Determinants of Rural Household Choice of Income Dependency Strategies in Eastern Mau, Kenya

Abstract

The rural households living in the forest fringes of Eastern Mau have become poor over the last five decades due to constraints related to socio-economic and demographic characteristics, low holding asset and contextual factors. As a result, rural households have been unable to make optimal decisions to pursue more remunerative income dependency strategies. These shortcomings are the causes of household regular and forest-based income underperformances. In this paper, we examine the factors that influence the rural household choice of income dependency strategies. Primary data in the study area were collected purposively selected from six-administrative locations that straddle Molo and Njoro sub Counties. Rural household respondents were those living in forest margins located in a four-kilometer radius away from the forest protected areas. Semi-structured questionnaires survey instruments and interviews were used to collect the data. The main objective of the study was to determine the factors that impede rural households from making optimal choices of income dependency strategies. Household income dependency strategies, like on farm income, off-farm income, mixed-income, transfer income and forest-based income. A multinomial logistic regression model was used to identify the predictor variables that influence the household choice of income dependency strategies. The variables of the model are household socio-demographics, asset holdings and contextual factors. In the analysis, the model used estimated coefficients, log-odd ratios, or odds-ratios and marginal effects to reveal the thirteen out of fifteen measured indicator variables. These predictor variables of the model influenced the choice of household income dependency strategies. The results of the analysis of the multinomial model show the likelihood ratio (LR) of the multinomial logit model analysis based on Chi-square tests show significance at the 1% level of significance (LR Chi2 (60) = 1680.04, Prob > Chi2 = 0.0000). Equally, the analysis of estimated coefficients, odd-ratios and marginal effects demonstrate that at least one of the predictor variables has a significant influence on the response variables. This study recommends that state-actor policymakers should invest in the embedment of household livelihood outcomes into efficient conservation and management of forest ecosystem resources. This strategy ensures that there is an increase in regular household on-farm income activities. Also, this will increase total household income which reduces household poverty and over dependence on forest ecosystem resources. The reduced dependence on forest ecosystem resources reduces its degradation and loss of biodiversity in the long term.