Economic Freedom and the Affordable Care Act

Abstract

A greater level of government involvement in the financing of health care is generally viewed unfavorably by organizations monitoring economic freedom. However, increased government provision of health insurance could be associated with improved economic freedom through enhanced labor market mobility. For example, job-lock alleviation accompanying a public insurance expansion could lead to increased innovation or a higher likelihood of self-employment. In this article, we use the Affordable Care Act (ACA)’s recent Medicaid expansions to examine the effect of an increase in public health insurance provision on labor market outcomes by gender and race/ethnicity. Our results lend support to the notion that state Medicaid expansions are associated with improved labor market autonomy for white men and white women; however, we find mixed results for black and Hispanic men and women. Notably, our findings cast doubt on earlier claims that the ACA would lead to large reductions in labor force participation and employment.

Keywords

economic freedom Medicaid labor mobility Affordable Care Act

On June 28, 2012, in the case of National Federation of Independent Business v. Sebelius, the US Supreme Court issued a ruling that would fundamentally alter the impact of President Obama’s landmark 2010 health insurance legislation known as the Affordable Care Act (ACA). A primary mechanism through which the ACA sought to increase health insurance coverage among the low-income uninsured was through a large-scale expansion of the Medicaid program. Although administered at the state level, the federal government mandates Medicaid coverage requirements for certain high-risk groups.¹ In addition to these mandatory minimum standards of coverage, states often establish Medicaid eligibility criteria that extend the scope and generosity of the program.² Because the availability of these extensions varied widely across states, a large number of low-income adults have historically been unable to qualify for Medicaid coverage. Originally, the ACA’s Medicaid expansion would have extended eligibility to anyone earning below 138 percent of the federal poverty level (FPL) on January 1, 2014.³ However, while ruling in favor of the ACA provision that required individuals to obtain health insurance coverage or face a financial penalty, the Court also ruled that adoption of the ACA’s Medicaid expansion would be at the discretion of state legislatures. Despite enticing financial incentives, intense debate over Medicaid expansion ensued and, as of May 1, 2016, only thirty-one states and the District of Columbia have adopted the expanded eligibility requirements for Medicaid under the ACA (figure 1).

Figure 1.

State Medicaid expansion adoption as of May 2016. Source: Kaiser Family Foundation Status of State Action on the Medicaid Expansion Decision: http://kff.org/health-reform/state-indicator/state-activity-around-expanding-medicaid-under-the-affordable-care-act/. All expansion states adopted the Medicaid expansion on January 1, 2014, except for Michigan (April 1, 2014), New Hampshire (August 15, 2014), Pennsylvania (January 1, 2015), Indiana (February 1, 2015), Alaska (September 1, 2015), and Montana (January 1, 2016). Louisiana adopted the Medicaid expansion on July 1, 2016, but is classified as a nonexpansion state in our analysis since Current Population Survey monthly data were only available through May 2016.

Proponents of the concept of economic freedom have voiced opposition to the increased government involvement of health insurance provision in the states that have adopted the ACA’s Medicaid expansion.⁴ In fact, states that ranked high on economic freedom indices prior to January 2014 tended not to adopt the Medicaid expansion.⁵ In this article, we argue it is plausible that, rather than restrict choice and reduce economic freedom, aspects of the ACA that weaken the link between employment and health insurance coverage actually increase labor supply mobility. This increased mobility potentially enhances the efficiency of the labor market, thereby improving welfare in multiple ways.

This article uses data from the Current Population Survey’s (CPS) basic monthly data files to estimate the impact of the ACA’s Medicaid expansion on several labor market outcomes including labor force participation, employment, self-employment, the average number of hours worked, involuntary part-time work, and rate of job changes. Our analysis uses the variation in changes in Medicaid eligibility brought about by the Supreme Court’s ruling to identify the labor market effects of increased Medicaid enrollment. We use a difference-in-differences (DiD) framework to compare changes in labor market outcomes for low-income residents in states that adopted the Medicaid expansions to changes for those living in nonadopting states. Our article provides several contributions to the literature that studies the impact of public insurance expansions on labor market outcomes. While the majority of the analyses in this area focus on private insurance crowd out associated with public insurance expansions, research on labor supply responses to these expansions is relatively sparse and has found mixed results (Dave et al. 2015). Additionally, much of this previous research has examined public insurance expansions that occurred in the 1980s and 1990s and were targeted to specific subgroups (e.g., pregnant women). Medicaid expansions associated with the ACA tend to be much larger in magnitude and affect a far wider range of the population. Furthermore, research on the labor market effects of the ACA Medicaid expansions has relied solely on data pooled by race and ethnicity. There are several reasons to expect labor market effects of these expansions to differ across race and ethnicity, and this article is the first to conduct separate analyses by both gender and race and ethnicity. Finally, using the CPS basic monthly data files allows us to include more than two years of postadoption data for most states in our analysis.⁶ Since labor market changes may occur over time, this extended postperiod allows us to be confident that we are observing the longer-term effects of expansion adoption, an especially important concern in light of the temporal differences in Medicaid take-up by race/ethnicity (Artiga, Damico, and Garfield 2015).

Our results indicate that, overall, labor force participation and employment rates increased in adopting states compared to nonadopting states; however, we find a high degree of heterogeneity in our estimated effects across gender and race/ethnicity. Findings for white men and white women are consistent with an increase in economic freedom for individual workers brought about by the adoption of the ACA Medicaid expansion. However, the economic freedom implications for black and Hispanic men and women are less clear.

Economic Freedom and the ACA

Proponents of the notion of economic freedom have been vocal critics of the ACA. While the precise definition of economic freedom and the metrics used to measure it differ across organizations that track it, a larger role of government in the economy, by all definitions, is considered detrimental to economic freedom. In particular, government rules or regulations restricting or prescribing labor market behaviors reduce economic freedom. For example, the Heritage Foundation defines economic freedom as “the fundamental right of every human to control his or her own labor and property. In an economically free society, individuals are free to work, produce, consume, and invest in any way they please. In economically free societies, governments allow labor, capital, and goods to move freely and refrain from coercion or constraint of liberty beyond the extent necessary to protect and maintain liberty itself” (see http://www.heritage.org/index/about).

It is not surprising, then, that several components of the ACA, including the individual and employer mandates requiring purchase of health insurance, result in lower measured indices of economic freedom. One Heritage Foundation policy brief on the ACA and economic freedom makes the claim that “policy makers (both elected and appointed) in the United States have expanded the federal government’s already intrusive reach into the health care market with the passage of the Affordable Care Act, which dramatically increases taxes, expands Medicaid, and coerces individuals into purchasing health care through negative monetary incentives. These obtrusive policies will have questionable outcomes at best. Instead of creating massive, inefficient bureaucracies, policymakers must recognize that adhering to the principles of economic freedom is a much more effective way of distributing health care” (see http://www.heritage.org/research/reports/2013/09/how-economic-freedom-promotes-better-health-care-education-and-environmental-quality).

The Heritage Foundation’s definition of labor freedom is employer centric: “The labor freedom component is a quantitative measure that looks into various aspects of the legal and regulatory framework of a country’s labor market. It provides cross-country data on regulations concerning minimum wages; laws inhibiting layoffs; severance requirements; and measurable regulatory burdens on hiring, hours, and so on.”⁷ However, often overlooked in discussions of the impact of public health insurance expansions on economic freedom are the potential labor market effects of separating insurance coverage from employment, which could increase individual autonomy.⁸ Economists have long argued that the US system of employer-sponsored health insurance provision reduces the efficiency of the labor market by discouraging job-to-job mobility that allows workers to find better job matches. Those who become eligible for Medicaid coverage under state expansions may be less likely to experience “job lock,” or the reluctance to leave a job that offers health insurance benefits for a more desirable job with no health insurance offer or for self-employment. Several studies have identified the presence of job lock and found evidence of increased labor mobility associated with additional opportunities for obtaining health insurance outside of employment (Cooper and Monheit 1993; Gruber and Madrian 1994; Buchmueller and Valletta 1999; Adams 2004; Antwi, Moriya, and Simon 2013; Heim and Lurie 2015).⁹ Focusing on the role of eligibility expansions to Medicaid and its related programs, Hamersma and Kim (2009) find evidence that increased Medicaid generosity reduced job lock for unmarried women, while Bansak and Raphael (2008) found that expansions in the State Children’s Health Insurance Program led to increased job mobility for low-income fathers whose wives were uninsured. Since job lock reduces the efficiency of the labor market by preventing or slowing the movement of workers to better job matches, the ACA could result in increased welfare for workers and improve the efficiency of the labor market, benefiting firm profitability and worker earnings (Hamersma and Kim 2009).

Similarly, those who are participating in the labor force solely to gain access to employer-sponsored health insurance—a condition known as “employment lock”—may find that with increased Medicaid eligibility they are able to drop out of the labor force altogether (Garthwaite, Gross, and Notowidigdo 2014). Using data from a large Medicaid disenrollment in Tennessee, Garthwaite, Gross, and Notowidigdo (2014) found evidence that a reduction in public insurance coverage led to a large increase in labor market supply for affected individuals. The authors also highlight the potential relevance to the recent Medicaid expansion and predict a decline in the aggregate employment rate of between 0.3 and 0.6 percentage points as a result of the ACA. Similarly, using data from a public insurance expansion in Wisconsin, Dague, DeLeire, and Leininger (2014) suggest that the increased availability of public insurance coverage for childless adults was associated with a reduction in employment of between two and ten percentage points. Importantly, the authors could not distinguish between moving to self-employment and exiting the labor force altogether. Also, Dave et al. (2015) find that Medicaid expansions in the 1980s and 1990s targeted at pregnant women were associated with statistically significant decreases in employment for these women. While these reductions in employment or labor force participation have been a cause of concern among policy makers and critics of the ACA, economists tend to weigh the potential reduction in tax revenue and labor market productivity against the gains to individual welfare due to the greater range of feasible labor supply choices available. Alternatively, the extension of benefits to families with income up to the 138 percent FPL could encourage an increase in labor supply since the increased earnings below this limit will not trigger loss of Medicaid benefits (Dave et al. 2015).

In addition to employment lock and job lock, the Medicaid expansion has the potential to allow the choice to increase entrepreneurial activity through greater self-employment opportunities. Madrian and Lefgren (1998) found that both continuation of health insurance coverage and the availability of spousal coverage led to an increased likelihood of self-employment, while Fairlie, Kapur, and Gates (2011) reported an increase in the rate of business ownership after individuals were eligible for Medicare coverage. Alternatively, Holtz-Eakin, Penrod, and Rosen (1996) found mixed evidence on the impact of insurance portability on self-employment.

It is also important to note that existing research finds differential effects of economic freedom across different demographic groups (Compton, Giedeman, and Hoover 2014; Hall, Humphreys, and Ruseski 2015; Hoover, Compton, and Giedeman 2015). Furthermore, labor market behaviors differ significantly across demographic groups, and thus, the labor mobility effects of the ACA could also differ across these same groups. For example, figure 2 plots four measures of labor market behavior by race/ethnicity and gender for 2013. Prior to the enactment of the ACA, white and Hispanic men and women had higher rates of self-employment compared to black men and women. Also, black men had relatively lower rates of employment compared to whites and Hispanics. Additionally, black women had a higher rate of labor force participation compared to white and Hispanic women (albeit the differences are not large), whereas black men had much lower rates of labor force participation than white and Hispanic men. Moreover, prior to the ACA, health coverage and poverty rates varied significantly across these groups and, thus, the impact on Medicaid eligibility of expansion adoption should differ by race and ethnicity. For example, figure 3 presents poverty statistics by race and ethnicity for states that adopted the ACA Medicaid expansion and states that did not for 2013. The data clearly demonstrate that the relative effect of the Medicaid expansion on Medicaid eligibility will be larger for blacks and Hispanics than for whites, as the share of blacks and Hispanics below the FPL is approximately double that of whites.¹⁰

Figure 2.

Labor supply by race and gender for those earning 138 percent of the federal poverty level or less, 2013. Source: Current Population Survey basic monthly data files.

Figure 3.

Share of the population at or below 138 percent of the federal poverty level, 2013. Source: Authors’ calculations using the American Community Survey, 2013.

In summary, the ACA has been condemned by individuals and organizations that promote the notion of economic freedom because of the large increase in government involvement in health-care markets and in labor markets the law entails. But the criticisms neglect the potential improvements in economic freedom resulting from the increased flexibility individuals will have regarding labor market decisions. In fact, some critics of the ACA, such as the Heritage Foundation, have focused on predictions of lower labor supply under the ACA as a problem with the law rather than a potential benefit to economic freedom (see http://dailysignal.com/2010/10/26/even-cbo-is-skeptical-of-obamacare/). Proponents of the ACA, on the other hand, have noted the potential of the law to enhance economic freedom as a result of the increased individual autonomy discussed above. How large an impact the ACA has on increased individual labor market mobility and how those effects differ across gender and race/ethnicity is largely an empirical question, which we turn to next.

Data

The data for our analysis come primarily from the CPS’s basic monthly data files. The CPS is a monthly survey of approximately 60,000 US households conducted by the US Census Bureau for the Bureau of Labor Statistics. We use the January 2008 through May 2016 waves of the CPS in order to include several years of pre-Medicaid expansion data and more than two years of postexpansion data. We calculate each respondent’s relative position to the FPL using information on the respondent’s family income and family size.

Since income in the monthly CPS files is reported categorically, we use the midpoint of the reported income category in our FPL calculation. The majority of our analyses are limited to those earning 138 percent of the FPL or less, as this would be the group most likely affected by a state’s Medicaid expansion adoption decision. Additionally, the CPS contains information on several demographic and socioeconomic characteristics including age, gender, race, education, marital status, and state of residence. For all analyses that follow, we limit our sample to those between the ages of eighteen and sixty-five.¹¹

One drawback of the CPS basic monthly data files is the lack of available information on insurance coverage. Because of this, we use data from the 2009 through 2015 Annual Social and Economic Supplement (ASEC) of the CPS to analyze changes in Medicaid coverage following the ACA expansions. Insurance coverage in the ASEC is measured in the prior year, so our analysis period for effects on Medicaid coverage spans 2008 through 2014, while our labor market outcomes analysis covers the period from January 2008 through May 2016.

Table 1 displays weighted means for those earning 138 percent of the FPL or less in states that adopted the ACA Medicaid expansion and those that did not. Compared to the nonadopting states, the adopting states have relatively fewer black residents and more Hispanic and Asian residents and higher levels of education.

Table 1.

Descriptive Statistics for Those Earning 138 Percent of the Federal Poverty Level or Less.

Variable	Adopting states	Nonadopting states
Age	38.95	39.05
Female	0.551	0.556
Race/Ethnicity
White	0.496	0.506
Hispanic	0.251	0.222
Black	0.150	0.212
Asian	0.066	0.025
Other race	0.037	0.035
Marital status
Never married	0.410	0.364
Married	0.391	0.413
Divorce/separated	0.172	0.191
Widowed	0.028	0.032
Education
Less than high school	0.261	0.264
High school	0.369	0.378
Some college	0.264	0.270
College graduate	0.081	0.069
Graduate school	0.025	0.020
Observations	937,759	583,011

Note: Means are constructed using survey sample weights.

Empirical Methods

We rely on the state-level variation in Medicaid expansion generated by the 2012 Supreme Court decision to identify the impact of increased Medicaid coverage on a variety of labor market outcomes. Specifically, we estimate a DiD model that compares relative labor market changes for low-income residents in states that adopted the ACA Medicaid expansion to low-income residents in states that chose not to adopt the expansion. A requirement of the DiD specification is that in the absence of the intervention, in this case Medicaid expansion, trends for both adopting and nonadopting states would have been stable. We provide evidence of similar trends for our outcome measures by plotting weighted means for each outcome over time in figure 4. We see no evidence of diverging trends in the preperiod between adopting and nonadopting states for any of the outcomes we examine. These parallel trends in the preperiod strengthen the validity of our DiD approach. We formalize our DiD estimation equation as follows:

Y_{i s t} = α + γ_{1} A d o p t_{s t} + X_{i s t} β + δ_{s} + τ_{t} + ∊_{i s t},

where Y_ist is the outcome of interest for person i in state s in month t. We analyze several outcomes associated with labor market behavior including labor force participation, employment, self-employment, average weekly hours worked, probability of working more than thirty hours per week, involuntarily working part-time, and the rate of job changes. Adopt_st is an indicator for state adoption of the ACA Medicaid expansion and the coefficient, γ₁, represents the DiD estimate of the differential effect in adopting and nonadopting states.¹² The vector X_ist includes other control variables such as age, education, dummy variables for gender, race, and marital status. The variables δ_s and τ_t are state- and year-fixed effects, respectively. We also include month-fixed effects in all of our regressions to control for seasonal changes in labor market behaviors. Finally, in most specifications, we include a state-by-year time trend to control for potential endogeneity in the adoption of the Medicaid expansion. All regressions are weighted using the sampling weights provided by the CPS. We use linear probability models for specifications with binary outcome variables.¹³

Figure 4.

Trends in labor market outcomes by year for those earning 138 percent of the federal poverty level or less, January 2008 to May 2016. Source: Current Population Survey basic monthly data files. The six states that adopted the Affordable Care Act Medicaid expansion after January 1, 2014, were excluded from the figures. See figure 1 for a list of these states and their adoption dates.

We first report estimates of the effect of the ACA’s Medicaid expansion on Medicaid coverage. We then examine several labor market outcomes that could plausibly be affected by the expansion and, depending on the direction of the impact, would also indicate increased labor market freedom for individuals. We include labor force participation rates and rates of self-employment to measure potential reductions in employment lock and job lock. Reductions in labor force participation or increases in the rate of self-employment would indicate increased labor market mobility following the Medicaid expansion. We also measure employment changes associated with the ACA Medicaid expansions. Theoretically, we might expect Medicaid expansions to increase employment since the expanded eligibility requirements would allow recipients to earn higher incomes and still qualify for Medicaid. However, although delayed by one year, the ACA’s employer mandate that required certain employers to provide health insurance to their employees was originally slated to take effect on January 1, 2014. If employers were reacting to the initial implementation date of the mandate, then it is possible that employment levels could fall as employers attempted to avoid mandated insurance provision. Additionally, we measure changes in the probability of working more than thirty hours per week and the share of workers who changed jobs in the past month.¹⁴ Some caution must be used in interpreting these results. For example, a finding of reduced hours is consistent with either employee-initiated changes—which should be interpreted as an increase in labor market freedom—or employer-initiated changes in response to the employer mandate (Garrett and Kaestner 2015). Similarly, higher rates of job changes after the adoption of the ACA Medicaid expansions could be because of reduced job lock or because employers are making changes to their labor force in response to the ACA. We also examine changes in involuntary part-time employment, which the CPS defines as working part-time for an economic reason such as slack work, unfavorable business conditions, inability to find full-time work, or seasonal declines in demand. Changes in involuntary part-time work could clarify whether a change in hours worked was driven by employers or by workers themselves. In addition to the full sample, we present results separately by gender and race since, as noted previously, these groups may have different preferences for health insurance or may have experienced differential impacts of the Medicaid expansion.

Results

We begin by examining changes in Medicaid coverage associated with the ACA’s Medicaid expansion that took effect on January 1, 2014. Table 2 presents regression estimates of the change in Medicaid coverage for respondents in states that adopted the Medicaid expansion compared to those in states that decided against adopting the expansion. Results presented in column (1) include demographic controls along with state-, year-, and month-fixed effects while column (2) adds a state-year time trend. Estimates indicate that, compared to nonadopting states, states that adopted the expansion saw an increase in Medicaid enrollment of between 25 percent and 32 percent. It is this large relative increase in Medicaid enrollment that gives rise to potential labor market effects for those earning 138 percent of the FPL or less in adopting states. Because the inclusion of the state-year trend appears to have a significant impact on our estimates, we include state-year trends in all of the following analyses.

Table 2.

Estimates of the Effect of the Affordable Care Act (ACA) Medicaid Expansions on Medicaid Coverage.

Regression Elements	(1)	(2)
Adopt	0.036*** (0.007)	0.028*** (0.005)
State fixed effects	Yes	Yes
Year fixed effects	Yes	Yes
State-year time trend	No	Yes
Mean of dependent variable	0.114	0.114
Observations	721,989	721,989

Source: Annual Social and Economic Supplements of the Current Population Survey from 2009 to 2015.

Note: All specifications include controls for age, gender, race, education, income, and marital status. “Mean of dependent variable” denotes the mean of each outcome for adopting states in the period prior to ACA Medicaid expansion. All regressions use survey sample weights. Standard errors are clustered at the state level.

*p < .10.

**p < .05.

***p < .01.

In Table 3, we turn to estimates of the impact of the Medicaid expansion on our measures of labor market activity. Column (1) displays results for the full sample that includes respondents over the entire range of reported income. We find evidence that, compared to nonadopting states, labor force participation in states that adopted the Medicaid expansion increased slightly. This increase in labor force participation is inconsistent with the notion that an expansion in Medicaid eligibility led to a reduction in employment lock; however, it also contradicts claims that the Medicaid expansion would lead to large declines in labor force participation. Column (2) limits the sample to those earning 138 percent of the FPL or less. Since the Medicaid expansions were targeted at those earning at or below this threshold, we would expect to find larger labor market changes for this group. In column (2), we find a slightly larger positive effect on labor force participation associated with adoption of the Medicaid expansion, although the estimate is not statistically significant.

Table 3.

Estimates of the Effect of Affordable Care Act (ACA) Medicaid Expansion on Labor Force Participation and Employment.

	Full sample	Below 138 percent federal poverty level
Outcomes	(1)	(2)
Labor force participation	0.006** (0.003)	0.008 (0.006)
Mean of dependent variable	0.768	0.576
Observations	8,032,317	1,520,770
Employment	0.004* (0.002)	0.012*** (0.004)
Mean of dependent variable	0.923	0.794
Observations	6,119,463	871,645
Self-employment	0.000 (0.002)	0.001 (0.006)
Mean of dependent variable	0.107	0.122
Observations	5,726,072	723,102
Average weekly hours worked	–0.015 (0.077)	–0.149 (0.254)
Mean of dependent variable	38.99	34.88
Observations	5,286,845	640,132
Work more than thirty hours per week	0.001 (0.003)	–0.004 (0.008)
Mean of dependent variable	0.872	0.759
Observations	5,286,845	640,132
Involuntary part-time work	–0.004** (0.002)	–0.010** (0.004)
Mean of dependent variable	0.044	0.123
Observations	5,700,685	713,109
Job change	0.001 (0.001)	0.002 (0.003)
Mean of dependent variable	0.020	0.030
Observations	3,717,141	426,502

Notes: All specifications include controls for age, gender, race, education, income, and marital status. All specifications include state fixed effects, year fixed effects, month fixed effects, and a state-year time trend. Employment and self-employment outcomes are conditional on labor force participation. “Mean of dependent variable” denotes the mean of each outcome for adopting states in the period prior to ACA Medicaid expansion. All regressions use survey sample weights. Standard errors are clustered at the state level.

*p < .10.

**p < .05.

***p < .01.

The second panel in table 3 examines employment effects of expansion adoption. For the full sample, we find a small increase in employment associated with adoption, but this effect becomes much larger when we limit our sample to those earning 138 percent of the FPL or less. In this case, the positive change in employment for adopting states was approximately 1.5 percent greater than in states that chose against adopting the expansion. As discussed earlier, positive employment effects of an increase in Medicaid eligibility could result from the opportunity to achieve higher earnings while still qualifying for Medicaid benefits. We also see a decrease in the probability of involuntary part-time work for both the full- and low-income samples. This decline suggests an improvement in the economic freedom of workers who are increasingly choosing to work part-time after the expansion rather than working part-time for employer-specific reasons.

The remaining outcomes in table 3 show no relative changes between those living in adopting states compared to those living in nonadopting states after implementation of the Medicaid expansions. We find no differential effect on self-employment, average weekly hours worked, the probability of working more than thirty hours per week, or the likelihood of changing jobs in the past month.

Table 4 repeats the analysis reported in table 3 but limits the sample to women earning 138 percent of the FPL or less and examines differential changes by race and ethnicity.¹⁵ Given the large differences by race and gender in insurance coverage and labor market behaviors, we are interested in examining heterogeneous effects of the ACA Medicaid expansions on these groups. Prior analyses of the effect of health insurance on labor mobility have noted differential changes for men and women (Hamersma and Kim 2009; Bailey and Chorniy 2015). Explanations for the gender disparity tend to focus on the belief that women place a higher value on health insurance than men and are therefore more likely to be subject to job lock or employment lock. The first panel in table 4 shows that, for women, the increase in labor force participation that was observed for the full sample is concentrated primarily in white and Hispanic respondents. Compared to those in nonadopting states, white women in adopting states increased their labor force participation by 3.4 percent after the expansion, while the increase for Hispanic women was on the order of 5.2 percent. We find no such change in labor force participation for black women in our sample. Similarly, the second panel presents evidence of increased employment for white and Hispanic women in adopting states relative to those in nonadopting states. Our estimates indicate that employment for white women increased by nearly 2 percent and employment for Hispanic women increased by approximately 2.4 percent. We also find that the decrease in the probability of involuntary part-time work for the full sample carries over to both white and black women in table 4. White and black women in adopting states experienced, on average, a 14 percent and 22 percent reduction in involuntary part-time work, respectively.

Table 4.

Estimates of the Effect of Affordable Care Act Medicaid Expansion on Women’s Labor Force Participation and Employment by Race and Ethnicity.

	White	Hispanic	Black
Outcomes	(1)	(2)	(3)
Labor force participation	0.018** (0.008)	0.027** (0.012)	–0.006 (0.016)
Mean of dependent variable	0.524	0.520	0.529
Observations	410,833	199,396	160,396
Employment	0.016** (0.007)	0.019* (0.011)	0.002 (0.011)
Mean of dependent variable	0.833	0.801	0.718
Observations	210,989	103,253	87,251
Self-employment	0.017** (0.008)	–0.011 (0.015)	0.012 (0.010)
Mean of dependent variable	0.102	0.078	0.052
Observations	180,459	86,936	66,937
Average weekly hours worked	-0.297 (0.395)	0.433 (0.552)	0.187 (0.527)
Mean of dependent variable	31.63	33.35	33.45
Observations	159,900	79,257	61,115
Work more than thirty hours per week	–0.012 (0.015)	0.020 (0.019)	–0.003 (0.022)
Mean of dependent variable	0.651	0.735	0.738
Observations	159,900	79,257	61,115
Involuntary part-time work	–0.018* (0.009)	–0.011 (0.007)	–0.032* (0.016)
Mean of dependent variable	0.126	0.161	0.145
Observations	178,961	86,054	66,253
Job change	0.003 (0.004)	0.001 (0.005)	–0.019** (0.009)
Mean of dependent variable	0.031	0.028	0.027
Observations	112,021	50,066	39,383

Note: All specifications are limited to those earning below 138 percent of the federal poverty level. See footnote in table 3.

Unlike the results for the full sample, table 4 indicates that self-employment rates increased for white women in adopting states compared to nonadopting states. Greater rates of self-employment following an increase in public insurance generosity are indicative of a reduction in job lock for white women and a higher level of economic freedom. Estimates for changes in self-employment are positive for black women, but statistically insignificant, while Hispanic women show a statistically insignificant decrease in the likelihood of self-employment. Additionally, we find a large reduction in reported job changes for black women in adopting states compared to those in nonadopting states, which would not be consistent with an increase in economic freedom.

Table 5 examines racial differences in the labor market effects of the ACA Medicaid expansions for men earning 138 percent of the FPL or less. Unlike the results for women, we find no differential changes in labor force participation for white and Hispanic men between adopting and nonadopting states. However, we do find a relative increase of more than 8 percent in the labor force participation of black men. Again, this increase in labor force participation for black men could be due to the fact that a higher Medicaid eligibility cutoff allows for greater earnings while retaining public insurance benefits. The second panel of table 5 shows that a similar pattern holds for employment. We find no relative changes in employment levels for white or Hispanic men; however, employment for black men increased by 8.6 percent in adopting states compared to states that did not adopt the Medicaid expansion.

Table 5.

Estimates of the Effect of Affordable Care Act Medicaid Expansion on Men’s Labor Force Participation and Employment by Race and Ethnicity.

	White	Hispanic	Black
Outcomes	(1)	(2)	(3)
Labor force participation	–0.017 (0.012)	0.011 (0.011)	0.045** (0.019)
Mean of dependent variable	0.608	0.776	0.536
Observations	348,787	166,114	103,531
Employment	0.016 (0.010)	0.016 (0.011)	0.057** (0.022)
Mean of dependent variable	0.785	0.817	0.656
Observations	211,844	130,220	55,764
Self-employment	0.009 (0.013)	–0.021** (0.008)	0.010 (0.024)
Mean of dependent variable	0.198	0.114	0.109
Observations	175,685	112,698	39,771
Average weekly hours worked	–0.721** (0.319)	0.081 (0.227)	0.410 (0.770)
Mean of dependent variable	37.62	37.73	35.51
Observations	149,725	101,012	34,822
Work more than thirty hours per week	–0.015 (0.009)	0.005 (0.008)	0.006 (0.028)
Mean of dependent variable	0.812	0.879	0.791
Observations	149,725	101,012	34,822
Involuntary part-time work	0.009 (0.007)	–0.011 (0.008)	–0.014 (0.019)
Mean of dependent variable	0.100	0.122	0.137
Observations	172,017	111,113	38,806
Job change	0.009** (0.004)	0.006 (0.008)	0.000 (0.011)
Mean of dependent variable	0.031	0.032	0.030
Observations	104,944	63,229	21,558

Note: All specifications are limited to those earning below 138 percent of the federal poverty level. See footnote in table 3.

Additionally, for white men in adopting states, table 5 provides evidence that average weekly hours worked fell while job changes increased relative to white men in nonadopting states. Both of these findings are consistent with an increase in economic freedom resulting from the adoption of the Medicaid expansions. Coupled with no reported decrease in involuntary part-time work, the reduction in hours for white men appears to be a worker-driven decision and not the result of employers reducing their labor demand. We also find that white men in adopting states were nearly 30 percent more likely to report a job change than those in nonadopting states. Finally, with the exception of a reduction in self-employment for Hispanic men that reflects a similar reduction for Hispanic women, the remaining outcomes in table 5 do not appear to have been affected by the adoption of the Medicaid expansion.

Conclusion

Proponents of the concept of economic freedom are generally unsupportive of any expanded role of government in the provision of health insurance or health care. However, given the prominence of individual autonomy as a tenet of economic freedom, an expansion of government-sponsored coverage that severs the link between health insurance and employment could lead to a meaningful improvement in economic freedom, market efficiency, and individual welfare for the affected population. Evidence on the effect of public health insurance expansions on worker mobility has been equivocal, with some studies reporting improved labor market mobility associated with expansions (Bansak and Raphael 2008; Hamersma and Kim 2009) and others finding no effect (Blank 1989; Berger, Black, and Scott 2004; Bailey and Chorniy 2015).

Our article exploits the recent state-level Medicaid expansions brought about by the ACA to examine the impact of increased public health insurance coverage on several labor market outcomes by race and gender. Our empirical results generally do not support the idea that the ACA restricted individuals’ labor market autonomy. In fact, our findings suggest improvements in outcomes associated with increased autonomy for low-income white men and white women. The heterogeneity in the effects of expansion adoption by race and gender underscores the importance of analyzing the differential impacts of policies that affect economic freedom. Unlike recent studies of smaller-scale changes in government-sponsored insurance generosity, we find no evidence that the ACA’s Medicaid expansion led to lower rates of labor force participation (Dague, DeLeire, and Leininger 2014; Garthwaite, Gross, and Notowidigdo 2014). Instead, we find evidence that, for many of the demographic groups in our analysis, labor force participation and employment both increased in states that adopted the ACA Medicaid expansions compared to states that did not adopt the expansion. These findings contradict the widespread claims that the ACA would lead to undesirable labor market outcomes and are similar to findings reported in Garrett and Kaestner (2015) for those with low levels of education.¹⁶

Our study has a number of caveats that should be addressed by future research in this area. FPL definitions in our data did not match directly to requirements for inclusion in expanded Medicaid coverage, which relies on projected income during the enrollment period. Additionally, limiting the sample to those earning less than 138 percent of the FPL could be problematic if individuals are altering their labor market behavior in order to meet the eligibility threshold. One advantage of our study over the previous analyses of labor market effects of the ACA Medicaid expansions is our extended postperiod of more than two years for most states. However, it is possible that labor market adjustments associated with public insurance expansions may occur over a longer period of time.

Footnotes

Acknowledgments

The authors would like to thank four anonymous referees for helpful comments and suggestions.

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

References

Adams

Scott J.

2004. “Employer-provided Health Insurance and Job Change.” Contemporary Economic Policy 22:357–69.

Antwi

Yaa A.

Moriya

Asako S.

Simon

Kosali

. 2013. “Effects of Federal Policy to Insure Young Adults: Evidence from the 2010 Affordable Care Act’s Dependent Coverage Mandate.” American Economic Journal: Economic Policy 5:1–28.

Artiga

Samantha

Damico

Anthony

Garfield

Rachel

. 2015. “Estimates of Eligibility for ACA Coverage among the Uninsured by Race and Ethnicity.” Kaiser Family Foundation Issue Brief, October 2015. Accessed September 8, 2015. http://kff.org/disparities-policy/issue-brief/estimates-of-eligibility-for-aca-coverage-among-the-uninsured-by-race-and-ethnicity/.

Bailey

James

Chorniy

Anna

. 2015. “Employer-provided Health Insurance and Job Mobility: Did the Affordable Care Act Reduce Job Lock?” Contemporary Economic Policy 34:173–83.

Bansak

Cynthia

Raphael

Steven

. 2008. “The State Children’s Health Insurance Program and Job Mobility: Identifying Job Lock among Working Parents in Near-poor Households.” Industrial and Labor Relations Review 61:564–79.

Berger

Mark C.

Black

Dan A.

Scott

Frank A.

. 2004. “Is There Job Lock? Evidence from the Pre-HIPAA Era.” Southern Economic Journal 70:953–76.

Blank

Rebecca M.

1989. “The Effect of Medical Need and Medicaid on AFDC Participation.” Journal of Human Resources 24:54–87.

Buchmueller

Thomas C.

Valletta

Robert G.

. 1999. “The Effect of Health Insurance on Married Female Labor Supply.” Journal of Human Resources 34:42–70.

CBPP (Center on Budget and Policy Priorities). 2015. “Policy Basics: Introduction to Medicaid.” Accessed October 21, 2015. http://www.cbpp.org/research/health/policy-basics-introduction-to-medicaid.

10.

Compton

Ryan A.

Giedeman

Daniel C.

Hoover

Gary A.

. 2014. “A Distributional Analysis of the Benefits of Economic Freedom.” European Journal of Political Economy 33:121–33.

11.

Cooper

Philip F.

Monheit

Alan C.

. 1993. “Does Employment-related Health Insurance Inhibit Job Mobility?” Inquiry 30:400–16.

12.

Dague

Laura

DeLeire

Thomas

Leininger

Lindsey

. 2014. “The Effect of Public Insurance Coverage for Childless Adults on Labor Supply.” National Bureau of Economic Research Working Paper, Cambridge, MA.

13.

Dave

Dhaval

Decker

Sandra L.

Kaestner

Robert

Simon

Kosali I.

. 2015. “The Effect of Medicaid Expansions in the Late 1980s and Early 1990s on the Labor Supply of Pregnant Women.” American Journal of Health Economics 1:165–93.

14.

Fairlie

Robert W.

Kapur

Kanika

Gates

Susan

. 2011. “Is Employer-based Health Insurance a Barrier to Entrepreneurship?” Journal of Health Economics 30:146–62.

15.

Garrett

Bowen

Kaestner

Robert

. 2015. “Recent Evidence on the ACA and Employment: Has the ACA Been a Job Killer?” ACA Implementation—Monitoring and Tracking. Robert Wood Johnson Foundation and the Urban Institute. Accessed September 8, 2015. http://www.rwjf.org/en/library/research/2015/08/recent-evidence-on-the-aca-and-employment--has-the-aca-been-a-jo.html.

16.

Garthwaite

Craig

Gross

Tal

Notowidigdo

Matthew J.

. 2014. “Public Health Insurance, Labor Supply, and Employment Lock.” The Quarterly Journal of Economics 129:653–96.

17.

Gooptu

Angshuman

Moriya

Asako S.

Simon

Kosali I.

Sommers

Benjamin D.

. 2016. “Medicaid Expansion did not Result in Significant Employment Changes or Job Reductions in 2014.” Health Affairs 35:111–18.

18.

Gruber

Jonathan

Madrian

Brigitte C.

. 1994. “Health Insurance and Job Mobility: The Effects of Public Policy on Job-lock.” Industrial and Labor Relations Review 48:86–102.

19.

Hall

Joshua C.

Humphreys

Brad R.

Ruseski

Jane E.

. 2015. “Economic Freedom, Race, and Health Disparities: Evidence from the US States.” West Virginia University Department of Economics Working Paper No. 15-43, Morgantown, WV.

20.

Hamersma

Sarah

Kim

Matthew

. 2009. “The Effect of Parental Medicaid Expansions on Job Mobility.” Journal of Health Economics 28:761–70.

21.

Heim

Bradley T.

Lurie

Ithai Z.

. 2015. “Does Health Reform Lead to Increased Job Mobility? Evidence from Massachusetts.” American Journal of Health Economics 1:374–98.

22.

Hoover

Gary A.

Compton

Ryan A.

Giedeman

Daniel C.

. 2015. “The Impact of Economic Freedom on the Black/White Income Gap.” American Economic Review Papers and Proceedings 105:587–92.

23.

Holtz-Eakin

Douglas

Penrod

John R.

Rosen

Harvey S.

. 1996. “Health Insurance and the Supply of Entrepreneurs.” Journal of Public Economics 62:209–35.

24.

IRS (Internal Revenue Service). 2015. “Questions and Answers on Employer Shared Responsibility Provisions Under the Affordable Care Act.” Accessed September 8, 2015. https://www.irs.gov/Affordable-Care-Act/Employers/Questions-and-Answers-on-Employer-Shared-Responsibility-Provisions-Under-the-Affordable-Care-Act.

25.

Kaiser Family Foundation. 2013. Eligibility for Pregnant Women in Medicaid/CHIP by Income, January 2013. Accessed September 8, 2015. http://kff.org/medicaid/slide/eligibility-for-pregnant-women-in-medicaidchip-by-income-january-2013.

26.

Madrian

Brigitte C.

Lefgren

Lars J.

. 1998. “The Effect of Health Insurance on Transitions to Self Employment.” University of Chicago Working Paper, Chicago, IL.

27.

Miller

Terry

Kim

Anthony B.

. 2014. “Defining Economic Freedom.” In Heritage Foundation 2012 Index of Economic Freedom, Ch. 5, pp. 79–86. Washington D.C. and New York, NY: Heritage Foundation and Dow Jones & Company, Inc.