Determinants of Technical Inefficiency 0f Savings and Credit Co-Operatives in Kenya

Abstract

The co-operatives sector in Kenya has gone through a historical development which is known for inefficiency. This study is based on the census collected secondary data analyzed from the audited financial statements of the licensed deposit taking Saccos and the macro-economic indicator data sources over the research period. It focuses on environmental and Saccos specific independent variables affecting inefficiency of Saccos. Further, it measured the technical (pure technical) inefficiencies of Saccos during a period of pre-regulation and regulation. The explanatory research design was used. The study was designed to address the following3 objectives: to measure the extent of management inefficiency over a period of two eras, to determine the effect of Saccos specific predictor variables on Saccos’ inefficiency, and establish the effect of macro-economic variables on Saccos’ inefficiency. The specific financial institutions theories that guided this study include: the financial intermediation theory, x- inefficiency theory and agency theory. The econometric methods employed in measuring inefficiency include the non-parametric together with stochastic frontier analysis. These methods were subjected to a panel data census of 46 Saccos to determine the inefficiency during the combined eight years’ period covering 2007-2014. The result of the study shows that all predictors (given loan to members, dividend, net operating cash flows and total revenue dependent variable slacks) were significant as hypothesized in agency, x-inefficiency and intermediation theories. The impact of control variables, reflected mixed results. The number of women on the board, net profit to total assets, and capital adequacy predictors for instance, had a strong negative influence on dividends and loan to members’ slacks without the control variables effect (all these predictors reflected p-values of 0.00 at 1% level of significance).Further, the total revenue slack with control and without control variables regressions explain the significant management inefficiency as indicated by Gamma of 0.999956 and 0.999951 (with optimal of Gamma p-value of 0.00) respectively. Data envelopment analysis result of 368 observations indicated mean efficiency score of 0.976. The strong efficient Saccos with zero slacks across all output variables totaled 6.5% composed of small (1.4%), medium (0.8%) and large (4.3%) while, 26% (94 out of 368) were weak form efficient and the balance being inefficient. This study’s contribution to practice is that total revenue is key output slack in detecting management inefficiency or influence on inefficiency. For contribution to theory, it strengthens the conflicting prior studies on the effect of gender on the board of companies as the result supports the theory that higher number of women on the boards of Saccos reduces inefficiency in line with both agency and intermediation theories. On the contrary the higher technical inefficiency mean difference during regulation period conflict with the agency theory. Further, the significance of inefficiency predictor variables such as net profit to total assets, market power, capital adequacy, financial investment and technology expenditure in financial reports of Saccos and efficiency benchmarking using data envelopment analysis and stochastic mechanism are apt for decision making and will assist the regulators monitor better. Vigilant monitoring of the trend of dividend rates and total revenue levels, and coherent merger of small to medium sized Saccos are also recommended in order to reduce or eliminate the management inefficiency. The age predictor not in support of the learning effect had positive influence onloan to members, net operating cash flows and dividend slacks. Also to be researched in the future studies is the comparative inefficiency study on deposit and non-deposit taking Saccos