University of Wisconsin - Madison Institute for Research on Poverty

This paper chal lenges the widely c i t e d f inding of Groeneveld, Hannan, and Tuma t h a t the Seattle-Denver Income Maintenance Experiment provides evidence t h a t guaranteed income plans f o r poor husband-w i f e fami l i e s w i 11 inc rease mar i ta l d i s so lu t ions . The conclusion of t h i s paper i s t ha t the plans ( s p e c i f i c a l l y , the negat ive income tax plans in the experiment) had no e f f e c t on the r a t e of mar i ta l d i s so lu t ions among the "treatment" couples r e l a t i v e to the cont ro l couples. The con t ro l couples were e l i g i b l e f o r the e x i s t i n g program, Aid to Fami l i e s with Dependent Children. Our r eana lys i s of the experimental data dis t inguishes between the experimental treatment in the form of the "pure" negat ive income tax and the t reatment plans tha t involved an experimental t r a in ing program. We use a l l the time periods of the experiment, allow fo r the timing of the marit a l d i s so lu t ion i n our inferences, and allow for a t t r i t i o n and r e c o n c i l i a t i o n s . A Reanalysis of Marital Stability in the Seattle-Denver Income Maintenance Experiment by Glen G. Cain and Douglas A. Wissoker A LANDMARK STUDY IN POLICY RESEARCH An article by Hannan, Tuma, and Groeneveld in 1977 in the American Journal of Sociology was the first published report that the Seattle-Denver Income Maintenance Experiment (SIME-DIME) had the effect of increasing marital instability among couples who were participants in the experiment.' The results startled and dismayed advocates of the welfare reforms being tested in the experiment, because t,hey had expected that the reforms, which extended income transfer payrnent:~ to poor husband-wife families, would stabilize marriages relative to the existing program, Aid to Families with Dependent Children (AFDC), which essentially provided benefits only to poor families without a father present. AFDC had itself been frequently blamed for contributing to the rising trends in marital breakups and female-headed families. Although SIME-DIME, like three other social experiments with income maintenance plans sponsored by the (then) Department of Health, Education, and Welfare, was primarily designed to estimate labor supply responses, the find. 'Michael T. Hannan, Nancy B. Tuma, and Lyle P. Groeneveld, "Income and Marital Events: Evidence from an Income Maintenance Experiment," American Journal of Sociology, 82 (1977), 1186-1211. Subsequent articles in this journal by these authors reinforced their original findings. See "Income and Independence Effects on Marital Dissolution: Results from the Seattle and Denver Income-Maintenance Experiments," 84 (1978), 611-633; and "Dynamic Analysis of Event Histories," 84 (1979), 820-854. ings on marital breakups have had the biggest impact.' Groeneveld, Hannan, and Tuma subsequently wrote over twenty research papers and articles on this subject, and their research was presented as testimony before Congress during debates on legislative proposals to reform the welfare system. In their final report, published in 1983, they claimed that "the negative income tax (NIT) plans tested in SIME/DIME dramatically increased the rate at which marriages dissolved among white and black couples," and reported that the rate of marital dissolution increased by "40 to 60 p e r ~ e n t . " ~ These findings were and continue to be an important source of opposition to such reforms in the welfare system as the negative income tax, which provides a guaranteed income and cash transfer payments to low-income married-couple families. Gilbert Steiner, who reviewed the testimony in congressional hearings on welfare reform, wrote that "the SeattleDenver evidence has yfmuaded key politicians that a guaranteed-income plan at levels the leaders of the country think it can afford is incompatible with maximizing family stability in the affected p ~ ~ u l a t i o n . " ~ In this paper we present 'Citations to descriptions of all four experiments and a review of the labor supply responses in the experiments are provided by Robert A. Moffitt and Kenneth C. Kehrer, "The Effect of Tax and Transfer Programs on Labor Supply: The Evidence from the Income Maintenance Experiments," in Research in Labor Economics, Vol. 4, ed. R. G. Ehrenberg (Greenwic.h, Conrl.: JAI Press, 1981.), pp. 103-150. 3Lyle P. Groeneveld, Michael T. Hannan, and Nancy R. Tuma, "Marital Stability," in Final Report of the Seattle-Denver Income .'Maintenance Experiment, Volume 1 , Design and Results (Menlo Park, Cal.: SKI International, May 1983), pp. 257 and 383. Hereafter, Volume 1 will be cited as Final Report. 4Gilbert Y. Steiner, The Futility of Family Polzcy (Washington, D.C.: The Brookings Institution, 1981), p. 110. The research findings of Groeneveld, Hannan, and Tuma were also cited by Martin Anderson, John Bishop, George evidence to challenge the empirical findings of Groeneveld, Hannan, and Tuma. The research of Groeneveld, Hannan, and Tuma was also influential theoretically, because it appeared to show serious inadequacies in the previous interpretation, particularly by economists, of the causal linkage between income maintenance programs and marital instability among low-income families. SIME-DIME appeared to show that "ungenerous" (or low-benefit) NIT plans caused increased marital breakups relative to AFDC even though (a) these NIT plans provided less income to a mother whose husband left her than did AFDC, and (b) NIT plans provide benefits to the husband-wife couple if they stay together, whereas AFDC does not. Also adding to the puzzle was their finding that the generous, high-payment NIT plans, which offered more income to a mother whose husband left her than did AFDC, had no destabilizing effect on marriages. The theoretical framework used by Groeneveld, Hannan, and Tuma has two economic arguments. They hypothesize that welfare plans like AFDC or NIT (a) increase the economic independence of wives with children, which tends to destabilize marriages, and (b) provide income payments to intact husband-wife families, which tend to stabilize marriages. Groeneveld, Hannan, and Tuma depart from the conventional economic framework by hypothesizing that a stigma is attached to AFDC payments relative to NIT payments and that the "independence" and "income" effects of an NIT plan have complicated interactions for