Who Gets the Daddy Income Bonus? 127

WHO GETS THE DADDY INCOME BONUS AND WHO PAYS THE COST?*

Richard Hogan

Purdue University

Carolyn Cummings Perrucci

Purdue University

ABSTRACT

The literature suggests that: (1) married persons earn more, (2) this is particularly true for white men, and (3) white men with children earn even more, which is less true for black men and not true for women. We use 2012 March Supplement Current Population Survey data to examine these assertions and find consistent support for the first assertion but little for the second. We find that men, black and white, claim a marriage bonus. Similarly, both black and white men, whether single or married, claim the Daddy Bonus, while women, black and white, single or married, suffer the Motherhood penalty. Here, however, it seems that white men are particularly advantaged. We deconstruct and evaluate the theoretical and methodological significance of these findings in conclusion.

Sociologists, beginning with Budig and England (2001), have specified the wage cost of motherhood, attempted to partition the effects into discrimination versus domestic duties that interfere with labor force participation, traced changes over time (Avellar and Smock 2003) and through the life course (Hogan and Perrucci 1998; Hogan et al. 2000).

More recently, Glauber (2008) and Hodges and Budig (2010) have reoriented the literature toward consideration of the Daddy wage bonus, the extent to which it is not equally available to all men who father children, and how the masculine advantage in domesticity has changed over time (Cohen 2002). Meanwhile, even the mass media has proclaimed that marriage if not parenting is actually a good thing, even for women, financially as well as in almost every other way (Waite and Gallagher 2000) and, most recently, that the decline in marriage rates, particularly for black women, is a problem for everyone, even for white people (Banks 2012).

While not everyone will agree on these assertions, there seems to be adequate evidence to conclude that we now know three things: (1) marriage tends to be associated with higher annual earnings, (2) this is particularly true for white men, (3) having children tends to be associated with increased earnings for white men, but not so much for black men and even less for women.

Why? First, for some men, especially white upper-middle class men, marriage brings not just an unpaid, live-in housekeeper and life-partner but also a set of kin-based advantages (capital and job opportunities) at relatively little cost (since women do most of the work in maintaining kin relations: organizing weddings and funerals, sending out cards for birthdays and holidays, caring for the young and old, etc.—see Bianchi et al. 2000; Bittman et al. 2003; DiMaggio and Garip 2011; DiPrete et al. 2011; Reardon and Bischoff 2011). This seems to work less well for black men and seems to be a real burden for women (Glauber 2008). Women, particularly white women, do get the benefits but they have to provide the unpaid labor to keep kinship networks in shape (Tilly 1998; Hogan 2001; Hogan 2013).

Second, whites and blacks now work together (Farley 1984; Tomaskovic-Devey 1993; Tomaskovic-Devey et al. 2006), and some have had children together, dating back to slavery, but white men rarely marry or accept responsibility for caring for their black sex partners or the offspring of these unions (see DuBois 1998; Hunter 1997). Thus black women and children are excluded from the benefits of patronage from wealthy whites. Blacks can marry each other, but the kin-based benefits of doing so are severely limited, because whites hoard employment opportunities and capital, which is part of the reason for the enduring racial inequality in wealth (Oliver and Shapiro 1987; Tilly 1998; Hogan 2001).

Third, black women who work for pay outside the home, who are doubly disadvantaged in the marriage and children/labor market (Hogan and Perrucci 2007a), tend to be under-represented among married people with kids. Blacks working for pay outside the home tend to be single persons with no kids, the modal condition of the black labor force in the U. S. in the year 2011 (Banks 2012).

Here we analyze 2012 March Supplement Current Population Survey (CPS) data to compare black and white men and women who are “earners” (report more than no earnings for 2011) and are not enrolled as high school or college students, in order to offer some empirical support for these three assertions. We choose to use this cross-sectional survey data rather than the panel data that is most prevalent in the literature because we want to look at a representative national sample of earners, including the self-employed, and compare their annual earnings as opposed to their weekly wages. We are less interested in the short-term effects of giving birth than in the overall effect of marriage and parenthood on income inequality. We also want to use the data that is routinely used to evaluate the race and gender gaps in earnings (Hogan and Perrucci 2007b), so that we can begin with average earnings differences that can be readily translated into earnings gaps.

Obviously, class, occupation, labor market, and “human capital” or labor status differences interpret or specify a substantial amount of the observed (gross) race and gender gaps in earnings. As we shall see, when predicting earnings, the explained effects of race, gender, marriage and family status are dwarfed by these other effects, which are usually the focus of attention but are treated here as controls. Some scholars have suggested that we should stop looking at racial and gender inequality, because there is more variance within than between race and gender categories (Leicht 2008). We recognize that these other variables might seem to overwhelm the effects of race and gender, but they do not entirely specify the complex ways in which race and gender are confounded with marital and family status in predicting earnings. Thus we control for these effects in order to focus our attention on the net effects of race, gender, marital and family status as these combine in complex ways to predict earnings in Ordinary Least Squares (OLS) regression models that estimate these effects. The reader particularly interested in our control variables is referred elsewhere. Hogan (2013), for example, uses CPS data from 2001 to estimate a model that includes virtually all of these control variables but does not deconstruct marriage and family status as we shall here.

Here we use a series of dummy variables and interaction terms to estimate these confounded effects, as they are reported in the literature and hypothesized or predicted here, but we have less confidence in these OLS models as definitive tests of these hypotheses because they conflate “explained” and “unexplained” effects that we are able to separate with the Blender Oaxaca decomposition. As we elaborate in the methods section, below, we use this decomposition to distinguish these effects in order to better understand how race and gender inequality differ across marital and family relations.

Specifically, when analyzing family and marital relations, we conceptualize durable inequality associated with gender, following Tilly (1998) and Hogan (2001), primarily as “exploitation,” because women are routinely included in relations from which they are not able to secure their share of the available surplus (notably relations of marriage and parenthood). We conceptualize durable inequality associated with race primarily as “opportunity hoarding,” because blacks are routinely denied access to relations in which the participants claim surplus value on the basis of their monopoly (or status-group closure), notably in marriage and family relations.

In these terms, we uncover significant effects of the exploitation of women as wives and mothers, indicated by the unexplained effects of marital and family status in the Blinder-Oaxaca models comparing the earnings of white and black women to white men. Somewhat unexpectedly, we also uncover find a significant unexplained effect of having few children, when comparing the earnings of black and white males. Although this effect is less substantial and robust than the comparable effect for black women, it does suggest that the exploitation of women as mothers is, perhaps, intensified for whites, which may or may not be relevant in considering why black women are reluctant to marry white men (Banks 2011).

At the same time, we uncover substantial opportunity hoarding in the extent to which men hoard employment opportunities and family status, denying mothers and married women access to their own employment earnings. In this regard, patriarchy and the welfare state might be seen as redistributive mechanisms that provide shares of the earnings from husbands or other employed persons. Here we shall ignore these forms of recompense and focus simply on the extent to which marriage is racially exclusionary while parenthood is gender inclusionary, indicating the racial opportunity hoarding and gender exploitation that Hogan (2001) describes in more detail.

WHO GETS THE DADDY INCOME BONUS AND WHO PAYS THE PENALTY

The literature suggests that white men benefit from including wives and excluding blacks in kinship relations that might provide access to jobs or capital (Tilly 1998; Hogan 2001; Hogan 2013). What some have called motherhood and marriage penalties should really be considered fatherhood and husband benefits, which tend to be greater for white men (Glauber 2008; Hodges and Budig 2010).

Black women appear to be particularly disadvantaged in this regard (Hogan and Perrucci 2007a), but Budig and England (2001) find no significant racial difference in the effect of number of children and some evidence that black and Latino women with three or more children actually suffer a lesser penalty than do white women. There does seem to be evidence, however, suggesting that black men do not receive the same benefits from fatherhood that white men claim (Hodges and Budig 2010). Killenwald (2013) finds that only fathers who are living with their biological children gain the “Fatherhood Premium” of higher wage rates.

At the same time, it appears that marriage is associated with higher earnings but greater parenting penalties for women (Budig and England 2001, pp. 214-217). For men, however, both marriage and parenting are associated with higher earnings, especially for white men (Budig and Hodges 2010, pp. 730-736). In fact, Glauber (2008) finds similar racial differences but also finds that the Daddy bonus is limited to married men. This suggests that we need to more closely examine how parenthood and marriage affect the earnings of black and white men and women. The burden of parenthood is inequitably distributed, but it appears that the benefits of marriage compound this problem.

Since marriage seems to be associated with higher earnings, net of its effect on the Motherhood penalty (Budig and England 2001; Waite and Gallagher 2000), and marriage rates are declining dramatically, particularly for black women (Banks 2012), we expect that black women will be doubly disadvantaged in the marriage and employment markets, but we are not clear on how this might affect the Motherhood penalty for black women or the fatherhood (Daddy) bonus for black men. We need to look at parenting penalties, marriage benefits, and racial and gender inequality, separately, cumulatively, and in the confounded manner in which they are actually experienced by working men and women, as gross effects, as net effects, in models controlling for education, occupation, class, and labor market factors, and as components of racial and gender earnings inequality, which we might deconstruct empirically.

That is our goal in this analysis. We shall begin with baseline data on race and gender differences in annual earnings, averaged for white and black men and women who are single or married, with no children, one or two children, or many children. We expect to find evidence of white, male, marriage, and parenting earnings advantages and suggestions of how these advantages vary across categories of race, gender, marital and family status. Specifically, we expect to find that the parenting penalty is most severe for single parents with many (more than two) children, while the benefits of marriage are greatest for men, particularly for white men. We also expect to find that the parenthood (or Daddy) bonus is greatest for men, perhaps even greater for married men and especially for married white men. Since racial and gender differences in the parenting penalty for single persons have not been adequately explored, we are prepared to find effects that might be enlightening.

Based on what we find as gross effects, we estimate OLS models to test the significance of the observed effects, beginning with the net effects of race, gender, marital and family status and then adding interaction terms to evaluate the extent to which (1) the marriage bonus is greater for men, particularly for white men, (2) the parenting penalty is limited to women, perhaps to married women or even to married white women, and (3) the parenting bonus is really a Daddy bonus that is solely available to men, perhaps only to married men, or to married white men. These effects are then re-estimated in a model that controls for education, age, occupation, class, and labor market measures to determine how robust these effects really are. Based on the results of these three OLS models we remove insignificant effects that are not necessary as components of higher order interaction terms until we have the best prediction of logged annual earnings that OLS and these variables could provide.

Then, using this model, we decompose the race and gender earnings differences into explained and unexplained effects of race, gender and family status. This allows us to consider how the marriage bonus of white men is due to the fact that they are more likely to be married when working for pay (negative explained effect), because they benefit from having a housekeeper (negative unexplained effect), or both. Similarly, we can see whether and to what extent racial and gender earnings inequality can be attributed to Motherhood penalties or Daddy bonuses, for single and married persons, if the effects prove to be significantly different for single and married persons.