LOSERS AND WINNERS: THE FINANCIAL CONSEQUENCES OF SEPARATION AND DIVORCE FOR MEN PATRICIA A.MCMANUS THOMAS A.DIPRETE Indiana University Duke University Contrary to conventional thinking,the majority of partnered men in the United States lose economic status when their unions dissolve.Using data from the Panel Study of Income Dynamics,this analysis shows that for most men the primary source of economic decline after union dissolution is their inability to fully compensate for the loss of their partner's income.A secondary source of economic decline is an increase in compulsory and voluntary support payments.Welfare state tax and trans- fer mechanisms have a much smaller overall impact on changes in men's living standards following separation.Although most men experience a decline in living standards following union dissolution,men's outcomes are heterogeneous,and the minority of men who relied on their partners for less than one-fifth of pre-dissolu- tion income typically gain from separation and divorce.The data show a clear trend toward greater economic interdependence in American partnerships,and this trend appears to increase the proportion of men who suffer a reduced standard of living following separation. LARGE body of research has estab- and its implications for the well-being of lished that marital disruption has a children who experience a parental divorce substantial negative impact on women's (Duncan and Brooks-Gunn 1997;McLana- standard of living,and that this impact is han and Sandefur 1994)contribute to the worse for women than for men (Bianchi, heated public debate on divorce reform. McArthur,and Hill 1989;Burkhauser et al. Feminist scholars and commentators who 1990,1991;Duncan and Hoffman 1985; defend no-fault divorce laws nonetheless Hoffman 1977;Smock 1993,1994;Smock, condemn this gender disparity as "uncon- Manning and Gupta 1999;for a review,see scionable for a legal system and a society Holden and Smock 1991).Gender inequal- committed to fairness,justice and equality" ity in the economic consequences of divorce (Weitzman 1996:538,emphasis in original; also see Bradford 1997;DiTullio 1997; Faludi 1991). Direct correspondence to Patricia McManus, Department of Sociology,Indiana University, Though Weitzman's (1985)well-publi- Ballantine Hall 744,1020 E.Kirkwood Avenue, cized claim that women lose three-quarters Bloomington,IN 47405-7103 (pmcmanus@ of their previous standard of living while indiana.edu).This research was supported in part men gain over 40 percent has been proven by National Science Foundation grant NSF-SBR- erroneous(Peterson 1996),the assertion that 96-31944.Karen Segar provided assistance with men gain from divorce is still part of the con- data preparation.We thank Frances Gold- ventional wisdom about marital dissolution scheider,participants in the Political and Eco- (e.g.DiTullio 1997;Morrison and Ritualo nomic Sociology Workshop at Indiana Univer- 2000;Smock et al.1999).Yet while there is sity,and the ASR Editors and anonymous review- ers for providing helpful comments on early overwhelming evidence supporting the view drafts.The data used in this study were made that women's standard of living declines- available to us by the Cross-National Equivalent often precipitously-following separation or File (CNEF)project at the College of Human divorce,the financial impact on men is less Ecology at Cornell University,Ithaca,NY. well understood.Studies relying primarily on 246 AMERICAN SOCIOLOGICAL REVIEW,2001.VoL.66(APRIL:246-268) This content downloaded from 129.96.252.188 on Mon,15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
LOSERS AND WINNERS: THE FINANCIAL CONSEQUENCES OF SEPARATION AND DIVORCE FOR MEN PATRICIA A. MCMANUS THOMAS A. DIPRETE Indiana University Duke University Contrary to conventional thinking, the majority of partnered men in the United States lose economic status when their unions dissolve. Using data from the Panel Study of Income Dynamics, this analysis shows that for most men the primary source of economic decline after union dissolution is their inability to fully compensate for the loss of their partner's income. A secondary source of economic decline is an increase in compulsory and voluntary support payments. Welfare state tax and transfer mechanisms have a much smaller overall impact on changes in men's living standards following separation. Although most men experience a decline in living standards following union dissolution, men's outcomes are heterogeneous, and the minority of men who relied on their partners for less than one-fifth of pre-dissolution income typically gain from separation and divorce. The data show a clear trend toward greater economic interdependence in American partnerships, and this trend appears to increase the proportion of men who suffer a reduced standard of living following separation. A LARGE body of research has established that marital disruption has a substantial negative impact on women's standard of living, and that this impact is worse for women than for men (Bianchi, McArthur, and Hill 1989; Burkhauser et al. 1990, 1991; Duncan and Hoffman 1985; Hoffman 1977; Smock 1993, 1994; Smock, Manning and Gupta 1999; for a review, see Holden and Smock 1991). Gender inequality in the economic consequences of divorce Direct correspondence to Patricia McManus, Department of Sociology, Indiana University, Ballantine Hall 744, 1020 E. Kirkwood Avenue, Bloomington, IN 47405-7103 (pmcmanus@ indiana.edu). This research was supported in part by National Science Foundation grant NSF-SBR- 96-31944. Karen Segar provided assistance with data preparation. We thank Frances Goldscheider, participants in the Political and Economic Sociology Workshop at Indiana University, and the ASR Editors and anonymous reviewers for providing helpful comments on early drafts. The data used in this study were made available to us by the Cross-National Equivalent File (CNEF) project at the College of Human Ecology at Cornell University, Ithaca, NY. and its implications for the well-being of children who experience a parental divorce (Duncan and Brooks-Gunn 1997; McLanahan and Sandefur 1994) contribute to the heated public debate on divorce reform. Feminist scholars and commentators who defend no-fault divorce laws nonetheless condemn this gender disparity as "unconscionable for a legal system and a society committed to fairness, justice and equality" (Weitzman 1996:538, emphasis in original; also see Bradford 1997; DiTullio 1997; Faludi 1991). Though Weitzman's (1985) well-publicized claim that women lose three-quarters of their previous standard of living while men gain over 40 percent has been proven erroneous (Peterson 1996), the assertion that men gain from divorce is still part of the conventional wisdom about marital dissolution (e.g. DiTullio 1997; Morrison and Ritualo 2000; Smock et al. 1999). Yet while there is overwhelming evidence supporting the view that women's standard of living declinesoften precipitously-following separation or divorce, the financial impact on men is less well understood. Studies relving primarily on 246 AMERICAN SOCIOLOGICAL REVIEW, 2001, VOL. 66 (APRIL:246-268) This content downloaded from 129.96.252.188 on Mon, 15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
MALE LOSERS AND WINNERS AFTER SEPARATION/DIVORCE 247 data from the 1970s"decade of divorce"gen- An understanding of the financial after- erally found that men lost a small share of math of union dissolution for men requires their nominal income and enjoyed an in- knowledge of the mechanisms that contrib- crease in their standard of living following ute to change in economic status.Men's out- divorce or separation(Duncan and Hoffman comes depend not only on changes in house- 1985;Hoffman 1977;Sorensen 1992).But hold composition and the loss of partner's findings from more recent evidence are less income,but also on the extent to which state uniform.Two studies report post-disruption welfare policy-via tax and transfer pro- declines in men's living standards,or mixed grams-targets family formation and disso- results depending on the equivalence scale lution.Men's outcomes also depend on the used to calculate changes in standard of liv- characteristics of American family law (in- ing (Burkhauser et al.1990,1991).Another cluding enforcement practices)that affect study finds substantial gains in economic the size of mandatory support payments to well-being among separating men,but limits the ex-partner or to children.Finally,men's the scope of the study to noncustodial fathers outcomes are affected by“voluntary”cash (Bianchi,Subaiya and Kahn 1999).Two re- payments to their former partners (Edin and cent studies also find gains in per capita in- Lein 1997).Despite the potential importance come for separating men,but these studies of these mechanisms in determining men's limit the scope to men who were married to economic status following separation,only young women and do not report changes in a handful of studies address the impact of living standards using conventional equiva- the welfare state (Burkhauser et al.1990, lence scales (Smock 1993,1994). 1991),and a few more take compulsory fam- We believe it is time to reassess the finan- ily-support payments into account(Bianchi cial consequences of union dissolution for et al.1999;Burkhauser et al.1990,1991; men.The disparity between early research Duncan and Hoffman 1985;Smock 1994). and some recent research may indicate a These studies,however,are hampered either temporal shift in the financial consequences by inadequate coverage of the population of of union dissolution for men.More impor- separating men (e.g.,by restricting the study tant,recent research suggests a great deal of to noncustodial fathers),or by the use of in- heterogeneity in these consequences.Recent sufficiently broad measures of post-separa- demographic trends,including the decline in tion transfers. marital fertility,the rise in cohabitation,and The analyses presented here provide the the increase in stepfamily households,all first comprehensive and systematic assess- contribute to an increasing diversity of ment of the impact of market and nonmarket couple-headed households.The rise in mechanisms on men's financial outcomes women's labor force participation and the following union dissolution.They are de- decline in the gender gap in earnings signed to answer the following questions:(1) (Bernhardt,Morris,and Handcock 1995) Do men typically experience an increase or have increased women's economic contribu- a decrease in their standard of living follow- tions to the household,especially during the ing separation?(2)Is the economic impact 1980s (Hayghe 1993).More than ever be- of separation fairly uniform across men,or fore,men face the risk that separation will is there substantial heterogeneity in their impose a substantial financial burden in the outcomes?(3)How is the economic impact form of the loss of partner's income.At the of separation on men structured by market, same time,fathers may be assuming a larger welfare state,legal,and voluntaristic level- role in the post-disruption financial support ing mechanisms? of their children than was the case during the 1970s.National efforts to reform and en- LEVELING MECHANISMS force child-support decrees,along with a gradual increase in the rate of paternal and Men are often seen as economic winners in joint physical custody of children(Garfinkel a separation scenario in which material re- et al.1998),suggest that fathers may be less sources and financial obligations are un- likely to realize substantial material gains equally distributed between the two ex-part- from separation and divorce. ners.But any potential economic windfall This content downloaded from 129.96.252.188 on Mon,15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
MALE LOSERS AND WINNERS AFTER SEPARATION/DIVORCE 247 data from the 1970s "decade of divorce" generally found that men lost a small share of their nominal income and enjoyed an increase in their standard of living following divorce or separation (Duncan and Hoffman 1985; Hoffman 1977; Sorensen 1992). But findings from more recent evidence are less uniform. Two studies report post-disruption declines in men's living standards, or mixed results depending on the equivalence scale used to calculate changes in standard of living (Burkhauser et al. 1990, 1991). Another study finds substantial gains in economic well-being among separating men, but limits the scope of the study to noncustodial fathers (Bianchi, Subaiya and Kahn 1999). Two recent studies also find gains in per capita income for separating men, but these studies limit the scope to men who were married to young women and do not report changes in living standards using conventional equivalence scales (Smock 1993, 1994). We believe it is time to reassess the financial consequences of union dissolution for men. The disparity between early research and some recent research may indicate a temporal shift in the financial consequences of union dissolution for men. More important, recent research suggests a great deal of heterogeneity in these consequences. Recent demographic trends, including the decline in marital fertility, the rise in cohabitation, and the increase in stepfamily households, all contribute to an increasing diversity of couple-headed households. The rise in women's labor force participation and the decline in the gender gap in earnings (Bernhardt, Morris, and Handcock 1995) have increased women's economic contributions to the household, especially during the 1980s (Hayghe 1993). More than ever before, men face the risk that separation will impose a substantial financial burden in the form of the loss of partner's income. At the same time, fathers may be assuming a larger role in the post-disruption financial support of their children than was the case during the 1970s. National efforts to reform and enforce child-support decrees, along with a gradual increase in the rate of paternal and joint physical custody of children (Garfinkel et al. 1998), suggest that fathers may be less likely to realize substantial material gains from separation and divorce. An understanding of the financial aftermath of union dissolution for men requires knowledge of the mechanisms that contribute to change in economic status. Men's outcomes depend not only on changes in household composition and the loss of partner's income, but also on the extent to which state welfare policy-via tax and transfer programs-targets family formation and dissolution. Men's outcomes also depend on the characteristics of American family law (including enforcement practices) that affect the size of mandatory support payments to the ex-partner or to children. Finally, men's outcomes are affected by "voluntary" cash payments to their former partners (Edin and Lein 1997). Despite the potential importance of these mechanisms in determining men's economic status following separation, only a handful of studies address the impact of the welfare state (Burkhauser et al. 1990, 1991), and a few more take compulsory family-support payments into account (Bianchi et al. 1999; Burkhauser et al. 1990, 1991; Duncan and Hoffman 1985; Smock 1994). These studies, however, are hampered either by inadequate coverage of the population of separating men (e.g., by restricting the study to noncustodial fathers), or by the use of insufficiently broad measures of post-separation transfers. The analyses presented here provide the first comprehensive and systematic assessment of the impact of market and nonmarket mechanisms on men's financial outcomes following union dissolution. They are designed to answer the following questions: (1) Do men typically experience an increase or a decrease in their standard of living following separation? (2) Is the economic impact of separation fairly uniform across men, or is there substantial heterogeneity in their outcomes? (3) How is the economic impact of separation on men structured by market, welfare state, legal, and voluntaristic leveling mechanisms? LEVELING MECHANISMS Men are often seen as economic winners in a separation scenario in which material resources and financial obligations are unequally distributed between the two ex-partners. But any potential economic windfall This content downloaded from 129.96.252.188 on Mon, 15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
248 AMERICAN SOCIOLOGICAL REVIEW from separation is likely to be leveled by separation income is split across two new mechanisms that have gained strength in re- households,the loss of economies of scale cent decades.Shifts in the labor market have means that the joint economic status of the reduced married and cohabiting men's share separated partners is lower than it would be of household income,and thereby raised the if they shared the costs of maintaining a cost of exit.Men's potential gains are often single household.We expect that the loss of further reduced by tax and social welfare partner's market income is the primary policies,by compulsory child support and mechanism for reducing men's economic alimony payments,and by informal support status following union dissolution. to members of the former family. WELFARE STATE MECHANISMS GENDER-ROLE SPECIALIZATION Welfare state tax and transfer policy takes AND MARKET MECHANISMS marital status and household composition The presumption that men reap financial re- into account,ostensibly providing preferen- wards in the aftermath of shedding their tial treatment to families.In fact,the well- families is partly rooted in the "specializa- publicized "marriage penalty"imposed on tion and trading model"(Becker 1981; many dual-earner couples obscures the fact Oppenheimer 1997),which arguably is the that,at least through the mid-1990s,the ma- most prominent contemporary model for jority of married couples paid lower federal marriage.This model conceptualizes the income taxes than they would have if they married couple as a production and bargain- were single (U.S.Congressional Budget Of- ing unit in which,for biological reasons,the fice 1997).As another example,low-income female partner has a comparative advantage men who separate from their families stand in household production,while the male part- to lose benefits (e.g.,food stamps)that are ner has a comparative advantage in paid la- contingent on both household income and bor.According to this model,the partners household size.We expect state policy to specialize in gender-specific tasks and then buffer income lost to men with low pre-sepa- trade the product of their labor to maximize ration shares of household income,and to their joint well-being.It seems obvious that level the incomes of men with the highest a male breadwinner who kept his job while pre-separation income shares. separating from his homemaker partner would be materially better off,though at the JUDICIAL MECHANISMS TO cost of doing more housework(Gupta 1999). PROTECT DEPENDENTS Because of the specialization-induced finan- cial risk,the wife demands a marriage con- Post-separation household income can also tract to protect herself from possible malfea- be reduced by court-ordered family-support sance by her partner. payments.Noncustodial parents are subject As Oppenheimer (1997)points out,com- to compulsory child-support payments,and plete gender-role specialization is a high-risk in rare instances the financially weaker part- strategy for both partners and is not typical ner may also win redress in the form of a of contemporary American households.In- spousal support award.Noncustodial parents stead,American men and women generally are overwhelmingly male,and alimony pay- share in the market provision for the house- ments flow almost exclusively from men to hold.Men typically earn more than their fe- women.We expect compulsory transfers to male partners,and the partner with greater reduce men's incomes,more so for men who economic strength is likely to have a com- provided the bulk of the couple's pre-disrup- parative financial advantage following any tion income. separation.But relative market advantages are imprecise tools for assessing the abso- INFORMAL NORMS OF OBLIGATION lute financial impact of an event as costly as AND RECIPROCITY separation,which can easily produce a de- cline in the living standards of both former Separation and divorce do not invariably partners (Sgrensen 1994).When the pre- sever informal financial ties between men This content downloaded from 129.96.252.188 on Mon,15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
248 AMERICAN SOCIOLOGICAL REVIEW from separation is likely to be leveled by mechanisms that have gained strength in recent decades. Shifts in the labor market have reduced married and cohabiting men's share of household income, and thereby raised the cost of exit. Men's potential gains are often further reduced by tax and social welfare policies, by compulsory child support and alimony payments, and by informal support to members of the former family. GENDER-ROLE SPECIALIZATION AND MARKET MECHANISMS The presumption that men reap financial rewards in the aftermath of shedding their families is partly rooted in the "specialization and trading model" (Becker 1981; Oppenheimer 1997), which arguably is the most prominent contemporary model for marriage. This model conceptualizes the married couple as a production and bargaining unit in which, for biological reasons, the female partner has a comparative advantage in household production, while the male partner has a comparative advantage in paid labor. According to this model, the partners specialize in gender-specific tasks and then trade the product of their labor to maximize their joint well-being. It seems obvious that a male breadwinner who kept his job while separating from his homemaker partner would be materially better off, though at the cost of doing more housework (Gupta 1999). Because of the specialization-induced financial risk, the wife demands a marriage contract to protect herself from possible malfeasance by her partner. As Oppenheimer (1997) points out, complete gender-role specialization is a high-risk strategy for both partners and is not typical of contemporary American households. Instead, American men and women generally share in the market provision for the household. Men typically earn more than their female partners, and the partner with greater economic strength is likely to have a comparative financial advantage following any separation. But relative market advantages are imprecise tools for assessing the absolute financial impact of an event as costly as separation, which can easily produce a decline in the living standards of both former partners (Sorensen 1994). When the preseparation income is split across two new households, the loss of economies of scale means that the joint economic status of the separated partners is lower than it would be if they shared the costs of maintaining a single household. We expect that the loss of partner's market income is the primary mechanism for reducing men's economic status following union dissolution. WELFARE STATE MECHANISMS Welfare state tax and transfer policy takes marital status and household composition into account, ostensibly providing preferential treatment to families. In fact, the wellpublicized "marriage penalty" imposed on many dual-earner couples obscures the fact that, at least through the mid-1990s, the majority of married couples paid lower federal income taxes than they would have if they were single (U.S. Congressional Budget Office 1997). As another example, low-income men who separate from their families stand to lose benefits (e.g., food stamps) that are contingent on both household income and household size. We expect state policy to buffer income lost to men with low pre-separation shares of household income, and to level the incomes of men with the highest pre-separation income shares. JUDICIAL MECHANISMS TO PROTECT DEPENDENTS Post-separation household income can also be reduced by court-ordered family-support payments. Noncustodial parents are subject to compulsory child-support payments, and in rare instances the financially weaker partner may also win redress in the form of a spousal support award. Noncustodial parents are overwhelmingly male, and alimony payments flow almost exclusively from men to women. We expect compulsory transfers to reduce men's incomes, more so for men who provided the bulk of the couple's pre-disruption income. INFORMAL NORMS OF OBLIGATION AND RECIPROCITY Separation and divorce do not invariably sever informal financial ties between men This content downloaded from 129.96.252.188 on Mon, 15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
MALE LOSERS AND WINNERS AFTER SEPARATION/DIVORCE 249 and their children and former partners. hold consumption patterns give similar re- Sometimes informal transfers are an expres- sults for our question.Some of our tables re- sion of personal responsibility for kinship port results using alternative living standard obligations that are perceived to outlast the measures;other tables rely on a single mea- partnership.Such transfers can also be seen sure that produces relatively optimistic esti- as an indirect response to the specific char- mates of the economic consequences of acter of the American welfare state.High-in- separation for men. come fathers may supplement compulsory We assess the economic consequences of support payments or provide additional di- separation in several ways.We use regres- rect support to children who have moved out sion models and other statistical strategies to of the mother's household.Ethnographic estimate the actual economic impact of sepa- studies of low-income fathers suggest that ration.Next,we use the full sample of mar- many who are not paying compulsory sup- ried and cohabiting men to simulate the po- port may help out informally,perhaps as part tential economic impact of separation for of a tacit economic bargain (Edin and Lein men with different household arrangements. 1997;Sullivan 1989).Edin and Lein (1997) The median values from the simulations are found,for example,that AFDC mothers in then compared to the actual outcomes among their sample received more money from the men who separated in order to better under- fathers through informal (covert)support stand the typical pattern of men's behavioral payments than through the formal child-sup- responses to separation.We also assess the port system.Whatever their explanation, relative impact of market and nonmarket noncompulsory money transfers to former mechanisms by measuring economic change partners and children who no longer live in both before and after taking account of taxes, the household will reduce any financial public transfers,compulsory support pay- gains from divorce or separation. ments,and other private transfers. ANALYTICAL STRATEGY DATA AND METHODS Following previous research,we measure Data for this paper are from the 1980-1993 economic growth among men who at time t waves of the Panel Study of Income Dynam- were married or in a long-term cohabiting ics (PSID)and the income measures are union by comparing annual household in- from the Cross-National Equivalent File come or standard of living (income adjusted (Wagner,Burkhauser,and Behringer 1993).1 for household size)in the previous calendar The PSID collects data from sample mem- year (t-1)to annual household income two bers on an annual basis,typically in the years later (t+1).We define the short-term spring of the survey year.Respondents are economic impact of union dissolution as the asked about marital status and household difference at time t +1 between economic composition at the time of the survey as well change for men who separated from their as changes from the previous survey.Exten- partners and men who remained in intact sive information is collected on individual unions.While our primary focus is on the short-term consequences,we also present 1 The income measures on the Cross-National selected results for a four-year economic Equivalent File(CNEF)are constructed by aug- change-between one year before and three menting the detailed income components in the years after the last intact interview (i.e.,be- PSID data with two components not available in tween t-1 and t+3).We report results the source data,namely the imputed rental value separately for white men and African Ameri- of owner-occupied housing and imputed payroll can men. and income taxes,calculated using the National Our analysis includes results for change in Bureau of Economic Research's TAXSIM rou- men's nominal income,but our chief inter- tine.We use the CNEF pre-government and post- est is change in men's living standards. government income measures in our analyses, along with additional measures of post-support While no single measure of living standards income constructed by combining the CNEF has won universal acceptance,measures measures with information on support payments based on plausible assumptions about house- from the source data. This content downloaded from 129.96.252.188 on Mon,15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
MALE LOSERS AND WINNERS AFTER SEPARATION/DIVORCE 249 and their children and former partners. Sometimes informal transfers are an expression of personal responsibility for kinship obligations that are perceived to outlast the partnership. Such transfers can also be seen as an indirect response to the specific character of the American welfare state. High-income fathers may supplement compulsory support payments or provide additional direct support to children who have moved out of the mother's household. Ethnographic studies of low-income fathers suggest that many who are not paying compulsory support may help out informally, perhaps as part of a tacit economic bargain (Edin and Lein 1997; Sullivan 1989). Edin and Lein (1997) found, for example, that AFDC mothers in their sample received more money from the fathers through informal (covert) support payments than through the formal child-support system. Whatever their explanation, noncompulsory money transfers to former partners and children who no longer live in the household will reduce any financial gains from divorce or separation. ANALYTICAL STRATEGY Following previous research, we measure economic growth among men who at time t were married or in a long-term cohabiting union by comparing annual household income or standard of living (income adjusted for household size) in the previous calendar year (t - 1) to annual household income two years later (t + 1). We define the short-term economic impact of union dissolution as the difference at time t + 1 between economic change for men who separated from their partners and men who remained in intact unions. While our primary focus is on the short-term consequences, we also present selected results for a four-year economic change-between one year before and three years after the last intact interview (i.e., between t - 1 and t + 3). We report results separately for white men and African American men. Our analysis includes results for change in men's nominal income, but our chief interest is change in men's living standards. While no single measure of living standards has won universal acceptance, measures based on plausible assumptions about household consumption patterns give similar results for our question. Some of our tables report results using alternative living standard measures; other tables rely on a single measure that produces relatively optimistic estimates of the economic consequences of separation for men. We assess the economic consequences of separation in several ways. We use regression models and other statistical strategies to estimate the actual economic impact of separation. Next, we use the full sample of married and cohabiting men to simulate the potential economic impact of separation for men with different household arrangements. The median values from the simulations are then compared to the actual outcomes among men who separated in order to better understand the typical pattern of men's behavioral responses to separation. We also assess the relative impact of market and nonmarket mechanisms by measuring economic change both before and after taking account of taxes, public transfers, compulsory support payments, and other private transfers. DATA AND METHODS Data for this paper are from the 1980-1993 waves of the Panel Study of Income Dynamics (PSID) and the income measures are from the Cross-National Equivalent File (Wagner, Burkhauser, and Behringer 1993).1 The PSID collects data from sample members on an annual basis, typically in the spring of the survey year. Respondents are asked about marital status and household composition at the time of the survey as well as changes from the previous survey. Extensive information is collected on individual 1 The income measures on the Cross-National Equivalent File (CNEF) are constructed by augmenting the detailed income components in the PSID data with two components not available in the source data, namely the imputed rental value of owner-occupied housing and imputed payroll and income taxes, calculated using the National Bureau of Economic Research's TAXSIM routine. We use the CNEF pre-government and postgovernment income measures in our analyses, along with additional measures of post-support income constructed by combining the CNEF measures with information on support payments from the source data. This content downloaded from 129.96.252.188 on Mon, 15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
250 AMERICAN SOCIOLOGICAL REVIEW and household income flows during the pre- ables that affect household economic status, vious calendar year.Our sample includes so that observations on men between the ages of 18 and 65 who identified themselves as either In yit =xiB+Ei (1) white or black (we excluded others because If we difference this model at two points sample sizes are too small)and who were in time,and if we collect all changes related partners in couple-headed households.Mar- to union dissolution and repartnering into ried men were included if they lived with dummy variables D and R,we obtain their spouse at time t.Men who were cohab- iting with a female partner were included if In(y+)-ln(y-)=△lny the same partner was in the household at =DYD+RYR+△xB+V, (2) time t-1 and time t.2 The analysis sample excludes unions that were dissolved through where yi-is a measure of economic status death or institutionalization between times t at time t-1,Ax;is the subset of control vari- and t+1. ables that change in value between t-1 and To be included in the sample,men who t+1,D is an indicator for union disruption were observed in unions at time t had to be between t-1 and t+1,and R is an indica- followed for the next two interviews,at time tor for a new partner between t-1 and t +1. t+1,by which time a disruption may or (To achieve greater clarity,we have sup- may not have occurred,and at time t+2 pressed explicit subscripts for calendar time when income data is collected retrospec- on the right side of equation 2.)Equation 2 tively for the previous calendar year.Sepa- assumes that change in economic status does ration is associated with a high risk of not depend on the level of economic status sample attrition among men in panel data at time t-1.Because this assumption may (D.Hill 1997;M.Hill 1992;Fitzgerald, not be correct,we estimated a slightly more Gottschalk,and Moffitt 1998),and this risk complex specification that relaxes this as- is particularly high for men with unstable sumption: work histories (Fitzgerald et al.1998).Our sample is therefore likely to underrepresent ln(+i)-ln(-i)=△lny men with the worst economic outcomes fol- lowing separation,and to overrepresent men =In(yi-Iy +DiYD+RYR who pay child support.However,a recent +△xB+V 3) study of attrition bias in the PSID found that the appropriate use of sample weights in re- Because pre-disruption income might be gression analyses produces consistent esti- correlated with other factors affecting in- mates despite the high levels of attrition come change,we estimated this equation (Fitzgerald et al.1998).To adjust for these using standard instrumental variables tech- unequal probabilities of sample selection niques (Greene 2000).One set of valid in- and attrition,we assigned each respondent a struments for equation 3 includes the con- single longitudinal weight equal to the cross- trol variables in equation 1,whose values sectional weight attached to that respondent were constant over time and thus do not ap- in the final year in which the respondent pear in the difference equation because they contributed to the data. cancel out of the right-hand side (e.g.,ques- Our regression model for the average im- tions about men's education are generally pact of union dissolution can be derived as asked only at the initial interview in the follows.Let yi:equal household economic PSID,so Ax =0 for schooling).An alterna- status at time t,and let xi be a vector of vari- tive strategy is to find instruments that do not appear at all in equation 1,such as in- 2 We include only cohabitants in long-term come at time t-2.We applied these two unions because the PSID prorates the income of strategies in turn to estimate instrumental the cohabiting partner to reflect the actual num- variables models,using education and ber of months spent in the same household,but twice-lagged income as the respective in- does not prorate the income of new marital part struments.The results from these two esti- ners. mation strategies were similar,and we re- This content downloaded from 129.96.252.188 on Mon,15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions
250 AMERICAN SOCIOLOGICAL REVIEW and household income flows during the previous calendar year. Our sample includes observations on men between the ages of 18 and 65 who identified themselves as either white or black (we excluded others because sample sizes are too small) and who were partners in couple-headed households. Married men were included if they lived with their spouse at time t. Men who were cohabiting with a female partner were included if the same partner was in the household at time t - 1 and time t.2 The analysis sample excludes unions that were dissolved through death or institutionalization between times t and t + 1. To be included in the sample, men who were observed in unions at time t had to be followed for the next two interviews, at time t + 1, by which time a disruption may or may not have occurred, and at time t + 2 when income data is collected retrospectively for the previous calendar year. Separation is associated with a high risk of sample attrition among men in panel data (D. Hill 1997; M. Hill 1992; Fitzgerald, Gottschalk, and Moffitt 1998), and this risk is particularly high for men with unstable work histories (Fitzgerald et al. 1998). Our sample is therefore likely to underrepresent men with the worst economic outcomes following separation, and to overrepresent men who pay child support. However, a recent study of attrition bias in the PSID found that the appropriate use of sample weights in regression analyses produces consistent estimates despite the high levels of attrition (Fitzgerald et al. 1998). To adjust for these unequal probabilities of sample selection and attrition, we assigned each respondent a single longitudinal weight equal to the crosssectional weight attached to that respondent in the final year in which the respondent contributed to the data. Our regression model for the average impact of union dissolution can be derived as follows. Let Yit equal household economic status at time t, and let xit be a vector of vari- 2 We include only cohabitants in long-term unions because the PSID prorates the income of the cohabiting partner to reflect the actual number of months spent in the same household, but does not prorate the income of new marital partners. ables that affect household economic status, so that lnyit x+it- 1 If we difference this model at two points in time, and if we collect all changes related to union dissolution and repartnering into dummy variables D and R, we obtain ln(Yit+i) - ln(Yi,t-i) = A In yi Di7D =D~yD+R~y+ + Ri7R Axf + vi, (2) +AiD i where yit-l is a measure of economic status at time t - 1, Axi is the subset of control variables that change in value between t - 1 and t + 1, D is an indicator for union disruption between t - 1 and t + 1, and R is an indicator for a new partner between t - 1 and t + 1. (To achieve greater clarity, we have suppressed explicit subscripts for calendar time on the right side of equation 2.) Equation 2 assumes that change in economic status does not depend on the level of economic status at time t - 1. Because this assumption may not be correct, we estimated a slightly more complex specification that relaxes this assumption: ln(Yit+i) - ln(Yi,t-1) = A In yj = ln(Yit-j ry + DiYD + RiYR + Axf'+vi. (3) Because pre-disruption income might be correlated with other factors affecting income change, we estimated this equation using standard instrumental variables techniques (Greene 2000). One set of valid instruments for equation 3 includes the control variables in equation 1, whose values were constant over time and thus do not appear in the difference equation because they cancel out of the right-hand side (e.g., questions about men's education are generally asked only at the initial interview in the PSID, so Ax = 0 for schooling). An alternative strategy is to find instruments that do not appear at all in equation 1, such as income at time t - 2. We applied these two strategies in turn to estimate instrumental variables models, using education and twice-lagged income as the respective instruments. The results from these two estimation strategies were similar, and we reThis content downloaded from 129.96.252.188 on Mon, 15 Feb 2016 15:26:54 UTC All use subject to JSTOR Terms and Conditions