Imperfections in Membership Based Organizations for the Poor

Imperfections in Membership Based Organizations for the Poor An Explanation for the Dismal Performance of Kenya’s Coffee Cooperatives

Objective n n To show that the success of membership based organizations of the poor is sensitive to the underlying sociopolitical environment and that in certain cases, releasing ownership and management to groups of individuals can unleash incentives that work against the benefits of collective coordination and result in the diminished welfare of its members. I highlight this issue from the prospective of the smallholder coffee industry in Kenya.

Institutional Changes Impacting the Smallholder Coffee Sub-Sector n Why Cooperatives? n n Liberalization n n Fragmented nature of production Large fixed costs Inadequate infrastructure, missing markets 1998: New Cooperative Act: Government relaxes control on cooperatives. More decision making power given to them Present situation n n Bankrupt Cooperatives Deteriorating Factories and Infrastructure Generalized Corruption and Political Manipulation Neglected coffee trees

Hypothesis n The deterioration of coffee cooperatives can be partly explained by the institutional changes in cooperative organization that gave full ownership and administrative control to its members.

Vulnerabilities in Institutional Design n Perfect Vote Signaling n n Practice of Mlolongo facilitates vote-buying Local Monopsony Power n Protects cooperatives from competition thus dampening the incentives for efficiency and increasing the returns to rent-seeking

Analytical Model n Brief Summary Large expected rents increase incentive for corrupt candidates to buy their way into office n Voters accept any bribe at least equal to the expected loss of welfare should they be pivotal voter. n Likelihood of being pivotal voter decreases with increasing members. n

Data Analysis n n Nine of 19 Coffee Cooperatives in Murang’a District purposively selected. Random selection of factories and coop members serving these factories for farm level surveys. Empirical Strategy n n Goal: To test for the presence of rent-seeking behavior in cooperatives and show that it has a detrimental effect of the technical efficiency of members. The separate but interrelated tests n Stochastic production frontier estimation to extract farmspecific technical efficiency n Determinants of Payments to Farmers n Determinants of Farm-Specific Technical Efficiency

Stochastic Production Frontier Estimates *** - Significant at the 99% level ** - Significant at the 95% level * - Significant at the 90% level

Density Technical Efficiency Kamacharia Gaturi Weithega Kanyenyaini Kahuhia Iyego Kiru Kanguno Kiriti

Determinants of Payments to Farmers *** - Significant at the 99% level ** - Significant at the 95% level * - Significant at the 90% level

Mean Cooperative Payments by Average Members per Factory Mean Cooperative Payments by Total Members per Cooperative

Sources of Inefficiency *** - Significant at the 99% level ** - Significant at the 95% level * - Significant at the 90% level

n Summary n n A portion of the decline in cooperative performance can be attributed to the gross level of corruption/management incompetency present Certain features of institutional environment underlying coffee cooperatives undermine its effectiveness Collective organizations do not always lead to paretoimprovements for their members. Policy Implications n n Require that elections are carried out by secret-ballot in the presence of objective election supervisors Remove legal monopsony protection and allow farmers to sell to highest bidder Creation of effective formal regulatory mechanisms with prosecuting powers Improved access to credit and extension advice

Thank you for your attention

Testing the Hypothesis n n Crisis of Kenya’s coffee sector cannot be explained by poor world prices Highlighting vulnerabilities in components of Institutional Design Mlolongo Voting Tradition n Local Monopsony Power n n n Analytical Model Empirical Evidence