Agricultural Technolongy in Bangladesh: a Study on Non-Farm Labor and Adoption by Gender

There is growing interest in learning the impacts of agricultural technologies especially in developing economies. Economic analysis may entail assessment of employment and time allocation effects of new technologies. An issue of importance in South Asia is the impacts of technological change on a specific type of occupation: rural non-farm activities. In order to fully understand these effects, the research must integrate gender differences and determine if the results would be similar irrespective of gender.

This paper particularly looks at the effects of HYV adoption on time allocation and labor force participation of men and women in non-farm activities. In estimating the effects of HYV adoption on non-farm labor supply, information on the dependent variable, supply of non-farm labor (or the number of days worked while engaged in non-farm labor), is not available for individuals who do not participate in non-farm labor. Hence sample selection or self-selection of individuals occurs. A feasible approach to the problem of sample selection is the use of Heckman's Two Stage Selection Correction Model. Income functions were estimated for males and females while correcting for the sample selection of non-farm wage earners.

An enhanced understanding of the conceptual links among HYV adoption, non-farm labor supply, and gender issues is achieved by discussing the Farm Household Model. The constrained maximization which is drawn from the Farm Household Model would bring about demand functions and reduced form functions for adoption and labor supply. The reduced-form equations are estimated at the individual level for the following: adoption of HYV technology in rice cultivation, and non-farm labor supply of both adult males and females. Regression results are presented for both Ordinary least squares (OLS) and Tobit estimates.

HYV adoption and non-farm labor supply of men and women are influenced by several factors in Bangladesh. The household characteristics assumed to potentially determine technology adoption and non-farm labor decisions are the following: non-farm wages per month of the males and females, farm size, asset value, ratio of yield per decimal land of high-yielding to traditional variety of rice, HYV yield, local variety yield, and the ratio of variance of yield per decimal land of HYV to traditional or local varieties.

The empirical findings suggest that the decision to adopt HYV technology is determined primarily by farm size, value of total assets of the household, ratio of yield per decimal of land of high-yielding to traditional variety of rice, and the ratio of variance of yield per decimal of land of high-yielding to traditional variety of rice. A larger farm size or land owned in decimal unit increases the non-farm labor supply of females, but not of men. HYV yield is significant and positive, while the local variety yield is significant and negative. This means that higher HYV yields increase the supply of non-farm labor of women, while higher local or traditional yields lower women's supply of non-farm labor.