factors affecting the performance of farmer companies in sri...

Factors Affecting the Performance of Farmer Companiesin Sri Lanka: Lessons for Farmer-Owned Marketing Firms

H. S. Rohitha RosairoDepartment of Agribusiness Management, Sabaragamuwa University of Sri Lanka. E-mail:[email protected]

Michael C. LyneDepartment of Accounting Economics and Finance, Lincoln University, New Zealand. E-mail:[email protected]

Sandra K. MartinDepartment of Accounting Economics and Finance, Lincoln University, New Zealand. E-mail:[email protected]

Kevin MooreDepartment of Social Science, Parks, Recreation, Tourism and Sport, Lincoln University,New Zealand. E-mail: [email protected]

ABSTRACT

Sri Lanka introduced farmer companies to link smallholders with high-value markets. This study examineshow the institutional, group, and management characteristics of these farmer companies affect theirperformance. Theories about causal relationships are tested using a combination of qualitative andquantitative methods. The results indicate that farmer companies perform better when shares are tradablebetween members, and patrons pay and receive market-related prices for their inputs and products.All directors should be nominated by shareholders and voting should be conducted by secret ballot.To strengthen accountability, the right to hire and fire executive managers should vest with the boardof directors and should not be appropriated by government agencies that facilitate these companies.Although directors and managers should monitor shareholder views, they must take policy and strategicdecisions themselves as failure to centralize decision-making results in severe influence problems. Farmercompany performance is compromised by the absence of well-defined and regularly observed processesto develop and implement strategies, and by inadequate or inappropriate management skills. Thesemanagement problems may diminish if government facilitators had a clear exit plan from the time thecompany is established as this would help to focus their attention on the critical task of empoweringfarmer shareholders to direct and administer their company. [EconLit Citations: Q130, Q010]. C© 2012Wiley Periodicals, Inc.

1. INTRODUCTION

The Comprehensive Framework for Action set out by the United Nations High Level TaskForce (UN-HLTF) on global food security in 2008 prioritized greater public investment in thedevelopment of producer organizations as a long-term strategy to address global food inse-curity (UN-HLTF, 2008). This recommendation is based on strong arguments that horizontalcoordination by farmers can improve smallholder access to markets and technology, whichin turn raise rural incomes and improve the food security of poor people (Dorward, Poole,Morrison, Kydd, & Urey, 2003). In theory, smallholders can gain bargaining power and re-duce unit marketing, processing, compliance, and transaction costs by pooling their produce,and can finance and manage value-adding assets by pooling their capital and centralizingmanagement (Holloway, Nicholson, Delgado, Staal, & Ehui, 2000; Markelova, Meinzen-Dick,Hellin, & Dohrn, 2009; Poulton, & Lyne, 2009; Reardon & Barret, 2000). Agricultural market-ing cooperatives have long been promoted as an appropriate organizational model to realizethese advantages. Evans and Meade (2005) define a farmer cooperative as “an organisation inwhich those who transact with the organisation – patrons – also own and formally control the

Agribusiness, Vol. 28 (4) 505–517 (2012) C© 2012 Wiley Periodicals, Inc.Published online in Wiley Online Library (wileyonlinelibrary.com/journal/agr). DOI: 10.1002/agr.21312

505

506 ROSAIRO ET AL.

organisation, and derive significant benefits from those transactions over and above any finan-cial returns they derive from their investment in the organisation” (pp. 1–7). Another politicallyappealing attribute of a cooperative is that control is democratic in the sense that all membershave equal voting power regardless of differences in their levels of patronage or investment.

Agricultural marketing cooperatives were introduced to Sri Lanka by the British early in the20th century. Historically, the cooperative movement was promoted and supported by the gov-ernment as a strategy to service smallholders and elevate them from subsistence to commercialfarmers (Winslow, 2002). More than 80% of Sri Lanka’s population live in rural areas (CentralIntelligence Agency [CIA], 2012), and smallholdings of less 0.10 hectares account for almost40% of the arable land (Department of Census and Statistics Sri Lanka, 2002). However, thecooperative movement failed to mobilize farm services and products (Senanayake, 2004; Ten-nakoon, 1999). According to Shepherd (2008), experience with the “linking farmers to marketapproach” suggests that attempts to coordinate smallholders have not been consistently replica-ble at scale. Shepherd (2008) attributes part of the blame for this to inappropriate organizationalmodels. Reasons given to explain patchy performance of smallholder marketing cooperativesinvariably include poor management, lack of capital, and political interference (Hoyt, 1989;Mugambwa, 2005; Ortmann & King, 2007). The new institutional economics suggests thatthese problems may well be symptoms of an organization’s weak institutional arrangements.Cook and Iliopoulos (1999) attribute the rapid spread of new generation cooperatives in theU.S. agrifood sector during the 1990s to the efficiency-robbing effects of ill-defined propertyrights in traditional marketing cooperatives. More specifically, Sykuta and Cook (2001) identifyfive institutional problems that discourage members from investing equity capital or retainingprofits in traditional marketing cooperatives; the free-rider, horizon, portfolio, control, andinfluence problems. Problems that constrain levels of equity capital in an organization alsoconstrain levels of debt capital because lenders prefer their loans to be covered by equity in theevent of default (Hendrikse, & Veerman, 2001).

In 1995, Sri Lanka’s National Development Council recommended that smallholders shouldcoordinate their marketing through farmer companies owned and patronized by smallholdersrather than through traditional marketing cooperatives (Esham & Usami, 2007). This re-sponse was motivated by concerns about governance, management, and incentive problems inSri Lanka’s ailing marketing cooperatives (Tennakoon, 1999). Government agencies facilitatedand supported the establishment of these farmer companies, and sometimes appointed advisorydirectors and executive managers from their own ranks (Tuovinen, 2001). Farmer companieswere expected to operate like private companies, owned and controlled by shareholders fromthe surrounding farming community with safeguards to prevent the sale of shares to outsiders(Esham, & Usami, 2007). In reality, these companies adopted a mix of institutional arrange-ments borrowed from traditional marketing cooperatives and investor-owned companies and,in so doing, created a novel form of business organization in Sri Lanka. Some farmer companiesare performing reasonably well, but many have failed (Esham, & Usami, 2007; Ranasinghe,2002). This study set out to capitalize on the Sri Lankan experience by identifying institutional,group, and management factors that distinguish successful from failed farmer companies. Theaim was to recommend best practice for policy and management purposes.

In the next section of this article, we develop a conceptual model outlining causal relation-ships anticipated between a firm’s performance and its institutional, group, and managementattributes. In section 3, we describe the methods used to collect and analyze data. In section 4,we present the key findings and in section 5 discuss some of their policy implications.

2. THE CONCEPTUAL MODEL

A major advantage of an agricultural marketing cooperative is that the interests of its farmer-suppliers (patrons) are well aligned with the incentives of its farmer-owners because the patronsand owners are the same group of farmers. Bekkum and Bijman (2006) point out that thisargument is not entirely true because large patrons may be small shareholders, and vice versa.

Agribusiness DOI 10.1002/agr

PERFORMANCE OF FARMER COMPANIES IN SRI LANKA 507

Clearly, the alignment of patron and owner interests is strongest when investment is directlyproportional to patronage, as in a proportional-investment cooperative (Chaddad & Cook,2004). The alignment of buyer and seller interests in a marketing cooperative is expected toreduce costs associated with designing, monitoring, and enforcing supply contracts, givingthese firms a competitive edge (Sykuta & Cook, 2001). This alignment falls away when farmerscontract with a firm owned by external investors who are more interested in maximizing returnson their equity capital than offering farmers favorable product prices. Marketing cooperativeswould therefore appear to be a useful organizational form to promote farmer access to dis-cerning markets. However, direct proportionality between investment and patronage—a featureof new generation cooperatives—is not a characteristic of traditional marketing cooperatives.Moreover, theory suggests that the institutional arrangements underpinning a traditional co-operative will discourage members and lenders from financing highly specific, capital-intensiveassets required for value adding.

New institutional economics literature has identified five institutional problems in a tradi-tional cooperative (Sykuta & Cook, 2001). There is an “internal free-rider problem” becausea traditional cooperative distributes most of its profits to members according to their levelof patronage rather than their level of investment. An “external free rider problem” arises ifnonmembers are allowed to transact with the traditional cooperative on the same terms asthose offered to members. There is a “horizon problem” because members are not permittedto trade shares at their market price. The implication is that the investors cannot realize cap-ital gains when they leave the cooperative. New members capture some of these gains by notpaying market prices for their shares. The absence of an internal market for member sharesalso creates a “portfolio problem” because members cannot diversify their own portfolios toreflect personal risk preferences. The “control problem” refers to the difficult task of monitoringand sanctioning management when there is no share market, whereas the “influence problem”distances investors from control when voting rights are democratic rather than proportional toinvestment. These equity constraining institutional problems also constrain the creditworthi-ness of traditional cooperatives as lenders prefer their loans to be covered by equity to reducetheir exposure to loan default. Borrowing is further affected by the influence problem, especiallywhen the loan is needed to finance relation-specific assets that expose the borrower to a hold-upproblem (Royer, 1999), as the lender’s interests tend to align more closely with those of majorityinvestors than with those of majority voters.

Sri Lankan farmer companies represent an attempt to encourage investment by farmerpatrons while preserving some of the advantages of well-aligned buyer and seller interests.Although there has been no previous research on the institutional arrangements adopted bythese farmer companies, their mixed performance and cooperative antecedents suggest thecontinued presence of some or all of the institutional problems that constrain access to capitalin traditional marketing cooperatives. Although it is true that these problems may not impairthe sustainability of a firm that does not require much capital for its core business, they arecertainly expected to constrain business growth and the sustainability of firms that do requiresignificant capital. Figure 1 illustrates the causal model of internal factors affecting farmercompany performance proposed in this study. The first (upper) dark lateral arrow representsthe direct effects of institutional weakness on farmer company performance.

Some institutional arrangements, including the firm’s governance arrangements, also haveindirect consequences for performance via their effects on management as they determine whogets to direct and manage the firm, who the directors and managers account to, and the ease andextent to which they can be held responsible for poor decisions. This causality is representedby the long, lightly shaded vertical arrow in Figure 1.

Producer-owned marketing firms represent the interests of a group of farmers as shareholdersand patrons. Homogeneity in member interests can establish solidarity that alleviates influenceproblems (Food and Agriculture Association [FAO], 2000). Conversely, heterogeneity mayencourage members to form subgroups that use their voting power to advance managementdecisions serving their own interests rather than those of the farmer company’s investorsand lenders. Unlike investment-proportional voting rights, democratic voting rights provide


508 ROSAIRO ET AL.

Figure 1 A Model of Internal Factors Affecting the Performance of Farmer Companies.

no safeguard against decisions that favor minority investors over majority investors. Flawedelectoral procedures (e.g., voting by show of hands rather than by secret ballot) and the presenceof nonelected directors on the board could also expose a firm to such an influence problem.The scale of this problem is likely to grow with the range of services offered to shareholderswith different vested interests. These group dynamics could also affect a farmer company’sperformance if ownership and control are not clearly separated (i.e., when ordinary shareholdersparticipate directly in policy and operational decision-making) because such collective actionmay allow powerful individuals and subgroups to influence strategic decisions. Failure on thepart of management to separate ownership from control could therefore usher in an influenceproblem that brings group diversity to bear on farmer company performance. These indirecteffects of institutional weakness on farmer company performance are captured by the twoshort, lightly shaded vertical arrows in Figure 1.

Functions performed by managers of a firm can be described in terms of two distinctivedecision areas; business policy and strategies, and operational processes. Strategy is a high-leveldecision process concerned with responsiveness to, and relationships within, the firm’s businessenvironment (Ansoff, 1969). Strategic choices made by a firm are therefore powerful determi-nants of its performance over time. Operational processes facilitate short-term decisions andare important in the implementation of strategy. In marketing firms where the directors arefarmers rather than business managers, hired managers often assume greater responsibility forstrategic choices (U.S. Department of Agriculture [USDA], 1997). The quality of managers(including directors) has direct and obvious implications for company performance. In thisstudy, management factors are considered to be a key determinant of farmer company perfor-mance and are examined in terms of business policy and strategies, operational processes, andmanagerial quality. The second (lower) dark lateral arrow in Figure 1 indicates the direct effectsof these subconstructs on farmer company performance.

The preceding discussion qualifies some of the propositions flowing from the analysis ofinstitutional arrangements in producer-owned marketing firms. Although farmer companiesare expected to grow faster if they are not burdened with institutional problems, the impactof these problems may be trivial in companies providing services that are not capital intensive.Similarly, farmer companies that separate control from ownership may not experience influenceproblems even if their members have heterogeneous interests and democratic voting rights. The



Figure 2 Sample Design for the Case Studies of Farmer Companies (FCs).

proposed model therefore recognizes that good management may be a cause and not necessarilya result of good governance.

3. DATA COLLECTION AND ANALYSIS

A qualitative investigation was considered most appropriate to understand what and howinstitutional, group, and management characteristics influenced the performance of farmercompanies, and to test propositions about relationships between these variables by comparingcase studies of successful and failed companies. A subset of these variables was also analyzedquantitatively using cluster analysis to triangulate the qualitative findings, and to identifyrelationships not revealed by the qualitative analysis.

Successful and failed farmer companies were paired to reveal common and differentiatingfactors thought to explain their relative performance. To strengthen these paired comparisons, apriori information about farmer companies in Sri Lanka was used to stratify the populations ofsurviving and failed cases into three subgroups according to their core business: companies thatprocess commodities and market value-added products, companies that market commodities,and companies that serve primarily to procure inputs and negotiate better prices for theirshareholders (Figure 2). It was anticipated that institutional problems constraining performancewould be revealed most clearly in the pair of processing farmer companies owing to their higherdemands for capital.

A single-farmer company was then purposefully selected from each stratum. Considerationwas given to the accessibility of shareholders and key informants, availability of annual reportsand financial records, and length of the company’s track record. The selection of farmercompanies was also guided by a priori information about external factors that may haveinfluenced company performance to control these influences when comparing each pair ofsuccessful and failed farmer companies.1

1Rosairo (2010, pp. 21–30) provides a detailed account of data collection. Primary data were collected throughsemistructured interviews with shareholders and key informants using interview guides. Company records (such asannual reports) and external records (such as income tax returns) provided secondary data. These internal and externalrecords triangulated the primary data. Direct observations were also used to validate the interview data. Field workwas undertaken from May to July 2008.


510 ROSAIRO ET AL.

Prior to the qualitative analysis, two checks were made on the validity of pair-wise com-parisons between successful and failed farmer companies. First, the financial and outreachperformance of each successful farmer company (Table 2) was examined to check that thesecompanies were indeed successful. Second, external determinants of farmer company perfor-mance were examined to ensure that each pair of farmer companies experienced similar shifts inmarket and climatic conditions. The qualitative analysis followed the approach recommendedby Yin (2003). Each farmer company was described with emphasis given to its developmentand its institutional, group, and management characteristics. Theoretical arguments were thenchecked by a process of “pattern matching.” Where successful and failed farmer companiesexhibited contrasting attributes consistent with theory, the argument was confirmed. Other-wise, the argument was questioned and the analysis entered its “explanation building” phaseto account for the inconsistency or to refine the theory. Together, these processes identified aset of attributes that distinguished successful and failed farmer companies. The opportunitythen arose to explore quantitative relationships between these attributes and a set of variablesmeasuring company performance. Hierarchical cluster analysis was applied as the objective wasto group-related variables (not cases) and the variables were all binary measures showing thepresence or absence of performance, institutional, group, and management attributes. Knight,Lyne, and Roth (2003) used this approach in their analysis of best institutional practice forfarm-worker equity sharing companies in South Africa.

4. RESULTS

Qualitative analysis of the case studies revealed that the farmer companies were hybrid firmswith different blends of the institutional arrangements that characterize private companiesand traditional cooperatives. The pattern matching exercise identified 34 attributes that dis-tinguished the successful and failed farmer companies. Table 1 expresses these attributes asproblems that undermined farmer company success and groups these problems into four con-structs: institutional arrangements reflecting company rules and practices that created free-rider,horizon, portfolio, control, and influence problems; governance arrangements—while still partof the company’s institutional framework—focussing on concerns about transparency andaccountability; group factors relating to issues that exacerbated influence problems; and man-agement factors reflecting weak processes and skills.

For the purpose of the cluster analysis, each of the problems listed in Table 1 was codedas a dummy variable that scored one if the farmer company did not suffer from the problem,and zero otherwise. Table 1 includes the names of these dummy variables. Table 2 presentsthe indicators of performance—also measured as dummy variables scoring one if the farmercompany satisfied the indicator, and zero otherwise. The cluster analysis identified positiverelationships between the dummy variables (measuring the absence of particular problems) andvarious indicators of performance.

When the six farmer companies studied were ranked based on the presence of favorableperformance attributes, it was found that the farmer company with the highest number offavorable performance attributes had the lowest number of internal problems. Conversely, thefarmer company with the lowest number of favorable performance attributes had the highestnumber of internal problems. These farmer companies were both processors with relatively highcapital requirements. This finding is consistent with the proposition that institutional problemshave less impact on performance when good performance does not require much capital.

Four of the variables listed in Table 1, NOPDIV, VOTING, REVIEW, and ADVSANC,were excluded from the cluster analysis as there was insufficient variation in the sample. Forexample, none of the farmer companies assigned voting rights in proportion to shareholding.The variable MATSUP in Table 2 was also excluded from the cluster analysis as all six farmercompanies received some material support from their respective facilitating organizations andmeasurement of the level of this support was considered too subjective for quantitative analysis.



TABLE 1. Internal Determinants of Farmer Company Performance

Construct Problem Variable

Institutional Non-patrons were retained as permanent shareholders NONPATarrangements Shareholders had to invest equal amounts of equity capital EQINV

Dividends were not proportional to equity invested NOPDIVThe company did not permit share trading between shareholders SHMKTNew shareholders paid the same price for equivalent voting and

benefit rights purchased by the original shareholders (shares werenot appreciated)

SHPRICE

Executive managers influenced policy decisions against majorityinvestors

MGRINFL

The board was not responsible for hiring and firing executivemanagers

HIREFIRE

Non-shareholder-directors influenced policy decisions againstmajority investors

ADVDPOL

The facilitator influenced policy decisions against majority investors FACPOL

Governance Annual General Meetings were not convened regularly AGMREGUarrangements Annual General Meetings were called without adequate notice and

without annual reportsAGMDOCS

Shareholders did not nominate all of the directors SHNOMIElected shareholder-directors were not chosen by secret ballot SECRETBShareholders could not access minutes of the board meetings ACCESSShareholders could not convene special general meetings CONVENENon-shareholder-directors influenced the selection of the chairman CHAIRThe chairman did not set the agenda for board meetings CAGENDAShareholder-directors were not given sufficient information and

time to prepare for board meetingsINFOTIME

Executive manager reported to external agents MREPORTBudgets were not approved by the board BUDGETThe company did not report independently audited financial

statements each yearFINSTAT

Shareholders did not receive an annual report with auditedstatements

ANREPT

Shareholder-directors did not nominate the auditor AUDITORThe company did not prepare financial documents for audit

purposesFINDOC

Group factors Voting rights to elect directors were not proportional toshareholding

VOTING

The company purchased multiple products from its shareholders MULTIPRShareholders participated directly in policy-making decisions SHPOLICYShareholders participated directly in managerial decisions SHMGRIALNon-investors with vested interests in the company had

representation and influence on the boardADVINFLU

Management Organisational structure impeded expansion of the enterprise ORGSTRfactors Company objectives were not developed with and for shareholders OBJECTV

There was no process to review business objectives of the company REVIEWProcesses to develop strategies were not well-established,

continuously applied or informed by managers and shareholdersPROCESS

Processes to implement strategies did not assign clearresponsibilities to managers

IMPLMNT

There were no performance-based incentives or sanctions forappointed officers

ADVSANC

Administrative processes did not produce timely informationrequired by directors, shareholders and legal authorities

ADMIN

The General Manager was not a skilled and a dynamic leader GMSKILSkills of senior managers did not match their portfolios SKILPORT


512 ROSAIRO ET AL.

TABLE 2. Indicators of Farmer Company Performance

Construct Performance indicator Variable

Performance indicators Equity did appreciate GROWTHThe company was solvent SOLVENTThe company was liquid LIQUIDDividends were paid DIVOngoing material support from the facilitating organisation was low MATSUPThe company created its own brand name BRANDThe company developed voluntary links with corporate clients LINKBUYThere was growth in the number of shareholders OREACH

Homogeneity within clusters diminished sharply when the number of clusters dropped belowfour, with the agglomeration coefficient increasing sharply from 1.27 to 1.50.

Figure 3 illustrates positive correlations found between performance indicators and the ab-sence of institutional, governance, group, and management problems. The variable MULTIPRdid not correlate strongly with the performance indicators or with variables in any of the otherclusters. This reinforced the qualitative finding that performance was not adversely affected bymarketing multiple products when company ownership was separated from control.

Cluster 2 includes the majority of indicators in each of the performance, management, group,and institutional (including governance) categories. At a cursory level, this lends support to theargument that good institutions promote company performance directly by improving accessto capital, and indirectly by promoting good management and avoiding costs and conflictsassociated with collective action. However, the results offer deeper insights that question someaspects of this argument.

The performance indicators in Cluster 2 are predominantly measures of financial perfor-mance; share appreciation (GROWTH), solvency (SOLVENT), liquidity (LIQUID), and thepayment of dividends (DIV). These financial indicators are positively related with one anotherand with BRAND, the fifth performance indicator in the cluster, suggesting a positive associa-tion between better financial performance and the introduction of own brands. The finding thatinvestment in branding (a long-term asset) and financial performance were positively linked tothe absence of horizon (SHPRICE), control (HIREFIRE), and influence (MGRINFL, ADVD-POL, and FACPOL) problems supports the proposition that farmer companies will grow fasterif they are not burdened with institutional problems. It is interesting, though, that the problemof executive managers influencing policy decisions against majority investors (MGRINFL) ispositively related to the board’s inability to hold these managers accountable through reportingrequirements (MREPORT) or by hiring and firing them (HIREFIRE).

The absence of influence problems introduced by (unaccountable) managers, advisory-directors, and facilitators also correlated positively with the inability of shareholders and nonin-vestors to engage directly in policy decisions (SHPOLICY and ADVINFLU). This reinforcedthe qualitative finding that centralized decision-making improved company performance bymitigating the adverse effects of group heterogeneity diversity and lends support to the propo-sition that separation of ownership and control can mitigate influence problems when membershave heterogeneous interests and democratic voting rights. The presence of governance vari-ables SECRETB and SHNOMI in Cluster 2 further suggests that failure to elect directors bysecret ballot from a pool of candidates nominated by the shareholders could lead to influenceproblems that constrain access to capital.

Cluster 2 also shows positive correlation between good financial performance and governancearrangements that promote financial transparency and accountability within the company. Theindicators of financial transparency included preparation of financial statements for annualaudit (FINDOC), distribution of an annual report containing audited financial statements(ANREPT), access to the minutes of board meetings (ACCESS) and providing adequate noticeof meetings and timely information (AGMDOCS and INFOTIME). Accountability indicators



Figure 3 Relationships Between Variables Measuring Farmer Company Performance and Institutional,Governance, Group, and Management Attributes.


514 ROSAIRO ET AL.

included provisions to call special general meetings (CONVENE), holding regular annual gen-eral meetings (AGMREGU), and shareholder nomination of the auditor (AUDITOR). Thesequantitative results were consistent with the qualitative findings and support the argument thatgood governance promotes company performance.

All but one of the management indicators appears in Cluster 2. Although the model illustratedin Figure 1 is not explicit about relationships between management indicators, it was notsurprising to find positive associations between good leadership (GMSKIL), the fit betweenmanagers’ skills and their portfolios (SKILPORT), good managerial processes (PROCESS,IMPLMNT, and ADMIN), and an appropriate organizational structure (ORGSTR). Thatthese variables appeared in Cluster 2 along with those reflecting centralized decision making,transparency, and accountability lends support to the proposition that good management maybe a cause and not necessarily a result of good governance.

Cluster 3 shows positive relationships between just one indicator of performance(LINKBUY), two governance variables (BUDGET and FINSTAT), and one group variable(SHMGRIAL). LINKBUY indicates the presence of a long-term, voluntary relationship be-tween the farmer company and one or more its corporate clients. BUDGET and FINSTASTindicate the absence of problems that tend to reduce financial transparency, whereas a positivescore on SHMGRIAL shows reduced exposure to influence problems and high transactioncosts anticipated when shareholders participate directly in managerial decision making. Therelationships highlighted by Cluster 3 suggest that corporate clients are more likely to commit toa farmer company that displays financial transparency and management that is not influencedby (minority investor) interest groups.

Cluster 4 includes just one performance variable, OREACH. This variable indicates growth inthe number of farmers who join the company as shareholders. OREACH is positively correlatedwith the institutional variables NONPAT, EQINV, and SHMKT. A positive score on NONPATindicates reduced exposure to the internal free-rider problem as patrons did not benefit frominvestments made by nonpatrons. A positive score on EQINV indicates reduced exposure tothe portfolio problem because investors were not required to contribute equal amounts ofequity capital. A positive score on SHMKT suggests reduced exposure to the horizon problembecause investors were at least entitled to trade shares with other members. These results supportarguments that farmers are more likely to buy shares in farmer-owned marketing organizationsthat are not exposed to internal-free rider, portfolio and horizon problems.

Cluster 4 also shows positive links between OREACH, the management variable OBJECTV,and the governance variables CHAIR and CAGENDA. The governance variables indicate thatboard decisions were taken by shareholder-directors and not by advisor-directors. These resultsreinforced the qualitative finding that company objectives were more likely to be establishedwith, and for, shareholders when the board was free of controlling influences from externallynominated directors. In addition, the quantitative results suggest that company objectivesestablished in consultation with prospective shareholders attracted more investors over time.

5. CONCLUSIONS AND RECOMMENDATIONS

The results of the qualitative and quantitative analyses indicate that farmer companies aremore likely to attract capital and invest in value-adding assets (like brands) when the horizonproblem is alleviated by making benefits directly proportional to investment. This means thatshares should be appreciable and that dividends should not be used to subsidize patronsby offering them better than market-related prices for products and inputs. This break withtraditional cooperative practices would address the internal free-rider problem, alleviate thehorizon problem, and allow farmer companies to extend their services to nonshareholderswithout creating an external free-rider problem.

Growth in the number of shareholders (outreach) was constrained by a portfolio problem infarmer companies that expected their members to invest equal amounts of equity capital. Toavoid this problem, equity shares should be tradable between members. Government agencies



that facilitate farmer companies (or new generation cooperatives) should help them to establishtrading platforms and to disseminate information about prices and quantities of shares traded.This information would signal company performance to shareholders and so alleviate thecontrol problem. Outreach was also constrained in farmer companies that arbitrarily limitedmembership to a small geographic area. Although geographic limitations on membershipshould be lifted, the right to admit new members should remain with the board of directors.

The analyses showed that existing shareholders are more likely to withdraw their patronageand support when farmer companies suffer control and influence problems. Perceptions thatpolicy decisions could be influenced against the interests of majority investors by advisory direc-tors and executive managers who were not nominated for election to the board by shareholderswere particularly damaging to investor confidence. Influence problems were also attributed toflawed electoral procedures. It is clear that all directors should be nominated by shareholdersand that voting should be conducted by secret ballot. To improve accountability, the right tohire and fire executive managers should vest with the board of directors and should not beappropriated by government agencies.

Some of the theoretical arguments could not be tested empirically owing to a lack of vari-ation in the case studies. For example, all of the farmer companies applied democratic ratherthan investment-proportional voting rights so it was not possible to test the argument thatdemocratic voting rights (one vote per member) undermine performance because they offer noprotection against influence problems. However, the results indicated that there is less need forthis protection when ownership and control are clearly separated. Centralized decision makingwas also shown to mitigate influence problems associated with heterogeneous interests (evi-denced by marketing multiple products), and to promote long-term business relationships withcorporate clients. Indeed, respondents identified management’s failure to separate ownershipfrom control as a leading cause of poor performance.

Leaving decisions in the hands of (accountable) directors and managers does not imply thatordinary shareholders’ views are unimportant when formulating company policy and businessstrategies. On the contrary, the results of this study suggest that directors and managers shouldestablish forums and processes to elicit the views of shareholders (and other stakeholders) onpolicy and management issues. The results also indicated that farmer company outreach ismore likely to grow over time if the facilitators canvas prospective shareholders widely whenestablishing the company, and ensure that there is broad agreement on its objectives andawareness of its institutional arrangements.

Viewed from an operational perspective, the analyses show that company performance (in-cluding its ability to forge long-term business relationships) is compromised when managementfails to comply with standard reporting and audit requirements, and does not facilitate meetingscalled by shareholders. This is consistent with the argument that the erosion of transparencyand accountability within a company tends to undermine the confidence of investors, lenders,and other strategic partners.

Performance was also compromised by the absence of well-defined and regularly observedprocedures to develop and implement new strategies. These weaknesses were linked to inad-equate or inappropriate management skills. Although the results of the study are consistentwith the argument that good institutions and governance promote good management, theyalso support the proposition that good managers promote good governance. The qualitativeanalysis suggested that management problems would diminish if facilitators had a clear exitplan from the time the company is established as this would focus their attention on the impor-tant task of empowering small farmers to manage their company. A farmer company that isable to sustain an internal market for its shares could also offer shares as an incentive for goodmanagement. This would have the added advantage of aligning the interests of managers withthose of shareholders.

The results presented in this article cannot be generalized to the population of farmer compa-nies in Sri Lanka owing to the small sample size. Nevertheless, they can be generalized to theory,and the recommendations do therefore warrant consideration in Sri Lanka and other developingcountries grappling with underperforming smallholder marketing organizations. In doing so, it


516 ROSAIRO ET AL.

is recognized that some of the key recommendations may be resisted at political, government,and community levels. Acceptability at these levels is a question for future research. Relatedto this is the need to prioritize problems constraining company performance—something thatwas not addressed in this study.

REFERENCES

Ansoff, H.I. (1969). Toward a strategic theory of the firm. In H.I. Ansoff (Ed.), Business strategy (pp. 11–40).Harmondsworth, England: Penguin.

Bekkum, O.F.V., & Bijman, J. (2006, May/June). Innovations in cooperative ownership: Converted and hybrid listed co-operatives. Paper presented at the 7th International Conference on Management in Agrifood Chains and Networks,Ede, The Netherlands.

Chaddad, F.R., & Cook, M.L. (2004). Understanding new cooperative models: An ownership-control rights typology.Review of Agricultural Economics, 26, 348–360.

Central Intelligence Agency. (2012). The world factbook. Retrieved March 22, 2012, fromhttp://https://www.cia.gov/library/

Cook, M.L., & Iliopoulos, C. (1999). Beginning to inform the theory of the cooperative firm: Emergence of the newgeneration cooperative. Finnish Journal of Business Economics, 4, 525–535.

Department of Census and Statistics Sri Lanka. (2002). Agricultural and environmental statistics. Retrieved March22, 2012, from http://www.statistics.gov.lk/agriculture/AGC2002/AGC2002.htm

Dorward, A., Poole, N., Morrison, J., Kydd, J., & Urey, I. (2003). Markets, institutions and technology: Missing linksin livelihoods analysis. Development Policy Review, 21, 319–332.

Esham, M., & Usami, K. (2007). Evaluating the performance of farmer companies in Sri Lanka: A case study of RidiBendi Ela Farmer Company. The Journal of Agricultural Sciences, 3, 86–100.

Evans, L., & Meade, R. (2005). The role and significance of cooperatives in New Zealand agriculture: A comparativeinstitutional analysis, report prepared for New Zealand Ministry of Agriculture and Forestry. Wellington: NewZealand Institute for the Study of Competition and Regulation.

Food and Agriculture Association. (2000). Small farmer group associations: Bringing the poor together, summary ofproceedings–Part I. Paper presented at the Small Farmer Group Associations: Bringing the Poor Together, e-mailconference, 24th September - 11th November 1998.

Hendrikse, G.W.J., & Veerman, C.P. (2001). Marketing cooperatives and financial structure: A transaction costseconomics analysis. Agricultural Economics, 26, 205–216.

Holloway, G., Nicholson, C., Delgado, C., Staal, S., & Ehui, S. (2000). Agroindustrialization through institutional in-novation: Transaction costs, cooperatives and milk-market development in the east-African highlands. AgriculturalEconomics, 23, 279–288.

Hoyt, A. (1989). Cooperatives in other countries. In D. Cobia (Ed.), Cooperatives in agriculture. Englewood Cliffs,NJ: Prentice Hall.

Knight, S., Lyne, M., & Roth, M. (2003). Best institutional arrangements for farm-worker equity-share schemes inSouth Africa. Agrekon, 42, 228–251.

Markelova, H., Meinzen-Dick, R., Hellin, J., & Dohrn, S. (2009). Collective action for smallholder market access.Food Policy, 34, 1–7.

Mugambwa, J. (2005). The saga of the cooperative movement in Papua New Guinea. Journal of South Pacific Law,9(1).

Ortmann, G.F., & King, R.P. (2007). Agricultural cooperatives II: Can they facilitate access of small-scale farmers inSouth Africa to input and product markets? Agrekon, 46, 219–244.

Poulton, C., & Lyne, M. (2009). Coordination for market development. In J. Kirsten, A. Doward, C. Poulton, & N.Vink (Eds.), Institutional economics perspectives on African agricultural development (pp. 143–183). Washington,DC: International Food Policy Research Institute.

Ranasinghe, I. (2002). Failure of farmer companies of Mahaweli areas of Sri Lanka. Unpublished Master’s thesis,Sabaragamuwa University of Sri Lanka, Blihuloya, Sri Lanka.

Reardon, T., & Barret, C.B. (2000). Agroindustrialization, globalization and international development: An overviewof issues, patterns and determinants. Agricultural Economics, 23, 195–205.

Rosairo, H.S.R. (2010). Factors affecting the performance of farmer companies in Sri Lanka. Unpublished doctoraldissertation, Faculty of Commerce, Lincoln University, New Zealand.

Royer, J.S. (1999). Cooperative organizational strategies: A neo-institutional digest. Journal of Cooperatives, 14, 44–67.Senanayake, M.S. (2004, Nov. 13–14). What is ailing farmer companies in their transformation into successful business

entities? Paper presented at the Third International Conference of the Japan Economic Policy Association, MeijiUniversity, Tokyo, Japan.

Shepherd, A.W. (2008). Experiences with the ‘linking farmers to markets’ approach in enhancing the performance ofhorticultural supply chains in the transition economies. In P.J. Batt (Ed.), Proceedings of the Second InternationalSymposium on Improving the Performance of Supply Chains in the Transition Economies. Acta Horticulturae,794, 309–316.

Sykuta, M.E., & Cook, M.L. (2001). A new institutional economics approach to contracts and cooperatives. AmericanJournal of Economics, 83, 1273–1279.



Tennakoon, D. (1998, Sept. 21–29). Sri Lanka – Country paper on role of institutions in rural community development.Paper presented at the Asian Productivity Organisation Study Meeting on Role of Institutions in Rural CommunityDevelopment, Colombo, Sri Lanka.

Tuovinen, K. (2001, Feb. 16–17). Flows of the Mahaweli: The formation and reformation of farmer companies andorganizations. Paper presented at the 16th Annual South Asian Conference, University of California at Berkeley.

U.S. Department of Agriculture. (1997). Cooperative management (No. 1). Cooperative information report, RuralBusiness-Cooperative Service. Washington, DC: Author.

United Nations High-Level Task Force (UN-HLTF). (2008). Elements of the comprehensive framework for actiondeveloped by the UN System High Level Task force (HLTF) on the global food security crisis. Retrieved August 3,2008, from http://www.un.org/issues/food/taskforce/cfa.shtml

Winslow, D. (2002). Co-opting cooperation in Sri Lanka. Human Organization, 61, 9–20.Yin, R.K. (2003). Case study research: Design and methods (3rd ed.). Thousand Oaks, CA: Sage.

H. S. Rohitha Rosairo is a senior lecturer in agribusiness at Sabaragamuwa University of Sri Lanka.He received his BSc in Agriculture from the University of Peradeniya, Sri Lanka, his MBA from ColomboUniversity, Sri Lanka, and his PhD from Lincoln University, New Zealand. Current research interests includeagribusiness and development in developing economies, institutional arrangements to promote smallholderaccess to markets, and farmer organizations.

Michael C. Lyne is an associate professor in international rural development at Lincoln University, NewZealand. He is also an honorary professor at the University of KwaZulu-Natal in South Africa. He received hisBSc in Agricultural Economics, an MSc in Agricultural Economics, and his PhD in Agricultural Economicsfrom the University of KwaZulu-Natal, South Africa. His current research interests include land tenureand land markets, rural finance and financial products to improve access to land markets, institutionalarrangements to promote smallholder access to supply chains and to manage shared resources, and institutionalarrangements to promote partnerships between land reform beneficiaries and agribusiness investors.

Sandra K. Martin is an associate professor in agribusiness for Development at Lincoln University, NewZealand. She received her BEcon from the University of Queensland, Australia; her MA and PhD from theUniversity of Canterbury, New Zealand. Her current research interests include agribusiness and developmentin emerging and developing economies, agribusiness value chains, supply chains, indigenous entrepreneurshipand poverty alleviation, and linking producers to markets through pro-poor initiatives.

Kevin Moore is a senior lecturer in psychology at Lincoln University, New Zealand. He received his BSc(Hons) and PhD from the University of Canterbury, New Zealand. His current research interests includetourist decision making and experience and theoretical psychology, especially social constructionist anddiscursive theories of consciousness and the person, and conceptual issues in evolutionary psychology.


factors affecting the performance of farmer companies in sri...

Documents