Urban-Rural Income Transfers in Kenya: An Estimated-Remittances Function

Abstract

The quantitative magnitude and empirical determinants of urban rural remittances for Kenya were investigated using data on the average monthly amount of money urban workers send to rural areas along with the joint distribution of several of their socioeconomic characteristics. In the spring of 1971 the Institute for Development Studies of the University of Nairobi conducted a survey of African households in Nairobi. The sample was confined to low and middle income areas of the city. The survey schedule included questions on basic socioeconomic variables as well as questions on income remittances. Of the 1140 males in the sample who had some income in December 1970 88.9% responded that they regularly sent some money out of Nairobi. The average amount remitted was 85.7 shillings per month; 20.7% of the sample urban wage bill was remitted. Most of the money was intended for consumption by the extended family. 2 sets of regression results are presented regarding the relation between the total income transferred out of Nairobi each month (T) as a function of monthly income in Kenya shillings. For incomes in excess of about 1600 shillings per month the estimated marginal propensity to remit was negative but was not significantly different from zero. Each child residing in a rural area increased the fraction of income remitted by .0164; each child in Nairobi reduced the fraction by .0188. The absolute values of these estimated coefficients were not significantly different from one another. Urban rural income transfers represent about 1/5 of the urban wage bill in Kenya. The analysis showed that the amount which individuals transfer is systematically related to income and other socioeconomic variables. To the extent that rural and urban residence is a useful distinction the magnitude of urban rural income transfers implies a very significant increase in rural welfare from what is implied by comparisons of relative incomes alone. The results imply that the welfare of the typical individual in Kenya depends rather significantly on the number and closeness of relatives working in the high wage sector. A general increase in the urban wage level has the effect of lowering the fraction of the wage bill remitted to rural areas and lowering the level of employment. The net effect of an increase in the urban wage on aggregate urban rural transfers is positive only if the elasticity of transfers with respect to income exceeds the absolute wage elasticity of labor demand.