Interest Group Conflict Over Medicaid Expansion: The Surprising Impact of Public Advocates

Timothy Callaghan is a PhD candidate in the Department of Political Science, University of Minnesota, Minneapolis. Lawrence R. Jacobs is with the Humphrey School of Public Affairs, University of Minnesota.

Correspondence should be sent to Lawrence R. Jacobs, Mondale Chair of Political Studies, Humphrey School of Public Affairs, University of Minnesota, 137 Humphrey School, 301 19th Ave. S., Minneapolis, MN 55455 (e-mail: ude.nmu@sbocajl). Reprints can be ordered at http://www.ajph.org by clicking the “Reprints” link.

T. Callaghan collected the data and conducted relevant quantitative analysis. L. R. Jacobs took the lead in writing the article. The authors worked collaboratively to develop the theory for this project, decide about statistical modeling and interpreting results, and make necessary edits to the article.

Abstract

Objectives. We examined the potential economic, policy, and political influences on the decisions of the 50 US states to expand Medicaid under the Affordable Care Act.

Methods. We related a measure of relative state progress toward Medicaid expansion updated to 2015 to each state’s economic circumstances, established policy frameworks in states, partisan control of state government, and lobbyists for businesses, medical professionals, unions, and public interest organizations.

Results. The 9201 lobbyists working on health care reform in state capitols exerted a strong and significant impact on Medicaid expansion. Controlling for confounding factors (including partisanship and existing policy frameworks), we found that business and professional lobbyists exerted a negative effect on state Medicaid expansion and, unexpectedly, that public interest advocates exerted a positive effect.

Conclusions. There are 3.1 million adults who lack coverage because they live in the 20 states that refused to expand Medicaid. Although political party and lobbyists for private interests present significant barriers in these states, legislative lobbying on behalf of the uninsured appears likely to be effective.

A key component of the Patient Protection and Affordable Care Act (ACA; Pub L No. 111–148) is to expand the health care coverage of low-income individuals by broadening Medicaid eligibility to those with incomes less than 138% of the federal poverty level ($16 105 per year for an individual in 2014 as determined by the Department of Health and Human Services). The Supreme Court’s June 2012 ruling weakened the ACA’s power to expand Medicaid by giving states the option of adopting the expansion without the threat of losing existing funding. Although the Supreme Court declared unconstitutional the stick of financial penalty, the ACA continues to offer financial incentives for states: full funding of newly eligible individuals from 2014 to 2016 and 90% by 2020. The financial incentives have produced, however, uneven adoption because 20 states rejected expansion. The result is that approximately 3.1 million poor uninsured adults fall into the “coverage gap”—their income falls below eligibility for the ACA’s insurance exchange tax credits and above the cutoff for Medicaid in states that did not expand the program. 1 Pinpointing why states accepted or rejected Medicaid expansion has urgent implications for the funding of medical providers and the health of lower-income people. 2–5

The dominant account for the uneven implementation of the ACA’s Medicaid expansion is the preeminent influence of political partisanship. 6,7 This is consistent with research on national and state politics that identifies the partisan control of executive and legislative offices as the “workhorse” of decision-making. 8–11 The “party rules” explanation for Medicaid expansion captures an unmistakable pattern: all 13 states where Democrats had unified control over the executive and legislative branches in 2014 adopted it, whereas states where Republicans enjoyed unified control of government or wielded a veto through its control of 1 lever of constitutional power rejected it. If the party rules account is accurate, states with Republicans wielding power are not expected to expand Medicaid under the ACA. However, partisanship is not a sufficient account of Medicaid expansion: 23 states where Republicans wield power—including Arizona, Iowa, New Mexico, Ohio, Pennsylvania, New Jersey, and North Dakota—rejected their party’s national position and implemented reform.

There has been little systematic study, though, of what influences state adoption of Medicaid expansion beyond partisanship. Fortunately, a substantial body of research on state and national policymaking presents compelling empirical and theoretical grounds for focusing on 3 sets of alternatives explanations for Medicaid expansion. State affluence is a possible offset of partisanship—but not as health reformers expected. The ACA’s advocates confidently predicted that generous financial terms would entice states—especially the less affluent ones—to expand Medicaid, but research on state politics and health policy points to a more dour expectation: states with weaker economies are fiscally constrained from committing even modest additional resources, whereas more affluent states are better positioned to seize Washington’s financial incentives and pursue new federal initiatives. 12–15

A second potential break on partisanship is the inertial dynamics of the established policy frameworks in states. Social science research on the development of Medicaid since its foundation has focused on state variation in what is defined as “generosity” in medical coverage. 16–18 Also of relevance are studies of established government programs that point to “path dependency.” These bodies of research suggest that states with generous social provisions and competent administrative capacity, which earn the confidence of politicians, tend to develop along similar trajectories. 19,20 New Jersey illustrates a potential pattern that is consistent with previous research 21–23 : the path dependency account suggests that its history of administrative competence and generous programs for the vulnerable, including the State Children’s Health Insurance Program, pressure Republican lawmakers to continue to pursue these traditions by adopting Medicaid expansion.

The pressure of interest groups and lobbyists on legislators is a third potential influence on Medicaid adoption. Numerous statistical studies of policymaking and health reform detect the advantages that businesses and professional organizations have over advocacy groups for the poor and vulnerable to raise funds and hire more lobbyists with better access. 24–30 Resource advantages are not, however, synonymous with influence on policy. Social science research in a range of policy areas finds that as the number of lobbyists representing private interests rises, they tend to compete with each other and counteract their advantages. 31–34 In health care reform, for example, the American Hospital Association and members of the medical profession geared to general practice and community care pressed for Medicaid expansion. This pressure within the business community was counteracted, as statistical research indicated, by the opposition of specialized physician associations (representing orthopedists, urologists, and others), insurers, and employer organizations (including large umbrella groups like the US Chamber of Commerce), which protested new regulations and higher taxes to pay for expanded programs. 35–37

Social science research finds that the resource advantage of business and professional organizations can also be offset by lobbyists for consumers, unions, and religious or charitable groups who use alternative sources of political influence—grassroots protests, letter writing, investigations that instigate media scrutiny, and appeals to voters. 38 Consider the closely studied example of smoking: well-funded lobbyists for the tobacco industry won government subsidies and lax regulation for decades until businesses outside tobacco organized in response to the health costs they absorbed, and, especially, advocates for public health mobilized to put smoking on the agenda as a public health problem and pressured lawmakers to change policy. 39,40

In short, research suggests competing sources of political pressure: a rise in the number of lobbyists can increase the influence of business and professions as well as public interest groups and unions. There are caveats, however: the effect of business and professional organization lobbyists can be offset by intense internal disagreement, and the presence of public interest and union lobbyists is often muted because scarce resources limit their ability to afford lobbyists.

The decision of states to expand Medicaid has significant policy implications, and yet research on what drives those decisions has largely focused on political party. The large body of social science research on policymaking that reports influences beyond partisanship has been largely neglected. Drawing on this research, we examined whether state decisions regarding Medicaid expansion are influenced by state-level measures of partisanship, economic affluence, the inertial dynamics of the established policy frameworks, and the role of health lobbyists for businesses, unions, and groups representing the poor and vulnerable.

METHODS

The Kaiser Family Foundation and other organizations have tracked Medicaid expansion, but these measures omit important variations in state policymaking by collapsing decisions to adopt into the simple categories of accepting expansion, rejecting it, or remaining uncommitted. 41–43 A further hurdle is that previous research on Medicaid does not offer clear guidance for appropriate measures of expansion because it was geared to studying an established program. 18,44 We adopted a new approach for our dependent variable that is tailored to measuring relative state progress through 5 stages of development toward greater compliance with federal requirements.

The first stage is state application for 1 or more “level 1” federal grant to finance planning for health reform implementation. Although originally intended to help states plan for insurance exchanges, these grants were also used to plan for other aspects of health care reform, including Medicaid expansion. The application process for obtaining these grants requires a concerted effort from state government to detail how it will use federal funds to comply with the ACA. Although a federal planning grant indicates a general intent to prepare for reform and identify options, the second level is more important and moves toward actual state approval of the Medicaid expansion—the governor’s public announcement of support for the expansion. Gubernatorial support for reform is a necessary (but not sufficient) step; the executive branch takes the initiative for planning, and the state’s chief executive must sign (as opposed to veto) enacted legislation to authorize the expansion.

The third level of relative state progress accounts for states seeking a federal waiver to implement some sort of Medicaid expansion related to the ACA. All states in this category have made the decision to implement some form of Medicaid expansion without adopting the original model stipulated by the ACA and the Center for Medicare and Medicaid Services. The fourth step toward greater federal compliance is federal approval of a state’s waiver request, signifying that a state’s solution to reform complies with the intent of the ACA. Four states have reached this status at this point: Arkansas, Iowa, Indiana, and Michigan. Finally, the fifth and highest level of state implementation is its adoption of Medicaid expansion as expressly stipulated under the ACA to comply fully with its requirements. (An appendix, available as a supplement to the online version of this article at http://www.ajph.org, provides detailed information on each stage and on the variables.)

To explain the relative progress of states toward Medicaid implementation, we created 7 explanatory variables. The first variable accounts for political party control of state government. This additive measure ranged from 0 to 6, with 3 points awarded for Democratic control of the governorship or both branches of the state legislature and 1 point given for Democratic control of only 1 legislative chamber. We explored alternative measures of partisanship—including separate variables for each branch of government—and found that our key findings remain unchanged.

The second explanatory variable is economic affluence, which we measured as real per capita personal income on the basis of data from the Bureau of Economic Analysis. The next 2 variables measured elements of inertial dynamics—the trajectory of established Medicaid policy and state administrative capacity. Drawing on past research, we constructed an additive scale to measure the existence and generosity of past state Medicaid programs toward children, pregnant women, working parents, the medically needy, and childless adults. 17,45 Each group was scored on a scale of 0 to 2, with higher scores representing more generous benefits; we created a composite measure by combining these results.

Tracking state administrative capacity in the area of health policy is daunting; we built a rough gauge by cumulatively measuring the capacity of states in a common area of responsibility (insurance oversight) and more specific capabilities related to aiding the poor and vulnerable, such as opening high-risk pools to the medically needy. Critically, the measure shows strong face validity: states considered to have high administrative capacity in past research, such as Massachusetts and Connecticut, scored considerably higher than did states considered to have lower administrative capacity, such as Alabama. 22

We built the final set of variables on previous research by measuring the number of individuals registered to lobby on health policy in the most recent year publicly available in each state for 3 sectors: businesses and professional organizations, unions, and public interest and nonprofit organizations. 39 Registering to lobby is an indicator of participation in the legislative process—attending hearings, directly pressuring individual lawmakers, or supporting a political action committee. Our analysis concentrated on the lobbyists working for organizations with the most intense interests in health policy rather than on umbrella business associations from outside health policy, which harbored diffuse interests in health care. 46,47

Our data from each state’s roster of registered lobbyists identified 7300 lobbyists who worked for businesses or professional health organizations, 1675 contracted by health advocacy organizations, and 226 for unions in the health sector. We counted a lobbyist who worked for multiple organizations toward both lists. Because of concerns that simple counts alone may serve as a proxy for state size, the variables in our analysis used a measure of the number of each type of lobbyists per 100 000 residents in each state, as collected by the 2010 US Census. We collected the data for all our measures as states started to adopt Medicaid expansion, and we updated the data as they became available (3 of our measures have been updated to 2015—our dependent variable, party control, and state affluence).

RESULTS

Initial results support the party rules account of state decisions regarding Medicaid expansion. Our measure of state Medicaid expansion, which is presented in Figure 1 , shows that more Democratic-controlled states than Republican-controlled states fully adopted the reform. Further support for the importance of party comes from bivariate correlations—there is a strong and positive correlation between Medicaid expansion and party control of state government (Pearson r = 0.64; P < .01).

Relative State Movement Toward Adopting Medicaid Reform: US States, 2015

Note. 1 = state application for federal planning grant; 2 = governor’s public announcement of support for Medicaid expansion; 3 = state seeks federal waiver for Medicaid expansion; 4 = federal approval of state’s waiver request as Affordable Care Act (ACA) compliant; 5 = adoption of Medicaid expansion as stipulated under the ACA. White bars represent full Democratic control of state government, light gray bars represent partial Republican control of state government, and dark gray bars represent full Republican control of state government.

Figure 1 also shows, however, that partisanship fails to fully account for the variation in state decision-making by Republican lawmakers. Although all the states that most firmly opposed Medicaid expansion (scoring 0 or 1 in Figure 1 ) were controlled by a Republican governor and legislature, other states with unified Republican control adopted the full Medicaid expansion (e.g., Arizona, Ohio, North Dakota) or designed a waiver for an approach that received federal approval (i.e., Arkansas, Indiana, Iowa, and Michigan). Bivariate analysis points to alternative influences on state decisions about Medicaid expansion. Medicaid expansion is positively associated with economic affluence (Pearson r = 0.33; P < .05), past Medicaid generosity (Pearson r = 0.51; P < .01), administrative capacity (Pearson r = 0.28; P < .05), and unions (Pearson r = 0.37; P < .05). Health advocacy organizations and professional and business organizations have no significant bivariate relationship with Medicaid progress.

Multivariate regression analysis—specifically, ordinal logistic regression—is needed to sort out the competing effects of our independent variables on state Medicaid expansion. Ordinal logistic regression is justified because the dependent variable (relative state progress in implementing Medicaid expansion) is measured along 6 ordered points. Our multivariate analysis, which is presented in Table 1 , shows that political parties are a potent force in state Medicaid expansion, whereas several other theoretical accounts lack support. In particular, increasing Democratic control of state government is a positive and significant predictor of state decisions to progress toward adopting Medicaid expansion.

TABLE 1—

The Effects of Lobbying, Affluence, Party, and Inertia on Relative State Movement Toward Adopting Medicaid Reform: US States, 2015

Note. We used ordinal logistic regression to obtain estimates because the dependent variable is along 6 ordered points.

Table 1 shows that state party control is a positive and statistically significant effect on Medicaid expansion, whereas state economic circumstances, which advocates believed would lure states to accept the federal government’s generous financing, failed to reach significance at the .05 level. This finding suggests that the pull of partisanship overwhelms the lure of Medicaid’s financing and contradicts the expectations of the ACA’s advocates and previous research. In addition, state decisions on Medicaid expansion were not markedly driven by path dependence. The variable for existing administrative capacity was not statistically significant, indicating that state decisions on Medicaid adoption were not locked in by the inertial dynamics of past institutional competence. A state history of providing generous health care for the vulnerable also failed to reach the .05 level of statistical significance.

The party rules account does neglect, however, the effects of organized associations. The lobbyist variables in Table 1 detect a new dimension in the state battle for Medicaid expansion that extends beyond partisanship. We found strong evidence that the presence of professional and business groups contributes to blocking Medicaid expansion (ordered log odds of −1.176; P < .05), which confirms a longstanding finding of the impact of private interests on government policymaking. In substantive terms, higher concentrations of business and professional lobbyists in state capitols retarded state decisions to adopt Medicaid. Unions however, failed to exert a statistically significant effect. This reflects past research and unions’ generally waning influence on states.48,49

What is surprising, however, is the positive and significant finding for public interest advocates. As the number of lobbyists for these often outnumbered groups increased relative to state population, states moved closer to the full adoption of the Medicaid expansion stipulated by the ACA. This finding is particularly impressive because it holds up alongside powerful controls for alternative influences on lawmakers. Even amid the potentially confounding effects of partisan polarization, economic affluence, the inertial dynamics within states, and the well-funded pressure from businesses and professionals, lobbyists for public interest groups still wielded influence.

DISCUSSION

More than 3 million adults lack the coverage funded by the new Medicaid expansion because states chose not to adopt it and its generous matching formulas. The party rules account that has dominated analysis and commentary over health reform suggests that these uninsured Americans have little hope of gaining coverage in states where Republicans wield power and that the safety net providers who treat them face dire financial prospects. Our findings indicate, however, a quite different scenario: reformers can take concrete steps to widen the reach of Medicaid expansion in states with Republicans in power. Well-organized public interest groups have succeeded in states where they deploy lobbyists.

Although public interest advocates have opportunities to influence health reform, business and professional associations enjoy advantages that present 2—potentially correctable—challenges to further widening the number of states adopting the Medicaid expansion. First, public interest advocates often lack the resources to marshal even a rudimentary presence in state capitols. According to our analysis, business and professional lobbyists formed a disproportionate share of lobbyists—they accounted for 146 lobbyists in each state on average with no state armed with fewer than 25. By contrast, public advocacy lobbyists were severely outmuscled: some states had as few as 4, and they averaged just 34 across all states. The lobbyists for private interests outmatched them more than fourfold. Expanding support for public advocacy lobbyists in states where they are not meaningfully present is a concrete step toward advancing Medicaid expansion across the country—even in states with Republican lawmakers.

Second, business and professional groups currently enjoy a more consistently influential presence in state capitols, but they are vulnerable to divisions that dissipate their resource advantages. The initial trepidations toward health reform among certain large insurers and businesses (e.g., UnitedHealth and Walmart) shifted as Medicaid expansion has progressed in ways that bolstered their bottom lines and failed to hurt them as they feared. 50,51 The prospects for expanding Medicaid in states that have thus far refused and bringing coverage to the millions of vulnerable citizens without health care coverage can be improved with concrete attention to the nature and sources of influences on state policymaking.

This study has several limitations. First, the fast-paced and dynamic nature of state action on health reform necessitates caution when considering the durability of these effects. As government officials around the country learn about the impacts of the ACA on the coverage of vulnerable populations, hospital finances, and state budgets, the influence of lobbyists over Medicaid expansion may strengthen or wane. In addition, if Republicans take control of the White House and Congress in the 2016 elections and proceed to reduce funding for Medicaid expansion and widen the leeway for state waivers, the positions of lobbyists advocating on health issues in state capitols may change.

Finally, our selected variables are derived from existing empirical research and theory; future research might also consider whether public opinion and perhaps other factors exert an independent effect on state Medicaid expansion decisions.

ACKNOWLEDGMENTS

We would like to thank Virginia Gray, Colleen Grogan, David Jones, Jonathan Oberlander, and Frank Thompson for their valuable advice on this project. We remain responsible, of course, for the analyses and interpretation.