Your current web browser is outdated. For best viewing experience, please consider upgrading to the latest version.
Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
search DONATE
Close Nav

How to Increase Health-Insurance Coverage by Reducing ACA Crowd-Out

back to top

How to Increase Health-Insurance Coverage by Reducing ACA Crowd-Out

January 23, 2018
Health PolicyAffordable Care Act

Executive Summary

Public health-care entitlements in the U.S. have traditionally been designed to supplement rather than to supplant privately purchased health insurance. About 40% of the entitlement funds disbursed under the Affordable Care Act (ACA), however, have gone to individuals who already had private coverage. This displacement of private-sector spending by public-sector activity is called “crowd-out.” While the ACA has reduced the share of the American population without health insurance, its spending has been poorly-targeted to fill gaps in care, and 28 million remain uninsured.

This paper reviews estimates of ACA crowd-out and examines the potential for block grants to allow states to target assistance at individuals otherwise lacking coverage. Under such a reform, the same level of federal funding could do more to expand access to care and to provide protection from catastrophic medical costs for those who need help the most.



Chris Pope is a senior fellow at the Manhattan Institute. Follow him on Twitter here.

Executive Summary

Public health-care entitlements in the U.S. have traditionally been designed to supplement rather than to supplant privately purchased health insurance. About 40% of the entitlement funds disbursed under the Affordable Care Act (ACA), however, have gone to individuals who already had private coverage. This displacement of private-sector spending by public-sector activity is called “crowd-out.” While the ACA has reduced the share of the American population without health insurance, its spending has been poorly-targeted to fill gaps in care, and 28 million remain uninsured.

This paper reviews estimates of ACA crowd-out and examines the potential for block grants to allow states to target assistance at individuals otherwise lacking coverage. Under such a reform, the same level of federal funding could do more to expand access to care and to provide protection from catastrophic medical costs for those who need help the most.

The Challenge of Crowd-Out

Policymakers often talk as if the U.S. has a relatively “small government” approach to health-care policy. This is not the case. U.S. public spending on health care (8.3% of GDP in 2014) is already among the highest in the world—toward the upper end of a group of large developed countries that ranges from Spain (6.4%) to France (9.0%) (Figure 1).

Most developed countries provide a universal entitlement to publicly subsidized health care. The U.S., however, has traditionally limited public health-care assistance to those who could not be expected to fund their own insurance—specifically, the elderly, the disabled, and certain low-income families.[1] If assistance is not targeted, public-health spending mostly goes to displace coverage that otherwise would be purchased by able-bodied middle-class workers, rather than filling in gaps in coverage. The result has been that while public health-care spending in other OECD countries has mostly served to supplant private spending on health care, government spending on health care in the U.S. has largely supplemented it.

Despite popular perceptions, this approach has let the U.S. provide more generous healthcare benefits to its neediest citizens than are available to the middle class in countries with single-payer systems.[2] In the U.K. and Canada, similar public health-care budgets must bear the costs of treating all citizens—restricting access to essential care, specialty physicians, and cutting-edge treatments across the board. By contrast, the limits on eligibility for Medicaid and the Children’s Health Insurance Program (CHIP) have ensured that 74 million[3] low-income Americans can receive pathbreaking drugs, hospital care from the finest facilities, and treatment from an array of specialists—without more than nominal out-of-pocket expenses or such lengthy waiting lists.[4] But the absence of public assistance for able-bodied working-age adults had left many, such as those with preexisting conditions, unable to purchase insurance coverage.

The ACA sought to fill these coverage gaps with a combination of expanded entitlement spending and regulations on insurance pricing. While increased insurance coverage is credited for improving medical outcomes,[5] much of the spending under the law did not provide insurance to those who lacked it, instead assuming costs associated with existing private health-care utilization—a phenomenon known as “crowd-out.”[6]

Thanks to the ACA, the number of people who have enrolled in Medicaid or purchased individual health-insurance policies has increased. But some people who had already been purchasing insurance on the individual market stopped doing so as premiums soared. Some people abandoned their previous health plans for Medicaid, or shifted to subsidized plans on the exchange to enjoy lower cost-sharing. Some employers stopped providing health insurance to part-time workers; other employers reduced their projected hiring, while some individuals who had taken jobs to qualify for employer-sponsored coverage no longer needed to do so.[7] Trevor Gallen of Purdue and Casey Mulligan of the University of Chicago predicted that enrollment in employer-sponsored insurance would eventually decline by 20 million as a consequence of the ACA.[8]

Health-insurance crowd-out is not a new or an unstudied phenomenon. In 1995, David Cutler of Harvard and Jonathan Gruber of MIT sought to quantify the crowd-out effect brought about by the expansions of Medicaid between 1987 and 1992 to low-income children and pregnant women.[9] Their study provoked an extensive research literature, employing a host of statistical methods, identification strategies, and data sources. This research has yielded a broad range of crowd-out estimates for various policy changes and different sections of the population.[10]

For example, summarizing the estimates of crowd-out resulting from the creation of CHIP, the Congressional Budget Office (CBO) in 2007 concluded that “the reduction in private coverage among children is most probably between a quarter and a half of the increase in public coverage.”[11] Overall, research has consistently found greater crowd-out for higher-income individuals, as they were more likely to have had private insurance coverage.[12]

Assessing Crowd-Out from the ACA

Several provisions of the ACA likely increase subsidized insurance at the expense of unsubsidized insurance: the expansion of Medicaid to cover low-income but able-bodied adults; subsidies to individuals who buy insurance plans on the exchanges; and various taxes on private insurance plans. The effects of these provisions may be masked or offset by others, which may increase overall enrollment in unsubsidized insurance, such as penalties for firms that fail to cover their employees and the mandate that individuals acquire insurance.[13] The net extent of crowd-out resulting from the ACA is therefore not clear a priori and needs to be assessed empirically.

Aggregate enrollment shifts

We can make a simple initial estimate of crowd-out by comparing the growth of publicly subsidized insurance with the decline in the proportion of the population uninsured. To do so, it is crucial to note that the ACA was enacted in March 2010—when the unemployment rate was 9%—and has been implemented gradually during years when the unemployment rate declined. This makes it necessary to disentangle the effects of the ACA on insurance coverage from those of the business cycle. A prima facie estimate of crowd-out can therefore be obtained by comparing enrollment in subsidized and unsubsidized insurance in 2007 with 2017—two years with unemployment rates of 4.5% (Figure 2).[14]

According to data from the Kaiser Family Foundation (KFF) and the CBO, the proportion of the non-elderly U.S. population with publicly subsidized insurance increased from 14% to 31%, while the proportion of uninsured fell from 17% to 10%—implying that 57% of the increase in publicly subsidized insurance between 2007 and 2017 was offset by a decline in unsubsidized private insurance. If the CBO enrollment projections are correct, this 57% metric of crowd-out can be expected to increase to 66% by 2027, as the ACA is more fully implemented.[15]

Yet between 2007 and 2017, many other confounding factors occurred (both potentially inflating and deflating the estimate), including slowing wage growth, the shift by employers toward high-deductible health plans, the aging of the population, immigration, and other public policy changes. While 57% may serve as a rough high-level estimate of the crowd-out associated with the ACA, a more precise and detailed assessment is required.

To avoid the uncertainties associated with a broad, 10-year window, the Urban Institute compared 2013 and 2014. Its paper, “Employer-Sponsored Insurance Stays Strong with No Signs of Decay Under the ACA,” concluded that the share of workers earning below 250% of the federal poverty level (FPL) receiving employer-sponsored insurance rose from 70.1% in 2013 to 70.3% in 2014.[16] But this comparison fails to control for the confounding effect of the economic recovery.[17]

KFF found that the share of workers earning below 250% of the FPL receiving employer-sponsored insurance fell from 40% in 2013 to 39% in 2014 in states that expanded Medicaid under the ACA, while it rose from 41% to 45% in non-expansion states. Medicaid enrollment grew as a share of this population by 8 percentage points in expansion states and 1 percentage point in non-expansion states, while the uninsured fell by the same 9 percentage points in both. This suggests that much of the Medicaid expansion offset much of an increase in employer-sponsored insurance (ESI),which otherwise would have occurred. Yet the statistics generated by KFF as well as the Urban Institute fail to account for the business cycle altering the proportion of the overall population earning below 250% of FPL, which further skews both comparisons.[18]

Nonetheless, comparing expansion with nonexpansion states is potentially useful for the sake of disentangling causal factors. Indeed, it is necessary to distinguish the direct effect of changes in the law from the “welcome-mat” effect—by which the publicity surrounding the expansion of Medicaid, the ability of medical providers to presumptively enroll eligible lowincome patients in Medicaid, the automatic diversion of Medicaid-eligible individuals shopping for exchange plans, and the simplification of Medicaid eligibility criteria encouraged increased enrollment among previously eligible groups such as low-income disabled individuals or pregnant women.[19]

Comparing expansion states with non-expansion states

Between 2013 and 2014, 21 states expanded Medicaid (five states and D.C. had done so between 2011 and 2013), while 24 states did not. According to the Census Bureau’s Current Population Survey (CPS), the proportion of the population enrolled in Medicaid increased by 2.9 percentage points in 2014-expansion states and by 1.3 percentage points in states that had not expanded Medicaid in 2014. Meanwhile, the share of the population uninsured declined by 3.4 percentage points in states that expanded Medicaid in 2014 and 2.4 percentage points in non-expansion states.

In states that did not expand Medicaid eligibility, therefore, Medicaid enrollment grew by 45% of the amount that it grew in states that did. But the proportion of the population uninsured declined by 38% less than the expansion of the Medicaid program in expansion states relative to non-expansion states. This may serve as an estimate of crowd-out but is likely skewed by the fact that states that expanded Medicaid already had lower numbers uninsured, and hence less scope to reduce them further. Rates of private insurance in non-expansion states may also be inflated by the availability of exchange subsidies to individuals earning 100%–138% of FPL in non-expansion states.

Several analyses have attempted to disentangle the effects of the 2014 Medicaid expansion from other provisions of the ACA, and crowd-out estimates can be calculated from their regression tables. Figure 3 summarizes these estimates, using two parallel definitions of crowd-out, provided in the 1995 paper by Cutler and Gruber:[20]

Crowd-out statistics calculated with method B (the growth of the Medicaid expansion in excess of the reduction in the share of the population uninsured) are consistently slightly higher than those calculated with method A (the decline of the share of the population privately insured as a share of the Medicaid expansion)—possibly because some individuals become covered by Medicaid without losing employer-sponsored insurance.

While several of the above findings of crowd-out fall short of statistical significance, the findings from a variety of comparisons between expansion and non-expansion states cluster around a crowd-out effect of 25%–30%.

Early expansion states

In 2014, the ACA simultaneously revolutionized insurance-market rules, expanded Medicaid, and established subsidies for the purchase of insurance on the exchange. We can distinguish the crowd-out effects of Medicaid expansion from those of other ACA provisions by examining states that expanded Medicaid before 2014.

A study of early Medicaid-expansion states in Health Affairs found a crowd-out rate of 40% in Connecticut’s 2010 expansion of Medicaid, three-fourths of which was displacement of employer-sponsored insurance; one-fourth was a displacement of people who previously had insurance in the individual market. “Nearly all” of the Medicaid expansion among young adults served to displace private coverage.[21]

Massachusetts expanded its Medicaid program in 2007, as part of a reform (often dubbed “Romneycare”) that many saw as a prototype for the ACA—and so the impact of its Medicaid reforms is harder to gauge in isolation. However, a Cato Institute study of Romneycare found a 15% drop in the proportion of children and a 6% decline in that of adults in low-income households (under 150% of FPL) who were privately insured in Massachusetts relative to other New England states.[22]

Following the implementation of the ACA, Massachusetts took advantage of the enhanced federal matching rate to further expand its Medicaid program from 19.4% of its population in 2013 to 23.6% in 2014.[23] Yet the proportion of its population uninsured actually rose from 3.6% to 4.4%, despite the state’s economy recovering in line with national trends. Casey Mulligan has argued that Obamacare imposed work disincentives 14 times greater than Romneycare.[24] From coverage and fiscal standpoints, Obamacare represents over 100% crowd-out relative to Romneycare.

Exchange subsidies

Though several studies have assessed crowd-out of private coverage by the Medicaid expansion, much less attention has been paid to crowd-out due to the premium and cost-sharing reduction subsidies available for the ACA-compliant insurance available through the state health exchanges. In 2007, 14 million people were enrolled in individual market plans, all of them unsubsidized.[25] In 2017, 8 million people were enrolled in unsubsidized individual market plans, while 9 million received subsidies through the exchanges.[26] This suggests that a large fraction of the subsidized policies had replaced prior unsubsidized policies—an implication confirmed in the 2016 study by Charles Courtemanche and colleagues, which found that 73% of the 8 million people purchasing insurance on the exchanges in 2014 had unsubsidized coverage in 2013.[27]

This much higher crowd-out estimate for exchange-subsidy recipients (100%–400% of FPL) than for Medicaid enrollees (below 138% of FPL) aligns with the fact that individuals in that higher income cohort were three times as likely to have previously had private insurance coverage. (Figure 4)

Regulatory changes

In a 2014 NBER paper, Jeffrey Clemens of the University of California at San Diego argued that the Medicaid expansion could “crowd in” individuals to private coverage—reducing the premiums charged to individuals achieved by taking higher-risk individuals out of a community-rated individual market’s risk pool.[28] But if the Medicaid expansion did so increase private enrollments, it was only by mitigating the effect of the ACA’s own individual-market regulations, which forced healthier insurance customers to pay above actuarilly fair rates in order to subsidize less healthy ones.

Furthermore, Clemens assumed that Medicaid-expansion participants would resemble the disproportionately disabled pre-expansion Medicaid population, which cost an average $8,300 per person in 2011.[29] But the Medicaid expansion was focused on low-income able-bodied adults, who are disproportionately young and healthy. As a result, the average Medicaid spending per expansion beneficiary in 2016 was about $5,800.[30] Indeed, subsequent analyses of the ACA’s Medicaid expansion have tended to find the highest rates of crowdout among the young and healthy.[31]

Through crowd-out, the ACA’s Medicaid expansion may, if anything, have made the individual market’s risk pool more expensive—an effect exacerbated by the ACA’s “dependent care mandate,” which requires employer plans to cover adult children under age 26. A 2016 study estimated that 30%–100% of enrollment resulting from the dependent care mandate replaced preexisting coverage.[32] Both these effects would have served to further reduce enrollment in the post-ACA individual market.

It has been suggested that the 40% “Cadillac Tax” imposed (from 2020) on employer plans exceeding $10,200 for individuals and $27,500 for families may shift an increasing number of individuals from ESI to the exchange.[33] But this ought not to be seen as a form of crowd-out because rather than costing the government money, it mitigates the fiscal effect of the ESI tax exemptions.

Increasing effects over time

The above assessments of crowd-out under the ACA are likely to underestimate its full effects. It usually takes several years for high-quality data on health-insurance enrollments to be collected, so existing research findings are currently limited. Those who were sickest and lacked prior coverage may likely have been the first to enroll in ACA plans; subsequent enrollees are thus more likely to have had previous insurance, and hence be subject to a higher rate of crowd-out.

Furthermore, it takes time for incentives to have their full effect in shaping behavior. For instance, individuals who would otherwise move for a job to get health insurance, but gained Medicaid as a result of the program’s expansion, may not have their behavior visibly altered until the next recession. The program’s aggregate enrollment tends to ratchet up to permanently higher levels during economic downturns.[34]

Increasing coverage by reducing crowd-out

Congress should consider block-granting the ACA’s entitlement funding to states in proportion to the total number of low-income individuals with health insurance in each state.[35] This approach would allow most states to maintain the tax, subsidy, and regulatory arrangements currently prevailing under the ACA if they believed them optimal. States could continue to use funds to reduce cost-sharing, subsidize premiums, and to mandate the purchase of an expanded benefit package for individuals who already had health-insurance coverage.

But block grants would also allow states to focus public funds on filling gaps, rather than forcing taxpayers to pick up the costs of services already paid for by employers or individuals—i.e., to minimize crowd-out. There are a number of ways to do this.

First, states could tighten eligibility criteria for able-bodied adults enrolled in Medicaid. Comprehensive insurance coverage with the total absence of cost-sharing or premiums may be appropriate for some able-bodied adults on a temporary basis, such as newly unemployed individuals looking for jobs. But such coverage need not be a permanent and open-ended entitlement more generous than any employer would provide—that is, an entitlement that could act as a deterrent to work. While regulatory work requirements may be too rigid or ineffective, criteria such as lifetime time limits on enrollment may ensure that the most generous assistance is reserved for those who suddenly find themselves in acute need without greatly reducing insurance funded through work.[36]

Second, states could reserve exchange subsidies for individuals who are unable to afford insurance through a fully competitive market.[37] The ACA’s community- rating regulations (requiring that insurance be priced regardless of expected risks) make it hard to allocate subsidies according to unmet needs. By encouraging those who can buy market-rate, ACA-noncompliant insurance (such as short-term limited duration plans) to leave the subsidized exchanges, policymakers can better target subsidies at individuals with preexisting conditions who need them most. This would also help those without preexisting conditions (15 million on the individual market and most of the 28 million currently uninsured)[38] receive coverage with cost-sharing more appropriate to their risk profiles, at premiums a third of those prevailing on the exchange.[39]

Third, states could integrate funds for Medicaid with subsidies for the individual market. Currently, Medicaid funds are segmented, and states are able to receive federal matching funds only for individuals eligible for the program’s comprehensive benefits. If the federal government integrated Medicaid and exchange funding as a block grant, states would find it easier to subsidize private insurance for the sake of reducing premiums or cost-sharing for low-income able-bodied individuals. Given that federal funding for the Medicaid expansion ($72 billion in 2017) well exceeds subsidies for the exchange ($34 billion), this would help ensure that public assistance serves to supplement rather than to entirely supplant private spending.[40]

Based on the extant research, we may conservatively estimate the potential savings from reducing crowdout at 25% for spending on the Medicaid expansion (that is, to individuals earning below 138% of FPL) and 70% from subsidies to those on the exchange (received by those earning 100%–400% of FPL) and the Basic Health Program (an ACA option enabling states to provide a health plan for low-income residents outside the health exchanges and Medicaid). Although no estimates of crowd-out associated with cost-sharing reduction (CSR) payments (those earning 100%–250% of FPL) have specifically been made in the research literature, one can reasonably assume that the crowd-out rate for those receiving CSRs lies between these figures (because of the relative income of those eligible). For the purposes of demonstration, we will take it as 40%.

Over the 2018–27 CBO scoring window, full reversal of the ACA’s crowd-out would amount to a total of $718 billion—or 41% of ACA spending.[41] Obviously, not all the potential savings from crowd-out caused by the ACA are likely to be realized—nor would many policymakers want them to be. But even a small reduction in crowd-out could allow states to finance a major expansion of coverage—for instance, yielding an additional $10 billion per year if only 25% were realized.

The ACA has successfully increased the proportion of Americans with health insurance. Coverage improves individuals’ access to care, use of preventive services, rates of medical diagnosis, and treatments, along with protection from catastrophic medical expenses.[42] But a large proportion of the entitlement spending established by the ACA has displaced privately funded insurance that advances these objectives just as well. This crowd-out effect can be minimized and ACA spending better targeted to fill gaps in insurance coverage—in other words, to build on private health-insurance spending, rather than to substitute so much for it.



  1. Even Medicare generally requires individuals or their spouse’s employer-sponsored insurance to pay for health-care services consumed by eligible elderly or disabled beneficiaries before it does so. See Centers for Medicare & Medicaid Services, Medicare Secondary Payer.
  2. Chris Pope, “Stretching Medicaid Too Thin,” The American Conservative, July 26, 2017.
  3., “Medicaid and CHIP Total Enrollment Chart—September 2017.”
  4. Albeit with less choice of providers than is available to those receiving most employer-sponsored insurance; see Julia Paradise, “Data Note: Three Findings About Access to Care and Health Outcomes in Medicaid,” Kaiser Family Foundation, Mar. 23, 2017; Bacchus Barua, “Waiting Your Turn: Wait Times for Health Care in Canada, 2017 Report,” Fraser Institute, Dec. 7, 2017; Sarah Neville, “NHS Waiting Lists Climb to Highest Level in a Decade,” Financial Times, Aug. 10, 2017.
  5. Benjamin D. Sommers, Atul A. Gawande, and Katherine Baicker, “Health Insurance Coverage and Health—What the Recent Evidence Tells Us,” New England Journal of Medicine 377 (Aug. 10, 2017): 586–93.
  6. Crowd-out denotes the displacement of private- by public-sector activity, such as borrowing or the provision of health-care coverage.
  7. Denver Nicks, “Trader Joe’s Explains Why It’s Cutting Health Benefits for Part Timers,” Time, Sept. 17, 2013; Ben Gitis, Conor Ryan, and Sam Batkins, “Obamacare’s Impact on Small Business Wages and Employment,” American Action Forum, Sept. 9, 2014; Casey B. Mulligan, “The Employer Penalty, Voluntary Compliance, and the Size Distribution of Firms,” NBER Working Paper #24037, Nov. 2017.
  8. Trevor S. Gallen and Casey B. Mulligan, “Wedges, Labor Market Behavior, and Health Insurance Coverage Under the Affordable Care Act,” NBER Working Paper #19770, Dec. 2013.
  9. David M. Cutler and Jonathan Gruber, “Does Public Insurance Crowd Out Private Insurance?” NBER Working Paper #5082, Apr. 1995.
  10. David Card and Lara D. Shore-Sheppard, “Using Discontinuous Eligibility Rules to Identify the Effects of the Federal Medicaid Expansions on Low-income Children,” Review of Economics and Statistics 86, no. 3 (Aug. 2004): 752–66; Anthony T. Lo Sasso and Thomas C. Buchmueller, “The Effect of the State Children’s Health Insurance Program on Health Insurance Coverage,” Journal of Health Economics 23, no. 5 (Sept. 2004): 1059–82; Jonathan Gruber and Kosali Simon, “Crowd-Out Ten Years Later: Have Recent Public Insurance Expansions Crowded Out Private Health Insurance?” NBER Working Paper #12858, Jan. 2007; Anna Sommers et al., “Substitution of SCHIP for Private Coverage: Results from a 2002 Evaluation in Ten States,” Health Affairs 26, no. 2 (Mar.–Apr. 2007): 529–37; Laura Dague et al., “Estimates of Crowd-Out from a Public Health Insurance Expansion Using Administrative Data,” NBER Working Paper #17009, May 2011; David B. Muhlestein and Eric E. Seiber, “State Variability in Children’s Medicaid/CHIP Crowd-Out Estimates,” Medicare & Medicaid Research Review 3, no. 3 (July 11, 2013): E1–E21.
  11. Congressional Budget Office (CBO), “The State Children’s Health Insurance Program,” Pub. 2970, May 1, 2007.
  12. Gruber and Simon, “Crowd-Out Ten Years Later.”
  13. Steven D. Pizer and Austin Frakt, “The Effect of Health Reform on Public and Private Insurance in the Long Run,” HCFE Working Paper 2011-03, Feb. 17, 2011.
  14. See Bureau of Labor Statistics (BLS), Labor Force Statistics from the Current Population Survey, Series ID: LNS14000000: Unemployment Rate.
  15. CBO, “Federal Subsidies for Health Insurance Coverage for People Under Age 65: 2017 to 2027,” Pub. 53091, Sept. 2017, p. 10.
  16. Fredric Blavin et al., “Employer-Sponsored Insurance Stays Strong, with No Signs of Decay Under the ACA: Findings Through March 2016,” Health Reform Monitoring Survey, Urban Institute, July 13, 2016. Those earning below 250% of FPL are entitled to additional cost-sharing reduction subsidies under the ACA, which diminish the out-of-pocket payments that individuals are expected to pay for care.
  17. The unemployment rate fell from 7.5% in June 2013 to 6.1% in June 2014; see BLS, “LNS14000000: Unemployment Rate.”
  18. Alanna Williamson et al., “ACA Coverage Expansions and Low-Income Workers,” KFF Issue Brief, June 10, 2016.
  19. Julie Sonier, Michel H. Boudreaux, and Lynn A. Blewett, “Medicaid ‘Welcome-Mat’ Effect of Affordable Care Act Implementation Could Be Substantial,” Health Affairs 32, no. 7 (July 2013): 1319–25.
  20. Cutler and Gruber, “Does Public Insurance Crowd Out Private Insurance?”
  21. Benjamin D. Sommers, Genevieve M. Kenney, and Arnold M. Epstein, “New Evidence on the Affordable Care Act: Coverage Impacts of Early Medicaid Expansions,” Health Affairs 33, no. 1 (2014): 77–87. The study’s analysis of D.C.’s similar 2010 Medicaid expansion found a weaker increase in Medicaid enrollment.
  22. Aaron Yelowitz and Michael F. Cannon, “The Massachusetts Health Plan: Much Pain, Little Gain,” Cato Policy Analysis 657 (Jan. 20, 2010).
  23. U.S. Census Bureau, Current Population Survey, CPS Table Creator.
  24. Casey B. Mulligan, “Is the Affordable Care Act Different from Romneycare? A Labor Economics Perspective,” NBER Working Paper #19366, Dec. 2013, Appendix Table 3.
  25. John Holahan and Allison Cook, “Changes in Health Insurance Coverage, 2007–2008: Early Impact of the Recession,” KFF Issue Paper #8004, Oct. 2009.
  26. CBO, “Federal Subsidies for Health Insurance Coverage,” Pub. 53091, p. 4.
  27. Charles Courtemanche et al., “Impacts of the Affordable Care Act on Health Insurance Coverage in Medicaid Expansion and Non-Expansion States,” NBER Working Paper #22182, Apr. 2016.
  28. Jeffrey Clemens, “Regulatory Redistribution in the Market for Health Insurance,” NBER Working Paper #19904, Feb. 2014.
  29. Ibid., p. 15.
  30. CBO, “Federal Subsidies for Health Insurance Coverage for People Under Age 65: 2016 to 2026,” Pub. 51385, Mar. 2016, pp. 2–3; $5,800 = $64 billion cost of Medicaid expansion divided by 11 million enrolled.
  31. George L. Wehby and Wei Lyu, “The Impact of the ACA Medicaid Expansions on Health Insurance Coverage,” Health Services Research, May 18, 2017; B. Sommers, Kenney, and Epstein, “New Evidence on the Affordable Care Act.”
  32. Gopi Shah Goda, Monica Farid, and Jay Bhattacharya, “The Incidence of Mandated Health Insurance: Evidence from the Affordable Care Act Dependent Care Mandate,” NBER Working Paper #21846, Jan. 2016.
  33. Pizer and Frakt, “The Effect of Health Reform on Public and Private Insurance in the Long Run.”
  34. “Exhibit 8: Medicaid Enrollment and Spending, FYs 1966–2016,” MACStats: Medicaid and CHIP Data Book 2017.
  35. States should be paid based on performance in efficiently using funds to increase insurance coverage—so that federal assistance is distributed among them in inverse proportion to the share of their low-income population that is uninsured, rather than in proportion to the expense they incur on behalf of federal taxpayers.
  36. Chris Pope, “Should States Be Allowed to Semi-Expand Medicaid?” Washington Examiner, Sept. 13, 2017.
  37. Chris Pope, “Insuring the Uninsurable,” National Review Online, May 2, 2017.
  38. Chris Sloan, “Proposed High Risk Pool Funding Likely Insufficient to Cover Insurance Need for Individuals with Pre-Existing Conditions,” Avalere, May 4, 2017; CBO, “Federal Subsidies for Health Insurance Coverage,” Pub. 53091, p. 4.
  39. “Study Shows Major Premium Savings for Short-Term Health Insurance as Compared to Entry-Level Obamacare Plans,” AgileHealthInsurance Report, Feb. 18, 2015.
  40. CBO,“Federal Subsidies for Health Insurance Coverage,” Pub. 53091, p. 10.
  41. Ibid.
  42. B. Sommers, Gawande, and Baicker, “Health Insurance Coverage and Health.”