The Status of Public Sector Pay Equity for Women of Color in the United States

Abstract

Research by Reese and Warner and Llorens, Wenger, and Kellough assesses the relative pay of women in the United States. However, research on the relative pay of women of color remains scant. What kinds of factors predict relative pay for women of color, and are they the same as for White women? The author utilizes ordinary least squares (OLS) regression on an Equal Employment Opportunity Commission (EEOC) panel data set on public sector employment by state to analyze the pay of Black, Hispanic, Asian, and American Indian women relative to men for 2005-2013. The author reaffirms that whether a woman lives in a state that has implemented a major gender pay equity measure is a significant factor determining her relative pay. Furthermore, the intersectional nature of public sector pay for women of color is numerically verified. Women fare better in states where they are descriptively represented in terms of gender and race/ethnic group.

Keywords

gender pay equity equal pay state and local government equal pay for women

Introduction

The purpose of this research is to begin to assess the condition of pay equity for women of color. Pay equity for women was an issue about which both major U.S. presidential candidates spoke in 2016 (www.pbs.org.). The Women’s march that was held worldwide on January 21, 2017, emphasized pay equity, among such other issues as domestic violence, reproductive rights, and immigration reform. President Obama’s first act in office back in 2009 was to sign the Lilly Ledbetter Fair Pay Act into law, indicating the significance he placed on the issue of pay fairness. The issue of gender pay equity is salient, and publicly acknowledged, but little attention has been given to the status of pay equity for women of color.

Reese and Warner (2012) show that women in the public sector are paid more fairly compared with men than women in the private sector, that a significant predictor of women’s pay is whether the state where she lives has enacted a pay equity measure or not, and that women in the southern part of the United States are paid more fairly in relative terms than others. This study seeks to replicate and extend that study to encompass an analysis of the wage gaps for women of color; specifically, I wish to explore and further analyze wage gaps for African American, Hispanic, Asian, and American Indian¹ women employed in the subnational public sector in the United States.

The pay situation for women of color is often relegated to the proverbial footnote in wage studies. Independent research groups, such as the (now defunct) Women of Color Policy Network at New York University (NYU) Wagner and the National Committee on Pay Equity publish policy briefs on the subject. There has been some academic work in the area as well, notably Cornwell and Kellough’s (1994) analysis of federal pay; however, I am unaware of any recent similar work on state and local government pay equity for women of color. Therefore, my central research questions revolve around the status of public sector pay for women of color in the United States. More specifically, what factors can be identified that predict public nonfederal pay for women of color?

Literature Review

Why Do Women of Color Earn Less Money?

Many of the reasons women of color—apart from Asian women, that is, who sometimes earn more than males—may make less money are the same reasons that women in general make less money. For example, for many years it was true that women in general did not earn wages equal to those of men because they lacked the similar requisite qualifications, particularly education (Alkadry & Tower, 2006; Kelly, 1991). In 1970, only 11.2% of women were college graduates; by 2010, this figure had risen to 36.4% (U.S. Bureau of the Census, National Center for Education Statistics, 2016). Six percent more males than females had college degrees in 1980, but by 2010, 32% of women had college degrees, while only 24% of men do (U.S. Bureau of the Census, National Center for Education Statistics, 2016). The educational attributes of women of color have similarly improved in recent decades. In 1970, 4.3% of Hispanic females and 4.6% of Black females had college degrees; by 2010, these figures had risen to 14.9% and 21.4%, respectively (U.S. Bureau of the Census, National Center for Education Statistics, 2016).

Also, women often leave the workforce to have children (Correll, Benard, & Paik, 2007), are positionally segregated (Alkadry & Tower, 2011; Cornwell & Kellough, 1994; Guy, 1993; Kelly, 1991; Newman, 1993; Reid, Miller, & Kerr, 2004), and are occupationally segregated (Dey & Hill, 2007; Dinovitzer, Reichman, & Sterling, 2009). Positional segregation is the idea that women are often lacking in top—and better-paying—managerial positions within an organization. Occupational segregation is the idea that many so-called “women’s jobs” have historically not been evaluated as contributing as much as “men’s jobs” to organizational success. Jobs with such “male” attributes as strength, extensive travel, and so on, were once thought of as inappropriate or even impossible for women. At the same time, the male breadwinner mentality helped ensure that favorable wage treatment was given to men as heads of household. What is often referred to as occupational choice for women may not turn out to have had much of a voluntary element at all, as women were socialized into a few, generally low-paying sorts of positions such as teaching, sewing, and child care. Furthermore, women are often also segregated by agency, which is an extension of occupational segregation; this means that women are often relegated to jobs in “women’s agencies,” or those with more caring, nurturing missions (see Alkadry & Tower, 2011).

Women may also have historically earned less than men because of such factors as their lack of politically descriptive representation; in other words, women account for only 17% to 28% of state legislators and 10% to 20% of governors at any point in time although they are approximately 51% of the national population. Thus, part of the reason that pay equity laws that might help ensure wage equality are not passed and/or enforced may derive from the fact that so few women have been in political power (Bratton & Haynie, 1999; Carroll, 2004; Hogan, 2008; Orey, Smooth, Adams, & Harris-Clark, 2006; Reese & Warner, 2012; Reese, In press). The idea of descriptive representation is that one is best served by a political representative most like oneself, as pertains to gender.

By the same token, the descriptive representation literature implies that we should be able to expect behavior that is different from legislators who represent racial/ethnic minorities (Gay, 2002; Pantoja & Segura, 2003; Pitkin, 1967; Rocha, Tolbert, Bowen, & Clark, 2010; Schroedel & Aslanian, 2017). To the degree that politically descriptive representation matters, it does not seem too far of a reach to presume that minorities living in states with descriptive representation might fare better wage-wise than not. However, while all of the factors just mentioned may affect the relative pay for women of color as well as women in general, there are factors at work that are only experienced by women of color.

Intersectionality

Pettit and Ewert (2009) analyze the employment gains of Black women’s relative wages and report declines in Black women’s relative wages since the mid-1990s. Lapidus and Figart (1998) review pay equity by race and gender and find that being an African American or Hispanic woman significantly negatively affects pay relative to men’s. However, neither of these findings specifically addresses the multiplicative impacts of being both a woman and being a woman of color.

Why would women of color face additional obstacles to pay equity? At the very least, the literature implies that women of color would be descriptively underrepresented even if White women were in power: Descriptive representation is important not only for women in general but also of special and more specific import for women of particular racial and ethnic identities (Lien, 2015). The literature further suggests that this is due to discrimination of a sort that is in addition to that experienced by women in general in the workforce (Breslin, Pandey, & Riccucci, 2017).

We know that racial discrimination occurs in the workplace, despite its being illegal in the United States per such major laws as the Equal Pay Act of 1963 and Title VII of the Civil Rights Act of 1964. We know this both anecdotally and through qualitative research (see Bobo & Fox, 2003; Deitch et al., 2003; Paradies, 2006) and by the visible results of unequal pay (Reese & Warner, 2012). In addition, it is well established in the literature that racial discrimination does not exist in a vacuum. As Collins and Bilge (2016) state,

. . . we all think we are playing on a level playing field when we are not. The cultural domain of power helps manufacture messages that playing fields are level, that all competitions are fair, and that any resulting patterns of winners and losers have been fairly accomplished. (p. 11)

They then go on to develop their thesis of intersectionality more thoroughly: “Within intersectional frameworks, there is no pure racism or sexism. Rather, power relations of racism and sexism gain meaning in relation to one another” (Collins & Bilge, 2016, p. 27). Thus, we know that the effects of race and sex are interactive or multiplicative (see Crenshaw, 1989). This is primarily established thus far through qualitative research.

In sum, women in general may have lower wages than men because of occupational segregation, agency segregation, position segregation, lack of politically descriptive representation, leaving the workforce to have children, relegation to lower level jobs, and, previously, lack of relevant education and experience. Moreover, women of color experience not only all of these factors but also an additional discriminatory penalty, intentional or not, measured in the past with qualitative data.

State-Level Action on Pay Equity

Reese and Warner (2012) conducted an email survey to determine which states have passed laws related to gender pay equity or otherwise implemented related measures. They did not find any states that implemented pay measures specifically for any subset of women; that is, there were no minority pay equity measures discovered. Overall, they discovered that eight states—Connecticut, Iowa, Massachusetts, Minnesota, New Mexico, Oregon, Vermont, and Washington—have implemented an adjustment of 2% of total payroll or more on behalf of their female employees. One question asked was whether female employees in the states that had implemented pay equity measures enjoyed significantly better pay than those in states that had not implemented said measures. The answer to the question was unequivocally “yes.” This study differs from that of Reese and Warner (2012) in that this study attempts to predict income for different racial/ethnic groups, while the prior study looked at women on the whole.

Preliminary Analysis

Two central and preliminary questions present themselves at the outset of this inquiry. First, using Reese and Warner (2012) as a model, can we say that minority women have experienced pay equity the same as women overall? Specifically, do all racial/ethnic groups experience a boon due to state-level pay intervention? My second question is whether there does indeed appear to be a wage penalty associated with both being a woman and being a racial/ethnic minority.

Therefore, my central questions include the following hypothesis:

Hypothesis 1: There is no difference between (White, Black, Hispanic, Asian, American Indian) women’s pay relative to men’s in states with and without major pay equity programs for state employees.

I use EEO-4 data, which are state and local government reports collected biennially by the Equal Employment Opportunity Commission (EEOC), to calculate public sector wage gaps by state by racial/ethnic group; then the relative wages representing this gap are my dependent variables. Then divide the states into two groups: those that have had a major pay equity adjustment and those that have not. As Table 1 shows, differences in mean pay by racial/ethnic group of women is statistically significant for Black, Asian, and American Indian women. For White women and Hispanic women, the differences are in the expected direction, but are not statistically significant.

Table 1.

The t Test: Major Female Racial/Ethnic Breakdown by Whether a State Has Had a Major Pay Equity Adjustment or Not, 2005-2013.

Adjustment	White	Black	Hispanic	Asian	American Indian
No (164)	88.44 (7.04)	80.84 (5.29)	79.19 (7.54)	97.05 (13.43)	83.50 (7.83)
Yes (82)	90.52 (7.41)	85.50 (5.03)	81.30 (6.99)	92.02 (10.10)	86.53 (6.75)
T score	−1.52	−4.60**	−1.47	2.01**	−2.04**

Note. Standard deviations are reported in parentheses. Numbers in parentheses under “adjustment” column are numbers of observations.

significant at the .05 level.

Table 2 shows data to help shed some light on the second preliminary question; not only is the median pay of various racial/ethnic groups substantially less than that of White men but also minority women make less on average than their racial/ethnic male counterparts. In other words, the effects of racial discrimination and sexual discrimination together account for a greater effect—despite difficulties in quantitative measurement as currently construed—than either one would on its own. From the last column of the table, we can see that for every racial/ethnic group, for every year, women are paid less than their male counterparts. For example, in 2014 Black women earned a median income of US$40,438 or 75% of the White/male gold standard for public sector pay. For that same year, Black women also only earned 90% of the median wage of their Black male counterparts. Thus, we have verified numerically the intersectional nature of the wage penalty.

Table 2.

Median Salary by Race/Ethnicity and Gender, for 2009, 2011, and 2013.

Year	Gender/race or ethnicity	Median salary	% of W/M salary	Gender/race or ethnicity	Median salary	% of W/M salary	% of relevant M/Reference
2009	W/M	US$51,519	100	W/F	US$42,236	82	82
	B/M	42,691	83	B/F	38,550	75	90
	H/M	49,499	96	H/F	40,648	79	82
	A/M	61,525	119	A/F	54,217	105	88
	I/M	47,812	93	I/F	40,739	79	85
	Tot/M	50,254		Tot/F	41,474	81	83
2011	W/M	US$52,844	100	W/F	US$40,037	83	83
	B/M	44,176	84	B/F	40,297	76	91
	H/M	51,758	98	H/F	42,344	80	82
	A/M	64,125	121	A/F	57,269	108	89
	I/M	48,527	92	I/F	38,481	73	79
	Tot/M	51,782		Tot/F	43,137	82	83
2013	W/M	US$53,673	100	W/F	US$44,958	84	84
	B/M	44,822	84	B/F	40,438	75	90
	H/M	51,733	96	H/F	42,169	79	82
	A/M	66,489	124	A/F	56,698	106	85
	I/M	48,781	91	I/F	45,107	84	92
	Tot/M	52,539		Tot/F	43,838	82	83

Source. Data derived from EEO-4 reports for 2009, 2011, and 2013.

Note. W = White; B = Black; H = Hispanic; A = Asian; I = Indian (American Indian); Tot = Total; M = male; F = female. Relevant M/Reference means, for example, for B/F, that B/F make on average 90% of what B/M earn.

Next, we turn to the multivariate analysis of the wage structure, by which we hope to assess factors predicting women’s public sector wages by racial/ethnic group (the dependent variable).

Primary Analysis

Data

I chose several control variables as previously utilized by Reese and Warner (2012). Specifically, I control for several political, demographic, and union-related factors. Political variables include whether the governor of a state is a woman (lagged 2 years), the percentage of women in the state legislature (lagged 2 years), the percentage of self-identified Democrats in a state, the political culture of a state, and an index of the well-being of women by state.² Demographic controls are per capita personal income (PCPI; natural logarithm thereof), the male-to-female population ratio, and the percentage of the state population that is, variously, White, Black, Hispanic, Asian, or American Indian.³ The labor union-related control variable is whether the employees of a state have collective bargaining rights or not. These independent variables are further explicated below.

Political controls

I expect whether a state has had a woman governor and/or a higher percentage of female state legislatures to have a positive effect on female wages of all ethnic variations because of the descriptive representation literature, which shows that women in the legislature are more apt to introduce and pass woman-related legislation, among which pay equity would count (see Bratton & Haynie, 1999). I expect that the percentage of self-identified Democrats would be positively related to pay increases because the issue of pay equity appears to be more on that political party’s agenda than the Republican’s (Henderson, 2014). I anticipate that the relationship of pay to political culture, which is determined, according to Daniel Elazar, by the initial immigration patterns to different regions of the country, will be positive insofar as both his moralistic and individualistic states are generally expected to adopt more progressive policies than those in the traditionalistic, or southern, part of the country (Elazar, 1972; Mead, 2004). Finally, a state’s position in a women’s well-being index is included as a control variable because this may indicate a higher level of concern for issues pertinent to women.

Demographic controls

I expect that the PCPI of a state will be positively related to pay fairness; a state first has to have a sufficient income to pay people better (Reese & Warner, 2012). I also propose that states with relatively more women in their populations may be more likely to pay attention to women’s pay. The ratio is expressed as male-to-female, so I will expect the sign to be negative on this variable. Finally, I posit that the more of a particular racial or ethnic group is present in the state population, the greater the chance that the government will implement some sort of pay equity measure on behalf of those citizens. Thus, I propose to include percentage of the population that is Hispanic as a control in the Hispanic regression, percentage of the population that is Black in the Black regression, and so on.

Union control

I include the union variable, a categorical representation of whether a state’s employees have the right to collectively bargain, because I expect that the right to bargain would have a positive impact on pay for all groups of women, as suggested by the literature (Gardner & Daniel, 1998; Portman, Grune, & Johnson, 1984). In this case, states with collective bargaining rights in three major areas—firefighting, education, and police—are rated “3,” those with collective bargaining rights in only two areas are rated “2,” those with collective bargaining rights in only one area are rated “1,” and those with no collective bargaining rights are rated “0.”

My resultant grand hypothesis is as follows:

Hypothesis 2: Parameter estimates for the female-to-male pay ratio for each ethnic group (White, Black, Hispanic, Asian, American Indian) of women will be positive and significant for whether states (a) have had a major pay equity adjustment; (b) had a woman governor; (c) had a higher percentage of women in the legislature; (d) had a higher percentage of Democrats; (e) had a moralistic political culture; (f) had an individualistic political culture; (g) had a higher position in an index of women’s well-being; (h) had a higher PCPI (natural logarithm thereof); (i) has the right to collectively bargain; (j) has a higher percentage of the population that is White, Black, Hispanic, Asian, or American Indian (respectively) will all positively affect the female-to-male pay ratio for each ethnic group of women; and, conversely, (k) the greater the ratio of male-to-female population by state will have a negative effect on the female-to-male pay ratio.

Data Sources

I use EEOC data from the biennial EEO-4 reports entitled Job Patterns for Minorities and Women for State and Local Government for the years 2005, 2009, 2011, and 2013,⁴ which are the only years presently available from the agency, for the female relative wage data by racial/ethnic group by state. Data for political variables were obtained from sources as follows: data for woman governors is from the Center for American Women and Politics at Rutgers, data for women in state legislatures is from the National Conference of State Legislature, data for percent Democratic is from the Gallup organization (2017), data for political culture is from Elazar (1972), and data for the women’s index is from wallethub.com. Regarding demographic controls, data were obtained from these sources: PCPI and male-to-female population ratios, as well as percentage of ethnic distribution, are all from the U.S. Department of the Census. Finally, the union control variable is from the Center for Economic and Policy Research (Sanes & Schmitt, 2014).

Method

Ordinary least squares (OLS) regression analyses were performed, using the xtpcse command in Stata, for the panel that includes 2005, 2009, 2011, and 2013, between the dependent variable for each racial/ethnic group, which is their relative wage, expressed in terms of a White man’s US$1.00, and the independent control variables discussed. Although the advantages of panel regression are great because one can look at data over time, there are a few potential disadvantages (Hsiao, 2003). Specifically, panel data often demonstrate cross-case autocorrelation and heteroscedasticity. The xtpcse, or linear regression with panel-correction standard errors, is a more conservative choice overall for presentation of panel regression results. Multicollinearity is accounted for via examination of variance inflation factors (VIFs) and is not a significant issue with these models. No outliers were detected. Overall, no violations of linearity, normality, or homoscedasticity were indicated.

Results

The regressions yield strong, if somewhat mixed, results. Table 3 reports the overall panel regression results for White, Black, Hispanic, Asian, and American Indian females. The R² figures for the equations are .53, .26, .31, .58, and .36, respectively. Clearly, I can presently do a better job predicting pay for White and Asian females than for Black, Hispanic, or American Indian females. To reiterate, the dependent variable, or that which we seek to predict, is the relative wage of the racial/ethnic group of women being analyzed; I am trying to predict their wages relative to those of a White man. The median wages of the White man are set at US$1.00.

Table 3.

Regression: Determinants of Gender Pay Equity for Female State and Local Government Employees, Panel Analysis, 2005-2013.

Variable	White	Black	Hispanic	Asian	American Indian
Constant	103.12(15.79)	171.07(36.97)	118.34(20.36)	204.57(39.37)	121.23(33.37)
Log of PCPI	7.92*** (1.99)	−4.63(3.69)	5.11** (2.60)	13.93*** (3.47)	0.55(2.81)
Governor (1 = woman)	−1.63*** (0.51)	−1.83*** (0.65)	−1.55** (0.71)	−0.49(1.05)	−2.64*** (1.07)
Legislature (% women)	0.12*** (0.03)	−0.07* (0.04)	0.17** (0.09)	0.28*** (0.08)	−0.21*** (0.05)
Major pay adjustment	4.21*** (0.42)	5.76*** (0.46)	3.81*** (0.79)	−0.33(0.82)	5.68*** (0.90)
Democrat (%)	−5.15* (3.18)	8.27(6.04)	0.05(7.71)	−8.95** (4.47)	2.79(5.13)
Individualistic political culture	−9.06*** (0.53)	−2.00*** (0.70)	−4.23*** (0.62)	−5.59*** (0.90)	−3.85*** (1.15)
Moralistic political culture	−3.96*** (0.36)	−2.19*** (0.88)	−3.97*** (0.72)	−4.81*** (0.88)	−2.22*** (0.53)
Women’s well-being	−0.25*** (0.02)	−0.02(0.05)	−0.13*** (0.03)	−0.39*** (0.06)	0.01(0.03)
Collective bargaining	−0.89*** (0.14)	−0.23(0.18)	−0.74** (0.36)	−4.09*** (0.26)	−2.42*** (0.33)
Male: Female population	−0.81*** (0.09)	−0.40*** (0.11)	−0.86*** (0.10)	−2.27*** (0.12)	−0.31* (0.17)
Reference population	−0.01(0.01)	−0.07*** (0.03)	−0.19*** (0.03)	−0.58*** (0.18)	−0.59*** (0.15)
R ²	.53	.26	.31	.58	.36
No. observations	196

Note. Standard errors are reported in parentheses. PCPI = per capita personal income.

, **, *** indicate significance at the .10, .05, and .01 level, respectively.

All variables but one (percentage of the population that is White) are statistically significant for the equation with which I attempt to predict the White women’s wages relative to White men’s. Specifically, the state’s PCPI, whether the state government has implemented a major equity adjustment, whether women live in a traditionalistic state (because both individualistic and moralistic show negative here), and the ratio of women to men in the population all are in the anticipated direction and all are statistically significant. Women per this result actually fare worse with their relative pay when a woman has been governor, and the women’s well-being index also is not a good predictor of women’s pay.

For Black women, the R² is much lower at .26. Whether a state has had a major pay adjustment is a good predictor of relative pay, as is political culture, and the proportion of females in the population. However, the following variables were all either in the wrong direction or statistically insignificant: PCPI, whether a woman was governor, the percentage of the legislature that is female, percentage of the population that is a Democrat, the well-being index, and whether a state’s workforce has collective bargaining rights, as well as the percentage of the population that is Black, all are either in the wrong direction or statistically insignificant. Whether a state has had a major pay adjustment or not is still a good predictor, as is political culture, and the proportion of females in the population.

The equation for Hispanic women yields stronger results, with an R² of .31; all variables but one are statistically significant, the percentage of the population that identifies with the Democratic party. Thus, PCPI, percentage of the legislature that is female, whether the state has had a major pay adjustment, political culture, the male-to-female population ratio, all have a statistically significant effect in the predicted direction. However, whether there was a woman governor, the well-being index, collective bargaining, and the percentage of Hispanics in the population are all in the nonpredicted direction.

The Asian female equation yields a .58 R², the best of all the racial/ethnic groups of women I am analyzing. All variables are significant predictors, except for whether a woman was a governor and whether there was a major pay adjustment in the state. Again, the percentage of Democrats in the state was not a predictor in the anticipated direction, nor was the proportion of the state population that is Asian.

Finally, the American Indian equation is at a .36 R² and all predictors are significant but PCPI, percent Democratic population, and the women’s well-being index. Whether a woman was a governor is significant in the wrong direction, as is the percentage of American Indians in the population.

The consistently insignificant results on the percentage of Democrats in the population led me to estimate another regression equation. Specifically, my second iteration, shown in Table 4, shows the results of an equation where I added a variable to represent the percentage of the relevant racial/ethnic population in the state legislature. This improved all equations except that for Black women.

Table 4.

Regression: Determinants of Gender Pay Equity for Female State and Local Government Employees, Panel Analysis, 2005-2013.

Variable	White	Black	Hispanic	Asian	American Indian
Constant	79.62(13.36)	168.59(36.74)	89.41(13.71)	208.43(39.60)	123.65(36.07)
Log of PCPI	9.27*** (2.05)	−4.52(3.68)	9.48*** (1.79)	14.21*** (3.50)	0.42(2.97)
Governor (1 = woman)	−1.69*** (0.55)	−1.67** (0.69)	−2.87*** (0.51)	−0.72(1.07)	−2.61* (1.08)
Legislature (% women)	0.04(0.03)	−0.08* (0.05)	0.15* (0.08)	0.30*** (0.08)	−0.21*** (0.05)
Major pay adjustment	4.17*** (0.45)	5.79*** (0.47)	2.09*** (0.65)	−0.18(0.90)	5.66*** (0.90)
Democrat (%)	−3.785(3.29)	8.27(6.07)	−2.31(8.26)	−9.20* (4.46)	2.78(5.15)
Individualistic political culture	−9.85*** (0.51)	−2.40*** (0.84)	−2.14*** (0.61)	−5.00*** (0.71)	−3.84*** (1.15)
Moralistic political culture	−4.58*** (0.46)	−2.50*** (0.95)	−0.94(0.85)	−4.98*** (0.83)	−2.21*** (0.52)
Women’s well-being	−0.19*** (0.01)	−0.01(0.05)	−0.23*** (0.03)	−0.39*** (0.06)	0.00(0.03)
Collective bargaining	−0.96*** (0.15)	−0.11(0.22)	−2.19*** (0.33)	−4.23*** (0.31)	−2.42*** (0.34)
Male: Female population	−0.58*** (0.08)	−0.40*** (0.11)	−0.90*** (0.11)	−2.34*** (0.15)	−0.33* (0.17)
Reference population	0.25*** (0.02)	−0.14*** (0.04)	−1.15*** (0.08)	−0.90*** (0.28)	−0.55*** (0.17)
Reference legislature	−0.39(0.03)	0.10* (0.05)	1.43(0.08)	0.56* (0.25)	−0.10(0.17)
R ²	.58	.26	.44	.58	.36
No. observations	196

Note. Standard errors are reported in parentheses. PCPI = per capita personal income.

, **, *** indicate significance at the .10, .05, and .01 level, respectively.

We subsequently estimated equations for all five racial/ethnic groups utilizing an interactive term between race/ethnicity in the legislature and percent women in the legislature. This equation, not shown here, proved a slightly better fit; R²s for all racial/ethnic groups to be .59, .28, .45, .59, and .37, respectively. This interactive term was created to attempt to capture the race-gender intersectionality of state legislators.

Finally, I estimated one further equation in which we improved all R² figures except for Black women. In this iteration, also not shown, I added a variable representing party control of the state legislature. For this equation, R²s were .60, .26, .43, .58, and .39, respectively.

Discussion

In effect, White women who live in states that have had a pay equity measure implemented in the past two decades should enjoy an increase to their pay on scale relative to White men, as should Black women, Hispanic women, and American Indian women. Asian women are the only group for whom living in a state that has passed a gender pay equity law has no significant effect. Why might this be? I believe that, given that Asian women apparently experience the pay gap quite differently than other racial/ethnic minorities, Asian women do so well already that they do not “need” the boost from a major pay adjustment. They already make more money than White men, on average, and nearly as much as Asian men. They do not, in this way, appear to be victims of discrimination in such a way that affects their wages negatively. This research supports the prior findings of Reese and Warner (2012).

The PCPI of a state is a potent predictor of women’s relative pay for White women, Hispanic women, and Asian women, but not for Black women or American Indian women. The prior work of Reese and Warner (2012) suggested that PCPI would be a strong predictor of all women’s relative wages. Perhaps the negative finding for Black women and American Indian women might be because more Black women and American Indian women live in relatively poorer states; these states also pay women better on a relative scale (at least in the public sector) than other states do, despite being poor. (Note that paying more fairly in a relative sense does not preclude overall pay still being lower than in other regions.)

The effects of having a woman governor are, for the most part, either fairly negligible or insignificant; however, this may change over time, as more women come into office. As for the percentage of women in the state legislature, this measure is only just beginning to show effects, although these may be mitigated in the future as it appears that the percentage of women in the state legislatures is stagnating. In essence, this research at this time only partially supports the importance of descriptive representation as suggested by Hogan (2008), Carroll (2004), Orey et al. (2006), and Bratton and Haynie (1999). However, it still may be the case that a critical mass of women has not yet been in statewide office long enough to accurately gauge their impact.

The percentage of Democrats in the state is not effective for any racial/ethnic group in a statistically significant way. The subsequently designed percentage of minority state legislators is a better measure, and the party control of the state legislatures is effective as well. I theorize that this may be so because the percentage of the state legislature that is a racial minority is more directly tied to the idea of politically descriptive representation, as stated in the literature.

Counter to the literature as exemplified by Mead (2004) and supportive of the work of Reese and Warner (2012), these results affirm the significance of political culture. Specifically, as shown by Reese and Warner (2012) previously, women living in traditionalistic—basically southern—states are paid more fairly on a relative basis than those living in the allegedly progressive regions of the country exemplified by Elazar’s moralistic and individualistic regions. It should be noted, however, that just as with the findings of Reese and Warner (2012), the political culture variables are actually predictive in a way that is counter to rest of the literature. The literature suggests that the more progressive states, or those in moralistic and individualistic regions of the nation, will probably pay women better. However, at least in relative terms, this is not the case; rather, traditionalistic (mainly southern) states pay women in the public sector better than states primarily identified as being of the other two political cultures. Why might this be, particularly since these states also tend to be the poorest in the nation? I can only surmise that the states that initially fell under the supervision of the federal government under the Civil Rights Act of 1965 perhaps were inculcated with such a respect for fair treatment that they carried it over into their state and local government classification and compensation systems.

The women’s well-being index is not an effective predictor in any case. The measure may simply not be valid.⁵ It may also be the case that women are not able to be represented in a monolithic, one-fits-all manner. However, the male-to-female population ratio is significant for every ethnic group, meaning that the more women there are proportionally in the state population, the higher the relative median wage tends to be.

The percentage of relevant racial/ethnic population in the state where one lives would seem to be predictive from the perspective of descriptive representation, wherein one is best served by those whom one is “like.” And indeed, the ethnic population is significant for every group but White women. This is probably the case because percent White is virtually the same as “percent majority”—the measure does not convey difference of representation in the same way that percent minority does.

One of the most interesting and curious findings is regarding state collective bargaining status. One would expect that states whose employees have collective bargaining rights would have better paid women in their public service. For one thing, women are only slightly less likely than men to belong to a public sector union, and Blacks are more likely to belong than Whites. However, this is most definitely not the case for most groups in this study; in fact, the opposite appears to be the case. This finding seemingly is related to the one for political culture; that is, most states that are right-to-work states are southern, as are most in the traditionalistic region. These two factors seem to reflect a new-found uniqueness we can associate with southern states: they do not allow collective bargaining, but even in the absence of those rights, and as mentioned previously, they are also poor, but their minority employees are paid better in relative terms than those living elsewhere. (Note that the employees living in moralistic or individualistic cultures may very well still be paid more in absolute dollars.)

Finally, my use of an interactive term between race/ethnicity in the legislature and percentage of women in the legislature was my effort to capture the gender-race intersectionality of state legislators. Per the literature (e.g., Collins & Bilge, 2016; Crenshaw, 1989), there does appear to be an effect of the interaction between race and gender that is greater than the sum of its parts.

Conclusion

This study builds upon prior research by analyzing the relative gender pay equity of different racial/ethnic groups in the United States; specifically, relative wages in the public sector of Black, Hispanic, Asian, and American Indian women are examined. I find that whether a state has implemented a major gender pay equity measure is a significant predictor of pay for all groups except Asian women, who actually do better than the average male wage-wise as a group in some states and so may not need the major gender pay equity measures, as such. In addition, the intersectional nature of wage discrimination is verified and the importance of descriptive representation, both in gender and racial/ethnic terms, is confirmed. Furthermore, to ascertain the effects of the intersectionality of gender and race, I show that the financial well-being of women can be positively affected by electing women who are racial/ethnic minorities.

Limitations of the study include that it is only exploratory and does not fully explain everything about women’s relative wages. In particular, several of the key independent variables do not help explain relative wages for Black women. Why might this be the case? Clearly, further research needs to be done for Black women in particular, utilizing additional variables of unknown nature. One possibility is that while Black women have made vast educational strides in recent years, their college degree rates (14.9% for Black women; 21.4% for Hispanic women) are still below those of women on the whole (32%) or men overall (24%). Thus, these groups could still be experiencing a form of education-related employment disadvantage that White and Asian women have already overcome.

As the United States potentially moves into a postaffirmative action era, this research can be important when both private sector CEOs and public leaders contemplate their organizations’ racial/ethnic diversity—or lack thereof—and related organizational climates of inclusiveness. It is said that the drive for employment diversity, which may be defined as striving on a voluntary basis to hire varied people to either make more profit (in the case of the private sector) or remain politically responsive (in the case of the public sector) is the gist of modern-era employment. It is difficult to have confidence in the importance of employment diversity if all parties are not being invited to the table on an equitable basis.

I conducted this research not to highlight a negative so much as to emphasize an opportunity. The potential bonuses that are projected to occur for our (or any) economy, if the talents of all people are fairly recognized and paid, are astounding. The great underdeveloped promise of equality has, moreover, been actually realized in the eight key states mentioned previously⁶ that have implemented gender-based pay equity policies involving the expenditure of 2% or greater of their total payroll costs. In the case of every group of women of color but Asian women, the implementation of a gender-based pay adjustment has a significant effect on women’s average pay compared with men’s. Thus, the states have it in their power to change the lives of women and their families dramatically for the better, and at a comparatively low cost.

President John F. Kennedy said on June 10, 1963, the same day he signed the Equal Pay Act into law, “If we cannot end now our differences, at least we can help make the world safe for diversity” (Kennedy, 1963). Over half a century later, has his point been realized yet?

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Catherine C. Reese is professor and master of public administration (MPA) director at Arkansas State University. Her research has appeared in the Review of Public Personnel Administration, Public Administration Review, Politics & Policy, Public Budgeting & Finance, and the Handbook of Human Resource Management in Government.

References

Alkadry

M. G.

Tower

L. E.

(2006). Unequal pay: The role of gender. Public Administration Review, 66, 888-898.

Alkadry

M. G.

Tower

L. E.

(2011). Covert pay discrimination: How authority predicts pay differences between women and men. Public Administration Review, 71, 740-750.

Bobo

L. D.

Fox

(2003). Race, racism, and discrimination: Bridging problems, methods, and theory in social psychological research. Social Psychology Quarterly, 66, 319-332.

Bratton

K. A.

Haynie

K. L.

(1999). Agenda setting and legislative success in state legislatures: The effects of gender and race. The Journal of Politics, 61, 658-679.

Breslin

R. A.

Pandey

Riccucci

N. M.

(2017). Intersectionality in public leadership research: A review and future research agenda. Review of Public Personnel Administration, 37, 160-182.

Carroll

S. J.

(2004). Women in state government: Historical overview and current trends. In The Book of the States 2004 (pp. 471-478). Lexington, KY: Council of State Governments.

Collins

P. H.

Bilge

(2016). Intersectionality. Malden, MA: Polity Press.

Cornwell

Kellough

J. E.

(1994). Women and minorities in federal government agencies: Examining new evidence from panel data. Public Administration Review, 54, 265-269.

Correll

Benard

Paik

(2007). Getting a job: Is there a motherhood penalty? American Journal of Sociology, 112, 1297-1388.

10.

Crenshaw

(1989). Demarginalizing the intersection of race and sex: A Black feminist critique of antidiscrimination doctrine, feminist theory and antiracist politics. University of Chicago Legal Forum, 1989, Article 8.

11.

Deitch

E. A.

Barsky

Butz

R. M.

Chan

Brief

A. P.

Bradley

J. C.

(2003). Subtle yet significant: The existence and impact of everyday racial discrimination in the workplace. Human Relations, 56, 1299-1324.

12.

Dey

J. G.

Hill

(2007). Behind the pay gap. Washington, DC: American Association of University Women.

13.

Dinovitzer

Reichman

Sterling

J. S.

(2009). The differential valuation of women’s work: A new look at the gender gap in lawyers’ incomes. Social Forces, 88, 819-864.

14.

Elazar

D. J.

(1972). American federalism: A view from the states (2nd ed.). New York, NY: Harper & Row.

15.

Gallup. (2017). “GOP Maintains Edge in State Party Affiliation in 2016.” Retrieved from news.gallup.com/poll/203117/gop-maintains-edge-state-party-affiliation-2016:aspx

16.

Gardner

S. E.

Daniel

(1998). Implementing comparable worth/pay equity: Experiences of cutting-edge states. Public Personnel Management, 27, 475-489.

17.

Gay

(2002). Spirals of trust? The effect of descriptive representation on the relationship between citizens and their government. American Journal of Political Science, 46, 717-732.

18.

Guy

M. E.

(1993). Three steps forward, two steps backward: The status of women’s integration into public management. Public Administration Review, 53, 285-292.

19.

Henderson

(2014, January 29). Obama, Democrats put spotlight on gender pay gap. Will It Matter? The Washington Post.

20.

Hogan

R. E.

(2008, December). Sex and the statehouse: The effects of gender on legislative roll-call voting. Social Science Quarterly, 89, 955-968.

21.

Hsiao

(2003). Analysis of panel data (2nd ed.). New York, NY: Cambridge University Press.

22.

Kelly

R. M.

(1991). The gendered economy: Work, careers, and success. Newbury Park, CA: SAGE.

23.

Kennedy

J. F.

(1963). Commencement address at American University, June 10, 1963. Retrieved from http://www.jfklibrary.org/Asset-Viewer/BWC7I4C9QUmLG9J6I8oy8w.aspx

24.

Lapidus

J. F.

Figart

D. M.

(1998). Remedying “unfair acts”: U.S. pay equity by race and gender. Feminist Economics, 4(3), 7-28.

25.

Lien

(2015). Reassessing descriptive representation by women and men of color: New evidence at the subnational level. Urban Affairs Review, 51, 239-262.

26.

Llorens

J. L.

Wenger

J. B.

Kellough

J. E.

(2008). Choosing public sector employment: The impact of wages on the representation of women and minorities in state bureaucracies. Journal of Public Administration Research and Theory, 18, 397-413.

27.

Mead

L. M.

(2004). State political culture and welfare reform. The Policy Studies Journal, 32, 271-296.

28.

Newman

M. A.

(1993). Career advancement: Does gender make a difference? American Review of Public Administration, 23, 361-384.

29.

Orey

B. D.

Smooth

Adams

K. S.

Harris-Clark

(2006). Race and gender matter: Refining models of legislative policy making in state legislatures. Journal of Women, Politics & Policy, 28, 97-119.

30.

Pantoja

A. D.

Segura

G. M.

(2003). Does ethnicity matter? Descriptive representation in legislatures and political affiliation among Latinos. Social Science Quarterly, 84, 441-460.

31.

Paradies

(2006). A systematic review of empirical research on self-reported racism and health. International Journal of Epidemiology, 35, 888-901.

32.

Pettit

Ewert

(2009). Employment gains and wage declines: The erosion of Black women’s relative wages since 1980. Demography, 46, 469-492.

33.

Pitkin

H. F.

(1967). The concept of representation. Berkeley: University of California Press.

34.

Portman

Grune

J. A.

Johnson

(1984). The role of labor. In Remick

(Ed.), Comparable worth and wage discrimination (pp. 219-237). Philadelphia, PA: Temple University Press.

35.

Reese

C. C.

(In press). Fair pay for women (iBook). Jonesboro, Arkansas: Arkansas State University Digital Press.

36.

Reese

C. C.

Warner

(2012). Pay equity in the states: An analysis of the gender-pay gap in the public sector. Review of Public Personnel Administration, 32, 312-331.

37.

Reid

Miller

Kerr

(2004). Sex-based glass ceilings in U.S. state-level bureaucracies, 1987-1997. Administration & Society, 36, 377-405.

38.

Rocha

R. R.

Tolbert

C. J.

Bowen

D. C.

Clark

C. J.

(2010). Race and turnout: Does descriptive representation in state legislatures increase minority voting? Political Research Quarterly, 63, 890-907.

39.

Sanes

Schmitt

(2014). Regulation of public sector collective bargaining in the states. Washington, DC: Center for Economic and Policy Research.

40.

Schroedel

J. R.

Aslanian

(2017). ”A case of descriptive representation: the experience of native american elected officials in south dakota.” American Indian Quarterly, 441(3), 250-286.

41.

U.S. Department of Labor, Bureau of Labor Statistics. (2016). Retrieved from nces.ed.gov/fastfacts/display.asp?id=27

42.

Women of Color Policy Network. (n.d.). Retrieved from http://wagner.nyu.edu/wocpn/