Farm Wives' Labor Force Participation and Earnings

Abstract

The decisions of farm wives to work off the farm and the earnings they make in that off-farm employment should be considered simultaneously. Previous studies of wives' off-farm earnings have included only employed wives in their analyses of the factors affecting earnings, which results in biased estimates. This study tests, via Tobit analysis, a model which includes all farm wives and examines the effects of wives' human capital, farm and family constraints, and labor market characteristics on both their off-farm employment decisions and their earnings. Wives' off-farm earnings are found to be related to wives' education, labor market experience, presence of children, other family income, farm size, and debt/income ratio. Changes in these factors have a greater influence on the labor market participation decisions of farm wives than on the variation in their earnings, once employed.

The Effects of Metropolitan Residence on the Off‐Farm Earnings of Farm Families in the United States1

“…The bivariate probit, a variant of Heckman's method, is also inappropriate and too complex given the repeated theoretical and empirical evidence that husbands and wives' labor allocation decisions are not made jointly (Buttel and Gillespie 1984;Huffman and Lange 1989;Lass et al 1989;Simpson et al 1988). Second, tobit coefficients can be decomposed to assess the effects of individual variables on both participation and earnings and how changes in participation affect changes in earnings (Godwin and Marlowe 1990;McDonald and Moffitt 1980;Norris and Batie 1987). This has yet to be done with the Heckman specification and its modification, which at present can only indicate whether selectivity in the choice to participate affects earnings.…”

“…Estimates from a tobit model are normalized coefficients that must be transformed to be meaningful measures of the effect of changes in the independent variables on the dependent variables. The decomposition method used in this analysis is attributed to McDonald and Moffitt (1980) and to Godwin and Marlowe (1990) who refined it. Further details on the tobit specification and decomposition method can also be found in these publications.…”

“…Past studies that have taken explicit account of local labor market conditions show mixed effects. Local unemployment rates, industrial and occupational structure, population size, distance to the nearest city, and region have been used as measures of the impact of the local economy (Albrecht and Murdock 1984;Deseran and Dellenbarger 1988;Deseran et al 1984;Fuguitt 1958Fuguitt , 1959Godwin and Marlowe 1990;Huffman 1977;Lass et al 1989;McNamara and Gunter 1989). Generalized findings suggest that the effects of the local economy are weak, but some are in the expected direction.…”

“…Linked family records can be generated from the CPS because it is a household-level survey that gathers information on all members. Smaller surveys, such as those used by Buttel and Gillespie (1984), Simpson et al (1988), Deseran and Dellenbarger (1988), Godwin and Marlowe (1990), Lass et al (1989), and Huffman and Lange (1989), provide detailed information on household, individual, and farm characteristics but have limited geographic scope, which may not include metropolitan areas within the sampling frame. In addition, they cover specialized agricultural areas, which, in turn, are adjacent to distinctive nonfarm labor markets.…”

“…Mixed empirical evidence exists on how labor market characteristics as measured by a variety of indicators affect off-farm earnings (Deseran et aI. 1984;Godwin and Marlowe 1990;Huffman and Lange 1989;Lass et aI. 1989;Rosenfeld 1985).…”