Carmel, E. (2003). Introduction to the special issue: The emergence of software exporting industriesin dozens of developing and emerging economies. EJISDC 13(1). Retrieved August 2004 fromhttp://www.ejisdc.org

Checkland, P. (1990). Soft systems methodology in action. Chichester: Wiley.Checkland, P. (1991). From framework through experience to learning: The essential nature of action

research. In H. E. Nissen, H. K. Klein, & R. Hirschheim (Eds.), Information systems research:Contemporary approaches and emergent traditions (pp. 397–403). North-Holland, Amsterdam.



Duong, N. (2004). Software industry development in Vietnam. Retrieved August 2008 fromhttp://www.iipi.org/Conferences/Hawaii SW Conference/Nguyen%20Paper.pdf

Ein Dor, P., Myers, M., & Raman, K. (1997). Information technology in three small developedcountries. Journal of Management Information Systems, 13(4), 61–89.

Escobar, A. (1995). Encountering development, the making and unmaking of the Third World.Princeton, NJ: Princeton University Press.

Heeks, R., & Nicholson, B. (2004). Software export success factors and strategies in developing andtransitional economies. Competition and Change, 8(3), 267–302.

Kambhampati, U. (2002). The software industry and development: The case of India. Progress inDevelopment Studies, 2(1), 23–45.

Ketelhoehn, N., & Porter, M. (2002). Building a cluster: electronics and information technology inCosta Rica. Case 9-703-422, Harvard Business School.

National Association of Software and Service Companies (NASSCOM). (2008). Fact sheet. RetrievedAugust 2008 from http://www.nasscom.in/upload/5216/Strategic Review Feb2008.pdf

Nicholson, B., & Sahay, S. (2005). Human resource development in the context of software exports:Case evidence from Costa Rica. Proceedings of International Federation of Information Processing(IFIP), Working Group 9.4 Conference, Nigeria.

North, D. (1990). Institutions institutional change and economic performance. Cambridge: CambridgeUniversity Press.

O’Riain, S. (1997). The birth of a Celtic tiger. Communications of the ACM, 40(3), 11–16.Paus, E. (2005). Foreign investment, development and globalisation. New York: Palgrave Macmillan.Sautet, F. (2005). The role of institutions in entrepreneurship: Implications for development pol-

icy. George Mason University Mercatus Policy Primer No. 1. Retrieved August 2008 fromhttp://www.mercatus.org/publicationdetails.aspx?id=15800

Scott, W. R. (2004). Institutional Theory: Contributing to a Theoretical Research Program. In K. G.Smith & M. A. Hitt (Eds.), Great minds in management: The process of theory development.Oxford, UK: Oxford University Press.

Silva, L., & Figueroa, B. (2002). Institutional intervention and the expansion of ICTs in Latin America:The case of Chile. IT & People, 15(1), 8–25.

Brian Nicholson is a Senior Lecturer at Manchester Business School, UK. Since 1995, he hasbeen involved in teaching, research and consultancy projects in the broad area of managing globaloutsourcing of software and other business processes. This has involved work in India, China, CostaRica, Iran, Egypt, Malaysia, and Bangladesh. Research at the firm level has resulted in severalinfluential publications in international journals and a book coauthored with Sundeep Sahay GlobalIT Outsourcing (Cambridge University Press, 2003). Policy level consultancy studies have beenundertaken for the governments of Costa Rica and Iran to stimulate software exports. His most recentwork has been commissioned by the Institute of Chartered Accountants in England and Wales: “Riskand Control of Offshore Outsourcing of Accounting Services.”

Sundeep Sahay is a professor at the Department of Informatics at the University of Oslo, Norway.He has held research and teaching positions at the Universities of Cambridge and Salford in the UKand at the University of Alberta in Canada. Sundeep has been actively involved in researching social,organizational and managerial implications of information systems, both in developed and developingcountries. Current research activities include health information systems and issues relating to globalsoftware organizations.


software exports development in costa rica: potential for policy reforms

Documents