Annotated Bibliography (APA Format)

Wealth Privilege and the Racial Wealth Gap: A Case Study in Economic Stratification

Robert B. Williams1

Published online: 30 September 2017 # Springer Science+Business Media, LLC 2017

Abstract The emerging subfield of stratification economics is a response to the orthodoxy’s resistance to recognizing the role of racial and ethnic disparities and its penchant for adopting cultural explanations for intergroup differences. With this view, the literature on the racial wealth gap and its particular embrace of the Life Cycle Hypothesis (LCH) offers a clear example of this critique at work. Not only is the LCH incapable of explaining why the racial wealth gap is so much larger than the income gap, but its limitations restrict the range of explanations explored. As an alternative, this paper introduces the Wealth Privilege (WP) model. Unlike the LCH, the WP model can incorporate the effects of contemporary racism as well as the systemic sources that are a legacy of several centuries of racialized policies. Using evidence from the 2013 Survey of Consumer Finances (SCF), this article offers empirical corroboration as well. Since the SCF queries households on their attitudes toward saving and investment, this article investigates the extent that cultural differences explain the wealth gap. To limit the problem of skewness, which is inherent in wealth studies, the analysis uses an inverse hyperbolic-sine transformation of household net worth. The OLS regression results show scant support for key features of the LCH while demonstrating the importance of asset ownership and family support, both crucial facets of the WP model. Two different decomposition methods, Blinder - Oaxaca and DiNardo - Fortin - Lemieux, corroborate these conclusions. As wealth is easily transferable across generations, the evidence supports the contention that household wealth serves as a source of economic stratifi- cation as it functions to preserve and even widen the racial wealth gap.

Keywords Racial wealth gap . Stratification economics . Householdwealth .Wealth privilege model . Structural racism

Rev Black Polit Econ (2017) 44:303–325 DOI 10.1007/s12114-017-9259-8

* Robert B. Williams [email protected]

1 Guilford College, 5800 W. Friendly Avenue, Greensboro, NC 27410, USA

Introduction

In recent years, there has been increased attention devoted to the racial wealth gap. While White households continue to hold substantial advantages in educational attain- ment, income, and occupational status, these disparities pale in comparison to the current wealth divide. Whereas White households typically earn about 60% higher incomes than Black or Latino households do, White household wealth surpasses Black and Latino wealth by tenfold or more (Kochhar and Fry 2014). Even when controlling for household income, Black or Latino households rarely hold one third of the wealth their White peers have (Tippett et al. 2014). This vast discrepancy makes any compar- ison among White, Black, or Latino middle-class households largely illusionary. Even more than income, household wealth enables families to get ahead, whether by financing additional education or professional certification, funding business opportu- nities, providing the down payment on a home in desired neighborhoods with better- resourced schools, or paying their children’s college tuition. As the racial wealth gap has been rising over the past generation (Asante-Muhammed et al. 2016; Kochhar and Fry 2014), there is concern that this widening gap will limit further gains by Black and Latino households in reducing the continuing disparities Bin the classrooms, work- places, and paychecks^ (Shapiro 2004, p. 183). Without a thorough understanding of the racial wealth gap, one cannot fully grasp the underlying forces that are obstructing any movement toward racial equality.

While the existence of a substantial racial wealth gap is indisputable, its underlying causes are not. Past studies have attributed the gap to a medley of causes, including differences in household income, age, and other demographic factors, varying attitudes toward saving, risk, and bequests, dissimilar asset portfolios, disparate inheritances as well as the effects of both contemporary and past sources of institutional racism. Given the complexity and breadth of these potential causes, it is essential that one use a conceptual framework capable of encompassing these factors. Yet, the past literature has relied upon a theoretical framework, the Life Cycle Hypothesis (LCH), whose limits make it unsuitable for this role.

Those who are encouraging the development of an emergent field within economics, stratification economics, have aptly anticipated this problem. According to one view (Darity et al. 2015), stratification economics is a response to the orthodoxy’s determi- nation to view racial and ethnic disparities as simply peripheral and transitory rather than systemic and enduring. By viewing racial discrimination as the result of irrational behaviors that cannot persist in market competition, orthodox economics defends its lack of interest in the topic. Due to these self-imposed blinders, the orthodoxy then finds itself reduced to arguing that cultural differences among groups must explain their intergroup disparities. In contrast, stratification economics acknowledges the intention- ality of discrimination particularly as a vehicle for maintaining status and power among the privileged group. Importantly for this analysis, it encourages a more expansive understanding of intergroup disparities as it considers various institutional and systemic factors that offer sources of privilege to members of the favored group. These structures of privilege need not be linked explicitly with either racial prejudice or bigotry (Darity et al. 2006) and thus they need not appear as racialized.

To test the efficacy of stratification economics, this article investigates the range of potential causes of the racial wealth gap. To date, much of the literature on the racial

304 Rev Black Polit Econ (2017) 44:303–325

wealth gap has relied upon the LCH (Krivo & Kaufman, 2004), itself the preeminent model of household saving and wealth accumulation. Yet, its unwillingness to recog- nize sources of privilege in the wealth accumulation process (Brown 2012) makes it an easy target of the critique offered by stratification economics. By neglecting to consider the role of systemic racism, the LCH encourages its adherents to consider only cultural explanations for the racial wealth gap. As an alternative, I propose using the more expansive Wealth Privilege (WP) model. The model incorporates systemic realities that suggest that the opportunities for wealth accumulation vary widely for households based on their current wealth and racial status. This paper does not simply compare the relative strengths of the LCH and WP models. In addition, it examines an expansive range of factors, including those frequently offered as sources of household wealth. This empirical analysis reveals the relative significance of both models as it discerns which factors provide the most compelling explanation for the racial wealth gap. I end the paper with some concluding remarks regarding the promise offered by this new field of stratification economics.

Literature review

Until the advent of various household surveys over the past fifty years, household wealth and its distribution had largely remained veiled from public scrutiny. While these surveys have brought significant transparency to this issue, they are not without their own problems. Querying households about their personal assets, debts, and inheritances leads to serious concerns regarding recall bias and non-response bias. These difficulties are exacerbated by the skewed distribution of wealth. Ignoring the wealthiest households, even unintentionally, means missing a sizeable portion of the wealth pie. Each of these problems creates challenges in securing accurate wealth estimates. In addition, these early surveys, following the convention of their day, offered respondents limited choices regarding their racial identity. As such, much of the previous literature considered only Black and White households when making race comparisons.

As the dominant model of household saving and accumulation, the LCH has played a leading role in the racial wealth gap literature. Many past studies make explicit reference to using a life-cycle model (Altonji and Doraszelski 2005; Blau and Graham 1990; Hurst et al. 1998; Juster et al. 1999) while another refers to Friedman’s perma- nent income hypothesis (Smith 2001). Though not referencing either model specifical- ly, other researchers have implicitly used it as the basis of their theoretical framework (Barsky et al. 2002; Keister and Moller 2000). Others (Conley, 1999; Gittleman and Wolff 2004; Oliver and Shapiro 2006) acknowledge the importance of life cycle analysis, although they refuse to succumb to its limitations. Most recently, three studies (Emmons and Ricketts 2017; Maroto 2016; Thompson and Suarez 2015) have taken a broader perspective in examining the racial wealth gap, though without a compelling conceptual framework.

Initially articulated by Modigliani and Brumberg (1954), the LCH posits that households seek to levelize household consumption over their lifetime. Household wealth follows a predictable life cycle in which young householders dissave as they invest in human capital, engage in substantial saving during their peak earning years,

Rev Black Polit Econ (2017) 44:303–325 305

and liquidate their assets over their retirement years. Ideally, households would exhaust their wealth just as death occurs to maximize consumption. According to the theory, unexpected bonanzas, whether from inheritances, earnings spikes, or asset appreciation, will have a temporary impact on household net worth as households respond to these events by raising their consumption expenditures accordingly. According to the model, wealth accumulation is essentially about household saving and wealth is simply a store of future consumption.

Despite its elegance, the LCH has shown a troubling inability to explain certain evidence. One testable hypothesis of the LCH is its prediction that the distribution of wealth should reflect lifetime earnings, particularly as one controls for age (Hendricks 2002). Clearly, this is at odds with the evidence, as wealth distribution is far more skewed than is income. Indeed, the literature has struggled with explaining why the rich accumulate so much more wealth than their income would predict, even irrespective of race. Simply arguing that households engage in precautionary saving to overcome earnings, retirement, and life expectancy risks cannot explain the concentration of wealth as the rich will stop saving once they achieve their desired buffer stock (De Nardi 2015). Instead, others argue that cultural differences among the rich and poor toward time preference (Krusell and Smith 1998) or toward leaving bequests to one’s children (Heer 2001; Laitner 2001) can better explain the upper tail. De Nardi (2004) shows that differences in family transfers, both the transmission of human capital and voluntary bequests, can replicate the actual distribution of wealth. Others posit the wealthy may save prodigiously because they face extreme risks in the form of volatile and unpredictable income (Castaneda et al. 2003) or because as entrepreneurs operating under borrowing constraints they want to expand their business to capture the returns from capital (Cagetti and De Nardi 2006). Lastly, others (Carroll 2000; Francis 2009) argue that the wealthy develop a Bcapitalist spirit^ in which they recognize wealth’s intrinsic value as a source of power, giving them an even greater motivation for saving than the rest of us. It appears from this literature that differences in attitudes toward saving among the rich and poor offer the best explanations for the skewed distribution of wealth. However, these studies suffer from the blinders of the LCH in that they neglect to include inherited wealth or asset ownership as part of their explanatory variables.

The LCH faces greater challenges in explaining the racial wealth gap. Thompson and Suarez (2015) find little evidence of expected net worth age profile – the model’s central prediction - among Black households. Other studies attempt to reconcile why the racial wealth gap so faintly resembles the racial income gap. Some studies argue that differences in household earnings and demographics (e.g. education, marriage, kids) can account for much of the wealth gap as each influences household saving (Altonji and Doraszelski 2005; Avery and Rendall 1997; Barsky et al. 2002; Gittleman and Wolff 2004; Menchik and Jianakoplos 1997). Yet, these studies generate wide- ranging estimates, normally between 20 and 90%, for what portion of the wealth gap they can account for. These wildly divergent estimates result from their reliance upon alternative sources of household wealth and their virtual neglect, with one important exception (Gittleman and Wolff 2004) for family inheritance and asset portfolio variables as part of their regression covariates. Worse, when applying regression decomposition, they find very different results depending on whether the Black or White estimates are used. This discrepancy suggests that White and Black households function under very different circumstances.

306 Rev Black Polit Econ (2017) 44:303–325

Others wonder whether Bcultural values^ might cause Black households to save less for retirement because they exhibit less patience or tolerance for risk (Scholz and Levine 2004), make poorer financial decisions (Boshara et al. 2015) or hold increased expectations of relying upon family support (Shin 2010). Taking a different view, Smith (1999) finds evidence that poorer health and higher medical expenditures along with lower life expectancy and reduced earnings can explain some of the racial wealth gap. Regarding the importance of bequest motives, the literature offers a mixed message as one study (Smith 1995) finds such motives as significant while another (Menchik and Jianakoplos 1997) does not.

Looking beyond the narrow confines of the LCH, other studies have investigated whether family inheritances might explain the racial wealth gap. Conley (1999) concludes that parental wealth is themost important predictor of household wealth. Two other studies (Avery and Rendall 1997; Menchik and Jianakoplos 1997) each find that differences in family transfers explain between 10 and 20% of the racial wealth gap. Comparing inheritance patterns among Black and White households over a ten-year period, Gittleman and Wolff (2004) come to similar conclusions. Chiteji and Hamilton (2002) estimate that 27% of the racial wealth gap among middle-income households is explained by family background, particularly as one considers both parental and sibling need. Narrowing the lens, another study (Charles and Hurst 2002) discovers that 42% of White homebuyers received help from their families in collecting a down payment as compared to 10% of Black homebuyers. Lastly, Chiteji and Stafford (2000) demonstrate that parental wealth affects the transfer of financial knowledge across generations and thereby influ- ences asset ownership. Although they argue this effect does not contribute mightily to the racial wealth gap, they note that it could accumulate over several generations.

Lastly, several studies examine howportfolio choices might influence the racial wealth gap. Two early studies (Gittleman andWolff 2004; Hurst et al. 1998) suggest that around 10% of the racial wealth gap could be closed if Black asset portfolios resembled those of White households. Both studies attribute the primary role to differences in stock owner- ship rates. A third study examines the lower rate of self-employment within the Black community and concludes that it could also explain another 10% of the racial wealth gap (Menchik and Jianakoplos 1997). Three recent studies suggest an even strong role of asset portfolios. Thompson and Suarez (2015) examine homeownership alone and conclude it could close the racial wealth gap by between 13 and 16%. Maroto (2016) includes only homeownership, stock ownership, and credit card debt and estimates they could account for as much as 35 to 45% of the racial wealth gaps.1 Taking the most comprehensive approach to date, Emmons and Ricketts (2017) conclude that differences in household balance sheets account for over half of the wealth gaps between both Black and Latino households with White households.

Although the literature just discussed offers insights into the racial wealth gap, it is ultimately limited by its reliance upon a conceptual framework poorly suited to answer the key conundrum: how does an income gap of less than two become a wealth gap that exceeds ten? According to the LCH, the vast advantage that Whites hold regarding inherited wealth should lead to increased consumption, not a persistent wealth gap. Similarly, any financial advantages experienced byWhite households due to their greater levels of asset ownership should lead to disparities in consumption, not wealth. In

1 Unfortunately, she lumps these three covariates with a measure of inheritance.

Rev Black Polit Econ (2017) 44:303–325 307

limiting its focus on household savings behavior, the LCH is inclined to argue that reduced savings among households of color due to their shorter time horizons, increased risk aversion, or disinterest in leaving a legacy are the sources of the racial wealth gap. Further, the LCH ignores the dynamic elements of wealth accumulation as it views wealth as a store of consumption and underestimates its potential as a source of power. It ignores how the distribution of wealth and power from one generation influences their distribution in subsequent generations. Finally, it fails to recognize how the distribution of wealth is the result of past and present sources of institutional racism.

One final study deserves some additional attention due to its design. Unlike the prior studies that examined the differences in wealth holdings, Gittleman and Wolff (2004) examined racial differences in the accumulation of wealth. Tracking specific house- holds over a ten-year period, (1984–94), they compare White and Black households’ capacity to accumulate wealth based on their savings behavior as well as their asset portfolios and family inheritances. As such, they conduct a more expansive search for possible causes. They find that Whites accumulated wealth at higher rates than did Black households largely due to greater family inheritances and increased household saving. The authors note that this increased saving by White households resulted from earning higher incomes, not an increased motive for saving. Interestingly, they do not find that White households experienced higher rates of return on their portfolio, despite having far more assets. They conclude that if Black households earned equal incomes, received similar family inheritances, and held comparable portfolios to Whites, these changes would narrow the racial wealth gap, though moderately2.

The Wealth Privilege Model

The Wealth Privilege (WP) model offers an important improvement over the LCH to those interested in examining the racial wealth gap from the perspective of stratification economics. Rather than simply view wealth as a store of consumption, the WP model recognizes how wealth functions as a source of power for its owners. Possessing a car can expand one’s employment options. Buying a home in select neighborhoods offers access to better-resourced schools even as it provides comfort, stability, and financial security. Wealth can subsidize education and training needed to open career doors or serve as start-up capital for a business venture. It can advance one’s political or social interests either in the form of political donations or charitable giving. Quite simply, wealth expands the choices, opportunities, and agency of its holder. As Raymond Franklin (1991) provocatively argues: BOwnership carries with it domination; its absence leads to subordination^ (p.xviii).

As a source of power, wealth embodies two important characteristics, its durability and transferability across generations. We treasure various forms of wealth for their capacity to retain their value over time and during crises. Wealth helps us weather personal emergencies as well as secure a comfortable retirement. Wealth can influence the future as parents finance their children’s college tuition, assist them in home buying, and fund their grandchildren’s schooling. Labeled Btransformative assets^ by Thomas Shapiro (2004, p.2), these gifts have the capacity to expand the opportunities of their

2 The reader can find a fuller explanation of this model in Williams (2017).

308 Rev Black Polit Econ (2017) 44:303–325

recipients. Any remaining wealth will simply pass to the next generation largely unimpeded at one’s death. 3 Due to these two attributes, the distribution of wealth across society resists dramatic shifts. Over the past half century, the dismantling of Jim Crow and the resultant changes in public policies, employment practices, and social norms have expanded opportunities for persons of color. Yet, these changes have had little impact on the racial wealth gap. In 2013, the real absolute wealth gaps between White and both Black and Latino households is greater than those experienced in 1983 (Kochhar and Fry 2014). Due to its durability and transferability, Bwealth has the particular attribute of tending to reproduce itself in a multiplicative fashion from generation to generation^ (Conley, 1999, p. 25).

Not only does the WP model recognize the wealth itself is a source of power, but it also acknowledges the role of power and hierarchy within the wealth accumulation system. Each of the primary pathways of wealth accumulation – household saving, family support and inheritances, and asset appreciation – shares a common trait; each functions as a self-reinforcing feedback loop beyond some threshold. Along each of these avenues of wealth accumulation, labeled the Household Saving, Asset Appreci- ation, and Family Support pathways, the circumstances of wealth accumulation change as one accrues more wealth. This explicit recognition enables the WP model to explain why the racial wealth gap persists and grows over time.

The WP model is quite straightforward. In the Household Saving pathway, past saving can fund income-generating assets, from savings accounts to rental properties, which simply increase income and permit greater saving in the future. This virtuous cycle not only permits households to experience increased saving as their income rises, but a higher savings rate as well. This is an insight that only economists might find surprising (Dynan et al. 2004). The Asset Appreciation pathway functions in a similar fashion. As households accumulate assets, they have greater opportunities to diversify their portfolio enabling them to assume increased investment risk. This assumption of risk permits them to benefit from higher-return assets, thereby yielding a larger portfolio, increased diversification, and the assumption of further risk. Increased wealth offers greater access to credit, opening the opportunity for debt leveraging and even greater returns. Consequently, larger portfolios can generate disproportionately higher rates of return and fuel faster appreciation.

The Family Support pathway functions in a comparable way across generations. Affluent parents have multiple ways they can boost their children’s prospects. With the gifts of superior education, cultural experiences, and social contacts, their kids can parlay this support into higher salaries and increased saving. Wealthy parents can offer in vivo gifts at milestone events like college admission, weddings, starter homes, and their grandchildren’s schooling. They can transmit their financial knowledge to their children, giving them greater exposure and comfort with different investments. The most visible, but arguably not the most important, form of family transfer is the gift of wealth at one’s death. Each of these forms of family support enables the beneficiaries to profit more fully from the advantages offered by the saving and asset appreciation pathways. According to Williams (2017), households in the wealthiest quintile inherited 80% of all reported family transfers. Although these households tend to be

3 In 2017, estates below $5.490 million escape estate taxation. Tax loopholes allow larger estates to avoid taxation as well.

Rev Black Polit Econ (2017) 44:303–325 309

older giving them more time to collect an inheritance, they also express greater optimism about receiving future gifts than do their peers. As most parents simply ask their children to pay this help forward to the next generation, this pathway operates like a virtuous cycle across the generations.

Of course, all three pathways complement and reinforce each other. Graduating from college debt-free not only offers the possibility of an ample salary, but also the opportunity for immediate investment. Absent the drain of student loans, households can save immediately for their starter home. Getting additional family help with the down payment only sweetens the experience. Accessing the Asset Appreciation path- way at an earlier age enables those households to benefit more fully from its exponen- tial rewards. Rising asset values raise household income and boost household saving further. Each of these advantages stokes wealth accumulation, giving households greater opportunities to help their kids in turn.

Not only are these virtuous cycles beyond the reach of most wealth-poor households, but also these households experience starkly different circumstances. Low pay, insuffi- cient hours, and unstable employment force some households to liquidate saved assets to meet current expenses even as they recognize the bleaker prospects ahead. Without some improvement, they must continue to liquidate assets or draw upon credit to meet their present obligations. In either event, escape from this spiral becomes increasingly difficult. Even worse, nearly a quarter of low-income households remain unbanked, largely due to high fees and inconvenient locations (Bucks et al. 2006). This deprives them of earning income on their savings as well as key services that facilitate saving.

The Family Support pathway functions in a comparable way. Normally, we view family gifts as simply going from the older to younger generations. Reality is more complex. In many families, parents or other family members may suffer from poor health and generate unpaid medical bills. Parents and grandparents may outlive what- ever nest egg they were able to accumulate. These circumstances may generate requests for help from younger family members. In providing this assistance, these households diminish their own capacity to build a retirement fund, perhaps causing them to make similar requests of their children. This makes clear that households from wealth-poor families not only start out with less, but also they experience a greater likelihood of receiving requests for help. This type of family transfer has a clear racial pattern. Black and Latino households are almost twice as likely to report helping a parent or older family member in the past year as do White families (Williams 2017). Among these donors, Black households give on average twice the gift of White families, despite having fewer means. Not only does the wealth of one generation reach into the next, but also its dearth in one generation can retard the next.

In a similar, though less dramatic fashion, the Asset Appreciation pathway functions poorly for households of modest means. In most cases, the purchase of a car or truck serves as most household’s first major investment. Having a car is frequently essential to gaining and maintaining employment. Yet, the family car along with household furniture and appliances are depreciating assets. As many households have the bulk of their limited wealth invested in these three categories, they have little opportunity to benefit from asset appreciation. Given their limited means, asset thresholds deter their entry into asset appreciation. Down payments limit who can become homeowners by requiring years of saving. Even then, wealth-poor households encounter fewer choices regarding neigh- borhoods and homes they can afford, often leading to lower rates of appreciation. Among

310 Rev Black Polit Econ (2017) 44:303–325

mobile home buyers, most do not purchase the land their home sits on, causing them to miss the financial benefits of homeownership (Collins and Dylla 2001).

The confluence of institutions, policies, and behaviors that favor the affluent as all households strive to better their circumstances is what gives rise to the model’s name. In some cases, largely desirable behaviors encourage wealth privilege. It seems inescapable that higher-income households will find it easier to save a larger portion of their income. With larger andmore diverse portfolios, would we not expect wealthier households to take greater risks and earn higher returns? Regardless of their resources, parents will support their children as best they can in assuring them a head start in their lives; wealthier parents simply have greater means to do so. In each case, these sources of wealth privilege stem from predictable behaviors making these favors appear largely inevitable.

Other sources of wealth privilege stem from institutions and policies that favor the affluent. Not all workers have access to retirement plans with employer matches, thereby reserving this opportunity for involuntary saving to the better-paid employees. Similarly, most homeowners engage in some form of saving with each mortgage payment while renters gain no such benefit. Banks offer a bevy of services to help their customers save. With direct deposit, affluent households can direct portions of their paycheck into a savings account, making saving easy and automatic. Online banking services offer their customers bill payment reminders and automatic payment, thereby protecting them from late fees and interest charges. Yet, as Mullainathan and Shafir (2009) argue, banks treat their customers with

Ba built-in asymmetry in banks’ incentives between credit and savings for the poor and the rich. Regarding poor clients, banks have a greater incentive to promote debt (which can be lucrative, delayed, and compounded) rather than savings (which are bound to be modest), as opposed to the treatment of the wealthy, whose debt is likely to be repaid with little penalty and whose savings promise to be large and valuable^ (p. 134).

Interestingly, remaining unbanked may signal financial acumen among the wealth- poor. To compound the issue, financial lenders make credit more available (and cheaper) to the wealthy, providing them with greater opportunities for leveraged investments. In these and other ways, the wealthy gain financial favors that are not universally available.

Other public policies follow suit in favoring the wealthy over the wealth-poor. Generous tax exemptions target retirement accounts to encourage additional saving even though these accounts ignore the savings needs of low-income households who require flexible Brainy day funds^. Both the home mortgage and local property tax deductions help homeowners save while ignoring the needs of renters. Other tax deductions like the exclusions on home sales and capital gains enable the wealthy to shield more of their market gains from the tax collector. Even family estates receive special treatment as all but the largest estates pass through untaxed while the step-up in basis insulates the beneficiaries from capital gains taxes. Together these deductions comprise several hundreds of billion of dollars annually (Joint Committee on Taxation 2013) with the vast bulk of it going to the wealthiest households (Williams 2017). Lastly, repeated reductions in the estate and gift taxes over the past generation permit larger transfers of wealth from one generation to the next escape taxation.

Rev Black Polit Econ (2017) 44:303–325 311

Explaining the racial wealth gap

The WP model can explain the widening racial gap from two perspectives not readily available to the LCH. Although the WP model functions as a system of privilege and power without reference to race, its consequences have clear racialized outcomes. Indeed, its covert nature as a system of race privilege and oppression gives it an insidious fig leaf of social approbation. Yet, the link between household wealth and racialized status is inarguable. Given the capacity of wealth to endure and grow over time and across generations, it is no surprise that our current wealth gap is rooted in our nation’s past racialized history. Many (Conley, 1999; Darity and Nicholson 2005; Lui et al. 2006; Oliver and Shapiro 2006; Williams 2017) have shown how our nation’s history of enslavement, genocide, and expropriation of property has shaped the current racial disparities in wealth. In 2013, median net worth among White households is $141,900 while the comparable figures for Black and Latino households are $11,000 and $13,700 respectively (Kochhar and Fry 2014). Even more telling, Black and Latino households are concentrated in the bottom wealth quintiles while White families predominate among the top wealth quintiles as illustrated in the graph below. Raised in families with vastly reduced resources, young Black and Latinos will find fewer opportunities to earn a college degree. Without a diploma, they will find it challenging to earn sufficient income to save regularly for the future. Relative to their White peers, only a fortunate few can expect any inheritance. Indeed, they will face increased pressure to offer help to older family members in financial distress. Lacking the means for a down payment, they will find homeownership beyond their grasp, thereby missing its many benefits. Unable to access the Asset Appreciation pathway, they will find it difficult to offer their children additional help. Given the advantages of wealth as predicted by the WP model, this demographic pattern suggests that the racial wealth gap will widen, even in absence of persistent racial discrimination.

Racial Composition of Wealth Quintiles, 2013

0%

20%

40%

60%

80%

100%

Bottom Second Middle Fourth Top

P e r c e n

t o

f Q

u i n

t i l e

Black/Latino

White

Source: 2013 SCF, Author’s calculations.

Of course, there is extensive evidence that Black and Latino households continue to suffer from racial discrimination, particularly in labor, credit, and housing markets. African Americans and Latinos earn substantially lower salaries and experience higher unemployment, even accounting for similar education levels. Various studies (Bertrand

312 Rev Black Polit Econ (2017) 44:303–325

and Mullainathan 2003; Pager 2003) show that White applicants are more likely to get callbacks for potential jobs than are Black applicants. Adding to these challenges, Blacks and Latinos experience much lower rates of educational attainment, the result of numerous historical and contemporary causes.

In their efforts to get ahead, Blacks and Latinos experience additional challenges in gaining access to credit. Two different studies conclude that Black and Latino mortgage applicants are rejected far more frequently than comparable White applicants (Charles and Hurst 2002; Munnell et al. 1996). Even worse, many more potential Black appli- cants simply do not apply believing they will be rejected (Charles and Hurst 2002). Among those getting loans, Black borrowers pay higher interest rates on car loans, student loan debt, and home mortgages (Chiteji 2010). Black and Latino homebuyers are more likely to use FHA, VA, and FMHA loans that require smaller down payments and charge higher rates; on both counts, they restrict home equity (Krivo & Kaufman, 2004). Several studies show how restricted access to credit markets affects the survival rates of Black-owned businesses (Bates 1989; Blanchflower et al. 2001; Fairlie 1999). Another study concludes that discrimination among White customers can explain some of the observed differences in self-employment rates across race (Borjas and Bronars 1988). According to Fairlie and Meyer (2000), Black self-employment rates have remained at one third of the levels among Whites for over 90 years.

When looking for housing, Black and Latino households continue to face obstacles. In a paired-testing study, applicants of color received unfavorable treatment from real estate or rental offices over 20% of the time (Turner 2002). Prospective Black and Latino homebuyers were given less information and fewer opportunities to inspect advertised homes. Given the persistence of residential segregation (Iceland and Weinberg 2002), this permits the possibility that White homeowners realize greater returns from their homes than do Black or Latino homeowners. The empirical evidence is mixed as some find home appreciation rates favoring Black homeowners (Gittleman and Wolff 2004), neither racial group (Coate and Vanderhoff 1993; Long and Caudill 1992), or White homeowners (Flippen 2004; Oliver and Shapiro 2006; Williams 2017).

Method

To test these arguments, I use the household wealth figures collected in the 2013 Survey of Consumer Finances (SCF). The SCF is a single-wave, cross-sectional survey that queries households every three years. Although this design limits its capacity to examine how households accumulate wealth over time as it simply offers a one-time snapshot, the SCF does offer two important benefits. It combines a random sample of 4568 households with a selected oversample of 1458 households to ensure adequate representation of the very wealthy, a group often reluctant to disclose their family wealth. Given the wealthiest 3% of households own nearly half of the household wealth, this represents an important advan- tage. Unlike other household surveys, the SCF queries household wealth as its primary focus. The range and depth of its questions regarding household wealth offers researchers a unique and nuanced understanding of contemporary household wealth, causing some to refer to it as the Bgold standard of wealth data^ (McKernan et al. 2014, p. 5).

The breadth of questions included in the SCF allows one to examine the relative effectiveness of both the LCH and WP models using a broad range of variables.

Rev Black Polit Econ (2017) 44:303–325 313

Household net worth includes both real (including vehicles) and financial assets minus any household debt.4 To explain household net worth, I include standard variables like householder age, marital and parental status, and educational attainment. To create a comprehensive set of attributes that likely influence household saving, I combine a number of variables under two additional groupings. Under saving capacity, I include a measure of normal income,5 as queried by the SCF, as well whether a household reports good health, health insurance coverage, or bouts of unemployment over the prior year. In addition, I add whether the household reported actual saving over the previous year. All of these variables influence a household’s capacity to save. To ascertain saving motivation, I add variables that measure household attitudes toward the use of credit, length of planning horizons, life expectancy, importance of leaving a legacy, and predictability of their future income. Each of these offers insight into key factors that likely influence household savings behavior, particularly as suggested by the LCH literature. Additionally, I include whether households reported engaging in regular saving or being late in paying their bills over the previous year. Collectively, these 16 variables provide a comprehensive portrait of household saving.

In addition to measuring these influences on household saving, I also include variables that capture the Asset Appreciation and Family Support pathways. Regarding the former, I include variables on whether households are homeowners and mortgage holders as well as owners of retirement funds, directly held stocks, rental property, or a business. Each of these reflects the range of opportunities experienced by households as they strive to accumulate wealth. I also add a variable that measures attitudes toward risk to discern its impact. Together, these variables capture the importance of asset ownership, portfolio choices, and the risk return tradeoff. To measure Family Support, I use five variables, including two that measure the extent of past and future inheritances. Since many forms of family help are not always viewed as an Binheritance^, I add three other variables. One form of such help is paying for school; consequently, I use whether households report outstanding educational debt as one measure. Another is the percep- tion that households have access to a safety net from families and friends in an emergency. Lastly, I include whether the household has provided financial support to their siblings, parents, or grandparents. All five of these variables provide some insight into family resources and their potential impact on wealth accumulation.

Results

Descriptive statistics

To illustrate the different circumstances facing Black and Latino households from White households, I present the mean values for all three groups in Table 1. Frankly, there are few surprises. While Black and Latino households confront quite similar conditions, White households report very different circumstances. On average, White households hold nearly seven times the net worth of Black or Latino

4 The SCF does not measure future claims to Social Security. 5 Although this measure of normal income does not fully capture the permanent income concept, it likely represents the best driver of savings behavior..

314 Rev Black Polit Econ (2017) 44:303–325

households.6 In addition, White households are older, better educated, and less likely to have children. Where these differences are statistically significant (.05 error) with both Black and Latino households, I have noted it with an asterisk. One major

6 As wealth is slightly more concentrated in the black and Latino communities, the wealth gap at the mean is smaller than that at the medians.

Table 1 Mean values of key variables

Variable White Black Latino Households

Net Worth $677,542 $95,050 $111,215

Demographic

Householder Age (yrs.) 53* 49 44

Married (%) 60 36 63

Dependent Children (%) 38* 50 59

Has College Diploma (%) 43* 27 18

Household Saving

Normal Income $98,463 * $45,841 $48,722

Good Health (%) 74 68 73

Health Insurance (%) 83* 74 55

Unemployed (%) 12* 21 21

Did Save (%) 47* 32 32

Regular Saving (%) 56 58 54

Years Remaining 30* 39 38

Credit – Vacation Ok (%) 14 14 15

LR Plan Horizon (%) 35* 21 18

Legacy Important (%) 49 63 63

Variable Income (%) 23* 45 44

Late Payments (%) 06* 13 09

Asset Appreciation

Risk Taker (%) 18 15 12

Homeowner (%) 73* 44 44

Holds Mortgage (%) 45* 30 29

Retirement Account (%) 56* 34 25

Owns Stock (%) 17* 03 03

Owns Other Real Estate (%) 21* 09 08

Owns Business (%) 14* 06 06

Family Support

Amount Inherited ($) 139,715* 14,793 8304

Expecting to Inherit ($) 80,811* 9716 11,069

Has Education Debt (%) 18 32 14

Distress Giver (%) 05* 10 11

Safety Net (%) 70* 43 49

*Denotes Significance (.05) between Whites and both Black and Latino households

Rev Black Polit Econ (2017) 44:303–325 315

difference between Black and Latino households is the low marriage rate among Black households.

Regarding other factors that might influence household saving, White households earn twice the normal income, report slightly better health, and experience greater health coverage and lower risk of employment than do Black or Latino households. Interestingly, all three groups regularly save at similar rates while Whites report actually saving at higher rates. Clearly, intention versus agency is at work here. Among the attitudinal variables, White households report longer planning horizons and greater tolerance for risks. Conversely, both Black and Latino households attach a greater importance to leaving a legacy, expect more years remaining, 7 and report greater unpredictability in their incomes. All three groups report similar attitudes toward the use of credit while both Black and Latino households report making late payments at higher rates than do White households.

Between the remaining two categories of variables, these patterns largely continue. By wide margins, White households report much higher levels of ownership of key assets, whether real estate, stocks, or businesses. Similarly, White households are far more likely to hold any retirement accounts. Collectively, these figures demonstrate the degree that White households have greater access to these, key sources of wealth accumulation. Curiously, Whites report only a modest increase in their tolerance for risks. Regarding family transfers, White households report receiving substantially higher amounts of past family gifts and expect an even greater bounty in the future. Greater family resources mean that White households are less likely to carry educa- tional debt as well as engage in direct assistance to older family members, particularly compared to Black households. The relatively low rate of educational debt among Latino households is attributable to their relatively low levels of college attainment rates. It is in these two areas that we find the greatest differences among Black and Latino households. Lastly, White households report at much higher rates access to a personal safety net, in which they could get at least $3000 from family or friends in an emergency.

OLS regression results

Next, I run OLS regressions to discern which factors appear to affect household wealth. Given their similar circumstances, I combine Black and Latino households and com- pare their results to White households. Given the skewed distribution of household wealth, I use an inverse hyperbolic sine (IHS) function to transform the dependent variable. Similar to a log transformation, this function generates coefficients that can be interpreted as elasticity. Unlike log transformations, the IHS function can accommodate non-positive values. As nearly 10 % of households fall into this category, this offers a significant advantage. To capture a more realistic representation, I have log transformed the two direct inheritance variables and the normal income variable.

The results are presented in Table 2. To interpret these results, recall the essential predictions of both models. As a savings model, the LCH would gain increased corroboration as more of the demographic and household savings variables are signif- icant. According to this view, Bcultural differences^ between the household groups

7 Among black households, the size of this number is partially, but not wholly, explained by their younger age.

316 Rev Black Polit Econ (2017) 44:303–325

might explain how these variables might lead to varied savings rates that explain the racial wealth gap. Further, the LCH predicts that asset ownership rates and inheritances would increase consumption expenditures, thereby having a mixed impact on wealth

Table 2 Regression results

Households White Black & Latino

Sample Size 4421 1306

Variable Coefficient p-value Coefficient p-value

Demographic

Householder Age 0.044 .238 −0.017 .827

Age-Squared −0.000 .925 0.000 .700

Married 0.570 .012 0.823 .088

Has Children 0.764 .001 0.642 .141

Has College Diploma −0.288 .205 −1.010 .105

Household Saving

Ln Normal Income 0.478 .021 0.944 .017

Good Health 0.605 .011 0.550 .230

Health Insurance 0.359 .258 −0.100 .837

Unemployed −0.421 .265 −1.298 .027

Did Save 0.831 .000 1.456 .001

Saved Regularly −0.228 .239 −0.536 .243

Years Remaining −0.004 .711 −0.020 .242

Credit – Vacation Ok 0.072 .802 −0.105 .868

LR Plan Horizon 0.341 .102 0.191 .710

Legacy Important 0.091 .624 0.671 .140

Variable Income −0.125 .594 0.128 .766

Late Payments −3.241 .000 −1.864 .029

Asset Appreciation

Risk Taker 0.099 .702 0.041 .940

Homeowner 4.016 .000 5.534 .000

Holds Mortgage −0.696 .000 −3.522 .000

Retirement Account 1.426 .000 2.619 .000

Owns Stock 0.856 .000 1.389 .241

Owns Other Real Estate (%) 0.868 .000 1.058 .118

Owns Business 0.525 .063 2.681 .003

Family Support

Ln Amount Inherited 0.050 .001 0.123 .027

Ln Exp. to Inherit 0.066 .002 −0.057 .521

Has Education Debt −5.423 .000 −6.299 .000

Distress Giver 0.313 .499 1.093 .077

Has Safety Net 0.647 .008 0.874 .048

Intercept −1.958 .367 −5.088 .234

F-Test 59.45 .000 29.22 .000

Rev Black Polit Econ (2017) 44:303–325 317

accumulation. In contrast, the WP model argues that both asset ownership and family assistance are important avenues for wealth accumulation. This model would receive greater corroboration as more of these variables gain importance in explaining house- hold wealth.

According to the results, the five demographic variables play a modest role in explaining household wealth. In neither case are the age variables statistically signif- icant. While these results are not that surprisingly for Black and Latino households given the findings of Thompson and Suarez (2015), they contrast sharply with what the previous literature has found among White households. While these results appear to confound a major prediction of the LCH model, that is probably an overstatement. The inclusion of other income, saving, and asset ownership variables in this model simply capture the effects that the age variables reflect in less complete models. The remaining results indicate the economic benefits of marriage across the board while the coeffi- cients on having children suggest a legacy effect at work, although it is statistically significant only for White households. While the results suggest that a college educa- tion has an adverse role on household net worth, one must remember that a diploma’s primary benefit is captured by the household income variable.

Turning to the Household Saving variables, the results reveal remarkable similarity across both groups. Household income and actual saving behavior are strongly linked with increased household net worth for all households.8 The higher coefficients found in the Black and Latino regression results suggest that this source of wealth accumu- lation may have increased importance to these households. Across the board, there exists a clear penalty for households making late payments. To be sure, the circum- stances are not identical for both groups. Among Black and Latino households, a bout with unemployment has clearly adverse circumstances while White households expe- rience a gain in net worth if they enjoy good health and, perhaps, if they engage in long run financial planning.9

It is also worth noting the list of household behaviors and attitudes that have little role in wealth accumulation. Households that report saving on a regular basis have lower net worth, although the links are not significant. Given the results regarding actual saving, it appears that agency is far more important than simple intent. Merely engaging in disciplined behavior appears to matter little if one’s income is insufficient to overcome unexpected outlays. Attitudes toward leaving a legacy or towards the use of credit have no apparent impact on wealth accumulation. Further, the results offer no corroboration that either unstable earnings or longer life expectancy produce any precautionary savings effect as some have speculated. What is striking is how few of these attitudinal variables find corroboration in the regression results.

We find much greater confirmation of the variables in the remaining two sections. In both cases, asset ownership (and debt holding), whether one’s home or other real estate, or stocks, retirement funds, or business enterprises, is strongly linked to increased (decreased) net worth. On the other hand, the one attitudinal variable, tolerance for risk- taking, does not reach the standard of statistical significance. Among the Family Support variables, past inheritances as well as the holding of educational debt are

8 Although the income coefficients are high, they are within the range found by Thompson and Suarez (2015). 9 The p-value on long run financial planning narrowly misses the standard for statistical significance among White households.

318 Rev Black Polit Econ (2017) 44:303–325

clearly related to wealth accumulation, regardless of one’s race. Similarly, access to a financial safety net also is linked to increased household wealth. Among Whites, the expectation of future family gifts is related to household net worth, suggesting some form of family support has already taken place. Among Black and Latino households, the distressed giving variable is linked to household net worth. One might wonder about the particular sign of this coefficient. Presumably, family requests will be extended (and granted) only to (and by) those family members who are relatively prosperous. Despite the sign of this coefficient, this transfer undermines the efforts of the giving household to accumulate wealth.

Oaxaca – Blinder results

While the prior results illustrate the impact of each covariate on households’ efforts to accumulate wealth, we can use the Oaxaca - Blinder (O-B) decomposition to determine their role in explaining the racial wealth gap. This technique works in the following way. Using the regression analysis presented in Table 2, the O-B method recognizes that the racial wealth gap is the result of two factors: the differences in the observed characteristics (the covariates) as well as the differences in each characteristic in its contribution to household wealth. The O-B method decomposes the two effects by examining how much the wealth gap would be narrowed if Black and Latino house- holds experienced levels of each covariate equal to Whites. What would happen to the racial wealth gap if Black and Latino households earned the same income, held the same attitudes, attained the same breadth of asset ownership, and received the same level of family support? Not only does this counterfactual estimate how much of the gap would vanish, but also the contributions of each covariate to that narrowing. I use the coefficients from a pooled model that incorporates both groups as the reference group, thereby eliminating the problem of two sets of estimates mentioned earlier.

According to the results in Table 3, 85% of the racial wealth gap would be erased if Black and Latino households attained similar levels to White households in each of the observed covariates. Orthodox economics would argue that the remaining 15% unex- plained portion would measure the level of racial discrimination, as it would gauge the level of special treatment accorded to White households. Under the WP framework, this

Table 3 Oaxaca – blinder decomposition results

Variables O-B Estimate Percent Explained

Demographics 0.075 2%

Saving Capacity 0.546 14%

Saving Motivation 0.243 6%

Asset Ownership 1.787 45%

Family Support 0.689 17%

Total Explained 3.340 85%

Total Difference 3.940

Includes the normal income, health status and insurance coverage, unemployment, and actual saving variables.

Includes the remaining six attitudinal variables under Household Saving and the risk-taking variable.

Includes all of the Asset Appreciation variables less the risk taking variable.

Rev Black Polit Econ (2017) 44:303–325 319

would estimate the lowest value, since past and current discrimination have led to the disparities in the observed covariates as well.

Table 3 exhibits the relative contribution of each of the categories of observed traits as well. Differences in demographics, including age, education, and family structures only accounts for 2% of the racial wealth gap. Similarly, if Black and Latino households earned the same income, experienced similar levels of unemployment, good health, and health insurance coverage, and saved at comparable rates toWhite households, the racial wealth gap would be narrowed by an additional 14%. Sharing similar cultural attitudes regarding saving motivation would reduce the gap by a mere 6%. These figures place this study within the range of past studies, though admittedly at the lower end. Similarly, differences in the Family Support variables explain about 17% of the gap, well within the range of estimates cited in the literature. In contrast, the results indicate that differences in asset portfolios account for 45% of the wealth gap. While this figure far outstrips the early estimates, each of the three most recent studies (Emmons and Ricketts 2017; Maroto 2016; Thompson and Suarez 2015) offers corroboration. It is certainly possible that the widening wealth gap over recent decades has caused asset ownership to assume greater significance. More certainly, these studies have moved away from the self-liming LCH framework and included one or more asset variables in their models.

Non-parametric (DFL) results

One criticism leveled at the O-B decomposition method is that it assumes a linear relationship between the observed variables and the returns to wealth. Given the highly skewed nature of wealth distribution, this represent a serious concern (Barsky et al. 2002), despite using the IHS and log transformations. To address this concern, I use the DiNardo – Fortin – Lemieux (DFL) decomposition to serve as a check (DiNardo et al. 1996). This method allows for reweighting the individual samples without assuming any parametric relationship. Further, one can use this with unconditional quantile regression to examine the effects of the covariates along the full range of the wealth distribution, rather than simply at the mean (Firpo et al. 2007). In this way, the DFL decomposition can be used to check and expand the range of the O-B estimates.

The results in Table 4 generally corroborate the O-B decomposition results. Across much of the wealth distribution, differences in Asset Ownership represents the most important source of the wealth gap while differences in Saving Motivation have little impact. This indicates that actual opportunities and not simply discipline and good intentions are crucial. The declining importance of the Asset Ownership among the affluent households is likely due to the specification of these variables. Among the very wealthy, asset ownership is widespread making any differences increasingly modest.10 Interestingly, differences in Family Support have a waning influence while differences in Saving Capacity have a waxing influence as we move up the wealth scale. It appears that differences in the incomes among the very affluent are the cause of this trend.11 Also, the observed values appear to explain household wealth in the vast middle more than in the two tails.

10 Among this group, measuring each asset category by its share of total assets may offer a more insightful method. 11 Evidence for this point is not provided in the paper.

320 Rev Black Polit Econ (2017) 44:303–325

Conclusions

While the LCH has a long and fruitful legacy, it does not serve as an effective vehicle for examining the racial wealth gap. By viewing wealth simply as a store of future consumption, the LCH ignores wealth’s unique role as a source of power, thereby overstating its vulnerability to diminishing utility. Not only does the model ignore the key motive for accumulating large amounts of wealth, the LCH neglects the essential means by which households accumulate wealth as well. Although the model nominally includes family inheritances and asset appreciation as sources of wealth, neither is viewed as playing a significant role in its distribution across society. Instead, the LCH posits that household saving is the crucial source of accumulated wealth. Following this perspective, many of the earlier studies on the racial wealth gap simply ignored family gifts and household portfolios as they examined the wealth gap. Clearly, this represents a serious omission given the results reported here.

This paper redresses this omission. In addition to the usual demographic, income, saving, and attitudinal variables, this paper includes asset portfolio and family support as equally important sources of wealth accumulation. It models these areas in ways not done before. The asset portfolio section includes not only measures of homeownership, but also other asset ownership variables to capture some of the risk-return rewards. To model family support, this paper integrates not only the two direct inheritance variables, but also three other, less direct, measures of family support. Both the regression results and the decomposition figures corroborate the importance of not simply including both of these sources of wealth accumulation, but also specifying them in detail.

Further, this paper acknowledges the paradigm shift provided by the WP model. While this model shares with the LCH a mutual understanding of the primary ways that households accumulate wealth, it also makes significant departures that provide a fundamentally different understanding of the wealth accumulation process. The WP model recognizes wealth’s unique role as a source of power, both within and across generations. Its durability and transferability makes it an effective vehicle for reproducing economic stratification across generations. Accumulated family wealth can travel across the generations through the transmission of human capital, financial knowledge, and familiarity with investment opportunities, the granting of in vivo gifts and Btransformative assets^, and the bequeathing of estates. Not only does this intergenerational transmission create an unequivocal link between the present and our racialized past of enslavement, extermination, and expropriation, but it also foreshadows the future. Family

Table 4 DFL decomposition results

Wealth Percentiles 10th 25th 50th 75th 90th

Demographics 4% 5% 14% 8% −4% Saving Capacity 7% 9% 13% 41% 58%

Saving Motivation 6% 7% 5% 1% 2%

Asset Ownership 37% 62% 44% 39% 12%

Family Support 17% 21% 9% 5% 0%

Total Explained 71% 105% 86% 94% 67%

Rev Black Polit Econ (2017) 44:303–325 321

resources largely determine where young householders find themselves along each of the wealth pathways, thereby establishing howmuch they can benefit from the virtuous cycles found along each avenue. As the majority of Black and Latino households begin their adult lives on the least hospitable parts of the wealth pathways, they can expect to experience far more head winds than tail winds. Not only does this suggest strict limits on their prospects, but on those of their children as well. Even more than the disparities in education, occupational status, and household income, the racial wealth gap serves as a reminder of our White supremacist legacy and assures its reproduction into the future.

This paper corroborates the perspective of stratification economics in two other ways as well. As orthodox economics is reluctant to investigate how systems of privilege might function to maintain economic disparities, it must look toward cultural explana- tions as the underlying cause. This paper examines this issue carefully as it includes the full range of attitudes and behaviors that might influence household saving. As the evidence shows, attitudes toward credit, life expectance, risk-taking and leaving a legacy have little impact on household wealth. Even the practice of engaging in regular saving does not produce marked results. Making late payments does lower net worth across both groups, though this could result more from inadequate income than simply poor. Even the choice of taking a longer view in financial planning appears to make little difference. Collectively, all of these household attitudes and behaviors account for only 6% of the racial wealth gap. Clearly, accumulating wealth and gaining financial security is more than a matter of exhibiting financial discipline and Bdoing the right thing^.

Instead, the findings confirm the two household groups face vastly different oppor- tunities. Not only have White households inherited far greater wealth, but they also anticipate even greater family gifts in the future. Family resources explain how White households suffer less from paying off student loans even as they hold college diplomas at nearly twice the rate of Black and Latino households. All three of these family support variables play a significant role in determining net worth among White households. Yet, it does not stop there. Gaining a college education leads to not only higher income, but also stable employment and jobs that yield retirement benefits. The regression results demonstrate each of these plays a significant role in household wealth. Earning an ample income enables households to avoid getting behind on regular bills thereby strengthening their credit score. Together increased family support and improved prospects for household saving can explain the far higher levels of asset ownership enjoyed by White households. Both the income earned and the appreciation gained from these assets enables further wealth accumulation. Indeed, differences in asset ownership and family support explain more than half of the racial wealth gap, except among the upper tail of wealth holders.

The evidence suggests that agency and opportunity, not simply financial discipline and willful behavior, are the key contributors to household wealth. Buoyed by greater family wealth, White households can more effectively leverage the benefits offered by the Household Saving and Asset Appreciation pathways. In contrast, Black and Latino households benefit only marginally from all three pathways, given the much lower levels of wealth in their communities. The unique qualities of wealth, its durability and transferability, ensures that this racial stratification will persist into the future, even in the face of a decline in more overt forms of racial discrimination. Indeed, the racially covert nature of the wealth privilege system makes it a particularly insidious form of racial oppression.

322 Rev Black Polit Econ (2017) 44:303–325

References

Asante-Muhammed, D., Collins, C., Hoxie, J., & Nieves, E. The ever-growing gap: Without change, African- American and Latino families won’t match White wealth for centuries. 2016.

Altonji JG, Doraszelski U. The role of permanent income and demographics in black/white differences in wealth. J Hum Resour. 2005;40(1):1–30.

Avery R, Rendall M. BThe contributions of inheritances to black-white wealth disparities in the United States^, Working Paper 97–08. Ithaca, N.Y: Bronfenbrenner Life Course Center; 1997.

Barsky R, Bound J, Charles KK, Lupton JP. Accounting for the black–white wealth gap: a nonparametric approach. J Am Stat Assoc. 2002;97(459):663–73.

Bates T. The changing nature of minority business: a comparative analysis of Asian, nonminority, and black- owned businesses. Rev Black Polit Econ. 1989;18(2):25–42.

Bertrand, M., & Mullainathan, S. Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination (No. w9873). National Bureau of Economic Research. 2003.

Blanchflower DG, Oswald A, Stutzer A. Latent entrepreneurship across nations. Eur Econ Rev. 2001;45(4): 680–91.

Blau FD, Graham JW. Black-white differences in wealth and asset composition. Q J Econ. 1990;105(2):321– 39.

Borjas, G. J., & Bronars, S. G. Consumer discrimination and self-employment (No. w2627). National Bureau of Economic Research. 1988.

Boshara R, Emmons WR, Noeth B. The demographics of wealth: how age, education and race separate thrivers from strugglers in today’s economy. Louis: Federal Reserve Bank of St; 2015.

Brown T. The intersection and accumulation of racial and gender inequality: Black women’s wealth trajec- tories. Rev Black Polit Econ. 2012;39(2):239–58.

Bucks BK, Kennickell AB, Moore KB. Recent changes in US family finances: Evidence from the 2001 and 2004 Survey of Consumer Finances. Fed Res Bull. 2006;A1:92.

Cagetti M, De Nardi M. Entrepreneurship, frictions, and wealth. J Polit Econ. 2006;114(5):835–70. Carroll C. Why do the rich save so much? In: Slemrod JB, editor. Does Atlas Shrug? Cambridge: The

Economic Consequences of Taxing the Rich. Harvard University Press; 2000. Castaneda A, Díaz-Giménez J, Ríos-Rull JV. Accounting for the US earnings and wealth inequality. J Polit

Econ. 2003;111(4):818–57. Charles KK, Hurst E. The transition to home ownership and the black-white wealth gap. Rev Econ Stat.

2002;84(2):281–97. Chiteji NS. The racial wealth gap and the borrower’s dilemma. J Black Stud. 2010;41(2):351–66. Chiteji NS, Hamilton D. Family connections and the black-white wealth gap among middle-class families.

Rev Black Polit Econ. 2002;30(1):9–28. Chiteji NS, Stafford FP. Asset ownership across generations. Ann Arbor. 2000;1001:48106. Coate D, Vanderhoff J. Race of the homeowner and appreciation of single-family homes in the United States. J

Real Estate Fin Econ. 1993;7(3):205–12. Collins, M., & Dylla, D. Mind the gap: Issues in overcoming the information, income, wealth, and supply gaps

facing potential buyers of affordable homes. Neighborhood Reinvestment Corporation and LISC Center for Homeownership. 2001.

Conley, D. (1999). Being black, living in the red: Race, wealth, and social policy in America. Univ of California Press.

Darity WA, Hamilton D, Stewart JB. A tour de force in understanding intergroup inequality: an introduction to stratification economics. Rev Black Polit Econ. 2015;42(1–2):1–6.

Darity WA, Mason PL, Stewart JB. The economics of identity: the origin and persistence of racial identity norms. J Econ Behav Organ. 2006;60(3):283–305.

Darity WA, Nicholson MJ. Racial wealth inequality and the Black family. African American family life: Ecological and cultural diversity. New York: Guilford Press; 2005. p. 78–85.

De Nardi M. Wealth inequality and intergenerational links. Rev Econ Stud. 2004;71(3):743–68. De Nardi, M. Quantitative models of wealth inequality: a survey (No. w21106). National Bureau of Economic

Research. 2015. DiNardo J, Fortin N, Lemieux T. Labor Market Institutions and the Distribution of Wages. Econometrica.

1996;64:1001–44. Dynan KE, Skinner J, Zeldes SP. Do the rich save more? J Polit Econ. 2004;112(2):397–444. Emmons WR, Ricketts LR. College Is Not Enough: Higher Education Does Not Eliminate Racial and Ethnic

Wealth Gaps. Federal Reserve Bank of St. Louis. Review. 2017;99(1):7–39.

Rev Black Polit Econ (2017) 44:303–325 323

Fairlie RW. The absence of the African-American owned business: An analysis of the dynamics of self- employment. J Labor Econ. 1999;17(1):80–108.

Fairlie RW, Meyer BD. Trends in self-employment among white and black men during the twentieth century. J Hum Resour. 2000:643–69.

Firpo, S., Fortin, N., & Lemieux, T. (2007). Decomposing wage distributions using recentered influence function regressions. University of British Columbia (June).

Flippen C. Unequal returns to housing investments? A study of real housing appreciation among black, white, and Hispanic households. Soc Forces. 2004;82(4):1523–51.

Francis JL. Wealth and the capitalist spirit. J Macroecon. 2009;31(3):394–408. Franklin, R. S. Shadows of race and class. University of Minnesota Press. 1991. Gittleman M, Wolff EN. Racial differences in patterns of wealth accumulation. J Hum Resour. 2004;39(1):

193–227. Heer B. Wealth Distribution and Optimal Inheritance Taxation in Life-cycle Economies with Intergenerational

Transfers. Scand J Econ. 2001;103(3):445–65. Hendricks, L. (2002). Accounting for patterns of wealth inequality. Huggett M. Wealth distribution in life-cycle economies. J Monet Econ. 1996;38(3):469–94. Hurst, E., Luoh, M. C., Stafford, F. P., & Gale, W. G. (1998). The wealth dynamics of American families,

1984-94. Brookings papers on Economic activity, 267-337. Iceland, J., & Weinberg, D. H. (2002). Racial and ethnic residential segregation in the United States 1980–

2000. Bureau of Census. Joint Committee on Taxation (JCT). Estimates of federal tax expenditures for fiscal years 2012–2017. JCS-1-

13. Washington, DC: Government Printing Office; 2013. Juster FT, Smith JP, Stafford F. The measurement and structure of household wealth. Labour Econ. 1999;6(2):

253–75. Keister LA, Moller S. Wealth inequality in the United States. Annu Rev Sociol. 2000;26(1):63–81. Krivo LJ, Kaufman RL. Housing and wealth inequality: Racial-ethnic differences in home equity in the United

States. Demography. 2004a;41(3):585–605. Kochhar, R., & Fry, R. Wealth inequality has widened along racial, ethnic lines since end of Great Recession.

Pew Research Center, 12. 2014. Krivo LJ, Kaufman RL. Housing and wealth inequality: Racial-ethnic differences in home equity in the United

States. Demography. 2004b;41(3):585–605. Krusell P, Smith AA Jr. Income and wealth heterogeneity in the macroeconomy. J Polit Econ. 1998;106(5):

867–96. Laitner, J. Secular changes in wealth inequality and inheritance. Econ J. 2001, 691-721. Long JE, Caudill SB. Racial differences in homeownership and housing wealth, 1970–1986. Econ Inq.

1992;30(1):83–100. Lui M, Robles B, Leondar-Wright B, Brewer R, Adamson R. The color of wealth: the story behind the US

racial divide. New York: The New Press; 2006. Maroto M. Growing farther apart: racial and ethnic inequality in household wealth across the distribution. Soc

Sci. 2016;3:801–24. McKernan, S. M., Ratcliffe, C., Steurle, E., & Zhang, S. Impact of the Great Recession and Beyond. 2014. Menchik PL, Jianakoplos NA. Black-White Wealth Inequality: Is Inheritance the Reason? Econ Inq.

1997;35(2):428–42. Modigliani, F., & Brumberg, R. Utility analysis and the consumption function: An interpretation of cross-

section data. Franco Modigliani. 1954, 1. Mullainathan S, Shafir E. Savings policy and decision-making in low-income households. Insuff Funds:

Savings, Assets, Credit, Banking Among Low-Income Households. 2009;121:129–33. Munnell, A. H., Tootell, G. M., Browne, L. E., & McEneaney, J. Mortgage lending in Boston: Interpreting

HMDA data. Am Econ Rev. 1996, 25-53. Oliver, M. L., & Shapiro, T. M. Black wealth, white wealth: A new perspective on racial inequality. Taylor &

Francis. 2006. Pager D. The Mark of a Criminal Record. Am J Sociol. 2003;108(5):937–75. Scholz, J. K., & Levine, K. US black-white wealth inequality. In K. Neckerman, Social inequality (pp. 895-

929), Russell Sage Foundation. 2004. Shapiro, T. M. The hidden cost of being African American: How wealth perpetuates inequality. Oxford

University Press. 2004. Shin, E. H. Black-White Differences in Wealth Accumulation Among Americans Nearing Retirement. 2010. Smith, J. P. Racial and ethnic differences in wealth in the Health and Retirement Study. J Hum Resour. 1995,

S158-S183.

324 Rev Black Polit Econ (2017) 44:303–325

Smith JP. Healthy Bodies and Thick Wallets: The Dual Relation between Health and Economic Status. J Econ Perspect. 1999;13(2):145–66.

Smith JP. Why is wealth inequality rising? Causes Consequen Increas Inequal. 2001;2:83–116. Tippett, R., Jones-DeWeever, A., Rockeymoore, M., Hamilton, D., & Darity, W. Beyond Broke: Why Closing

the Racial Wealth Gap is a Priority for National Economic Security. Report prepared by Center for Global Policy Solutions and The Research Network on Ethnic and Racial Inequality at Duke University with funds provided by the Ford Foundation. 2014.

Turner, M. A. Discrimination in metropolitan housing markets: National results from phase I of HDS2000. 2002.

Thompson JP, Suarez G. Exploring the racial wealth gap using the survey of consumer finances. Finance and Economics Discussion Series 2015–076. Washington: Board of Governors of the Federal Reserve System; 2015.

Williams RB. The privileges of wealth: rising inequality and a widening racial gap. New York: Routledge; 2017.

Wolff EN. The Accumulation of Household Wealth over the Life-Cycle: A Microdata Analysis*. Rev Income Wealth. 1981;27(1):75–96.

Wolfson, M. The causes of inequality in the distribution of wealth: A simulation analysis. Ph.D. Thesis, Cambridge University. 1977.

Rev Black Polit Econ (2017) 44:303–325 325

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.