Policy Diffusion across Disparate Disciplines: Private and Public-Sector Dynamics Affecting State-Level Adoption of the ACA

Abstract

Introduction

The 2010 Patient Protection and Affordable Care Act (ACA) entails a number of provisions that are profoundly changing who accesses medical care and how it is provided in the United States. Many of the law's provisions and subsequent regulations allow, indeed require, individual states to make decisions about whether and how to participate. This includes the expansion of Medicaid and the development and maintenance of health insurance exchanges. As other contributors to this special issue suggest, there is considerable interest among policy makers and health policy researchers regarding whether, which, and when states adopt these provisions, the determinants of these decisions, and their impact on individual and aggregate well-being.

In this commentary, we argue that the decision to adopt specific ACA coverage provisions by the states must be viewed within a larger ecosystem. This ecosystem has two main components: (1) complementary and competing state and federal policies; and (2) medical provision by a variety of private actors. As many other scholars have pointed out, including those in this special issue, state policy diffusion operates within a larger federal system and within state and interstate dynamics (Karch 2007, 2012; McCann, Shipan, and Volden 2015). However, not as many scholars have detailed how the diffusion of specific policies may interact with a large set of simultaneously launched or existing local, state, and federal policies; nor have the potential intended (and unintended) impacts of private actors been fully considered, namely, the panoply of medical care providers and health care organizations that operate within the health care system. Like all interest groups, health care providers have their own objectives and are subject to specific state and federal policies, such as reimbursement and other regulatory requirements. However, due to ACA enactment, hospitals and local medical providers face decisions about whether to participate in alternative payment arrangements, such as an Accountable Care Organization, and to what extent to care for additional Medicaid and/or exchange-insured patients. The diffusion of decisions made by these providers in the private sector has important implications for the diffusion of policy decisions made by leaders in the public sector. This is importantly separate from any direct lobbying they do in state capitals. Viewed from this vantage point, empirical study of the diffusion of ACA-based coverage expansion decisions must consider how federal policy creates a new private-sector environment at the state level, and how this then impacts state decision making regarding ACA policy.

Political scientists have examined the impact of private-sector decision making on state policy designs and decision making. Titmuss (1951), for example, long ago highlighted the important role of the private sector in the provision of welfare. More recently, Hacker (2002) highlights what he calls a “Divided Welfare State” where government provides direct public provision through programs such as Medicare and Social Security, and indirect private provision through taxation policies such as the Earned Income Tax Credit or tax exemptions for Employer-based Health Insurance. Hacker details how government both encourages private-sector involvement and is impacted by its involvement, and examines the unique politics that emerges under private provision. Morgen and Campbell (2011) also highlight another form of private-sector impact in their book, The Delegated Welfare State, by analyzing how government contracts with private actors similarly encourage involvement in policy, which in turn creates a new private-sector environment that impacts the political process of Medicare policy making. In particular, private-sector involvement in Medicare serves several purposes, including a way for policy makers to mask the role of government; it also gives government leaders an opportunity to win buy-in from potential allies who have helped a given reform succeed. By building on these insights and drawing on research from health economics, which focuses on how public policy impacts private actors, we suggest two main areas for further consideration in future research: (1) studies of policy diffusion should move beyond the focus on the diffusion of one policy to incorporate the reality that states must consider multiple policy decisions all at once—what we call “policy packages”; and (2) we need to consider how federal policy directly changes the private marketplace, which then interacts with and impacts state policy decisions. We discuss each of these points in the context of the ACA in the following sections.

The Diffusion of Policy Packages

One of the most underappreciated challenges of understanding the determinants of ACA state coverage policies is the fact that there were and are multiple specific ACA policy decisions simultaneously under consideration. For example, after the passage of the ACA, states had to make high-level decisions about whether to run their own exchange, and subsequently a myriad of programmatic decisions about what type of exchange to establish and how to do this. States also had to decide whether to expand Medicaid, and along with the Medicaid expansion decision they had to decide how to define an essential health benefits package and establish new rate review procedures. The ACA also triggered a number of decisions not explicitly spelled out in the law but implicit in the changing terrain, including whether to continue the temporary increase in Medicaid reimbursement rates for primary care physicians, and how to regulate changes in health care delivery and financing.

These policy decisions were not isolated from each other. While the political process of each decision might appear separate because regulatory policies with low public salience are made largely within state agencies and highly salient policies involve complicated legislative politics, the dynamics of one policy debate frequently spilled over to debates about other policies. Grogan (1994), for example, has shown how the dimensions of Medicaid policy interact where generous benefit policies have an impact on eligibility policy through budgetary trade-offs. Similarly, focusing on the diffusion patterns of each individual ACA policy may miss the complex interactions between policies—not just through the budget constraint but by creating a changed political dynamic. For example, interest groups in New Hampshire wanted the state to extend the Medicaid fee increase, but chose not to bring it up on the political agenda because they cared more about Medicaid expansion and did not want to confuse or dilute their message framing to legislators (Wilk, Evans, and Jones 2016). Idaho is another illustrative example of political interactions between ACA policies. The political capital and time spent on passing the exchange legislation was so significant that there was nothing left when the Medicaid expansion legislation was introduced in 2013 or since (Jones, forthcoming).

Further, just as the collection of ACA policies should be thought of as a policy package, we should consider how federal planning grants shift the baseline conditions in each state. Whether a state receives and adopts a planning grant will change the capacity facing each state to adopt future policies—not just bureaucratic capacity but private-sector preparedness, as many of the planning grants intended. The grants to establish health insurance exchanges are an important example. The federal government badly wanted states to run their own exchanges, and so offered a $1 million planning grant and open-ended, noncompetitive establishment grants. It also offered a small number of states approximately $30–45 million for an “early innovator grant.” The goal was for a handful of states to take the lead on developing exchange technology and then to share lessons with the rest of the country. State leaders were involved in regular conference calls and meetings to learn from each other about what to put in these grants and how to move forward. At the same time, groups opposed to the ACA worked to undermine these grants. Florida and Louisiana led the resistance, with each governor deciding in early 2011 to return the planning grants they had just received. The governors of Oklahoma, Kansas, and Wisconsin soon returned their early innovator grants. In this way, state decisions regarding federal planning grants change the baseline conditions in each state for consideration of other ACA policies in subsequent years. All of these decisions are closely intertwined and have to be examined as such.

The Larger Ecosystem

State policy diffusion should also be viewed within a larger ecosystem of medical care organization and delivery. Health care providers are subject to specific state and federal incentive programs and regulatory policies, and the way they react to these policies shapes the private delivery system, which in turn impacts state policy decision making.

There are numerous examples in the health policy and economics literature suggesting the diffusion of certain types of medical care organization and delivery can be both an anticipated and unanticipated product of state or federal policy. Section 340B of the 1993 Public Health Service Act was intended to provide assistance to medical providers who serve poor, underinsured patients, and is a good example of how federal policy impacts the private market, which in turn impacts state policy. For purposes of illustration, we detail this example below and describe why it is important for state policy diffusion research.

The 340B program provides enrolled hospitals and other providers with deep discounts on the acquisition costs of outpatient drugs, whether those drugs are later administered by physicians or dispensed by pharmacies, to enable underfinanced medical providers to purchase otherwise expensive drugs for the outpatient treatment of their patients. By statute, the program does not require 340B entities to pass on the drug discounts to the patients they treat, or to the insurance plans that cover those patients. Neither does it require these entities to limit the patients who receive the discounted drugs to those who are poor and in need. Instead, 340B entities, alone or via their contract pharmacies, can dispense discounted drugs to all their patients (except, in some cases, those insured by Medicaid), and keep the profits they make when they bill insurers and patients for the drugs as if they had purchased them at full price.

Critics speculate that the opportunity to profit from this provision has created an impetus for 340B-qualified hospitals to push the envelope on the program's intent—by opening outpatient clinics or pursuing affiliations with outpatient clinics in affluent communities where most patients are well-insured. By so doing, hospitals increase their opportunity to profit from dispensing discounted drugs while being reimbursed at retail rates, but divert from the goal of the program, which is to provide services to the poor. Conti and Bach (2014) empirically evaluated this contention, using nationally representative data on program participants in 2012, matched to US Census Bureau data on local communities' socioeconomic characteristics. They found that 340B-qualified hospitals are expanding their base into communities that tend to be affluent and well-insured, consistent with the most profitable expansion strategy that counters the objectives of the program. Moreover, these activities drive up costs of providing care—and ultimately, commercial insurance premiums—since hospital outpatient contracts tend to be much more generous than physician office contracts and charge facility fees on top of service charges to payers and patients.

Similar to this example, there are a handful of ACA provisions that may have intended and unintended consequences for medical care organizations. For example, the ACA requires providers to collect and report quality metrics for their Medicare beneficiaries. The quality metrics attempt to encourage cooperation across different types of care, such as mental and physical health. The ACA also uses financial incentives to encourage providers to take care of patients in high-value, low-cost settings (e.g., outpatient over inpatient provision of care). Still further, the ACA encourages—through federal Medicare policy incentives—the use of alternative organizational structures for the practice of medicine, including but not limited to Accountable Care Organizations (ACOs) (Bekelman et al. 2013). The main idea behind this delivery model reform is that payers are able to reimburse defined, predictable payments for each patient for a set period of time, and providers have the freedom to practice medicine without being micromanaged by payers. Clearly, the intent of these policies is to change the US delivery system for the better. While the verdict is still out on whether the delivery system has substantively improved, we do know that the ACA has brought on enormous private-sector changes, including a number of new organizational forms—such as the adoption of ACOs—as hoped. Clearly, this changed private-sector landscape has important implications for state policy decision making regarding how to work with private insurers on the exchanges and in their Medicaid programs, such as how to set network requirements given changes in the supply of different providers and organizational types.

Conclusion

In sum, this commentary seeks to contribute to future research by highlighting two key areas for further consideration. First, the ACA is a major piece of legislation with multiple policies. Most studies of policy diffusion tend to focus on the diffusion of one or two policies as separate processes; the problem, especially in the case of the ACA, is that states need to consider multiple policy decisions all at once, and a decision in one domain almost certainly impacts the decision in another domain. We need to take account of the reality that states are adopting “policy packages.” Therefore, future research should theorize and model the diffusion of policy packages. Second, future research should incorporate how ACA policies are targeted at private actors and the ultimate impact on state decision making. In particular, a large set of federal regulatory and incentive policies have changed private behavior, creating a new private delivery system environment, which in turn has an impact on state policy decision making.

Conti's effort on this article was generously funded by the Robert Wood Johnson Foundation and the Health Care Cost Institute in collaboration with the National State Health Policy Association and the John and Laura Arnold Foundation. Neither the funders nor the University of Chicago are responsible for the conduct of the study, model interpretations, or results.

References