Skip to main content Skip to secondary navigation

How to Heal Obamacare

Every argument for and against the Affordable Care Act came to a head this summer as opponents tried to repeal the 7-year-old law that is both loathed and lauded as Obamacare. And despite their failure during the summer, the political debate continues.

Supporters of the ACA come with their most compelling figures: The number of Americans without health insurance fell from 49 million in 2010 to 29 million in 2016 (HHS, 2017). The growth rate of health care spending slowed significantly after ACA passage. As shown in Figure 1, after rising from 13.3 percent to 17.4 percent during the 2000 to 2010 period, health care spending as a share of GDP rose to just 17.8 percent by 2015.

Figure 1. Health Care Spending as a Share of US GDP

The law’s detractors are armed with their strongest data, as well. They focused on rising prices in many of the state-based health insurance exchanges and a decline in the number of private insurers participating in the exchanges. The median premium increase from 2016 to 2017 was 19 percent. Additionally, 21 percent of exchange enrollees had just one insurer from which to choose in 2017. Just one year earlier, only 2 percent of enrollees faced such a lack of competition (Kaiser Family Foundation, 2017). These changes are consistent with earlier research that less competition in private health insurance markets leads to higher premiums (Dafny, Duggan, and Ramanarayanan, 2012).

At least one loud criticism of the law — that it was killing jobs — was discredited by research that I recently conducted with colleagues at Stanford (Duggan, Goda, and Jackson, 2017).

The ACA is far from a perfect piece of legislation. But it’s now time for policymakers to dig into the law and find ways to make improvements — even as some critics are still looking for ways to scrap it.

In this policy brief, I describe in greater detail the impact of the ACA — focusing especially on provisions that were designed to expand health insurance coverage. I also discuss some potential reforms that could address some of the key shortcomings of the legislation — focusing mainly on stimulating competition and expanding choice in the health insurance exchanges.

Medicaid Expansion

The effect of the Affordable Care Act has varied substantially across the country. That’s largely because of states’ decisions about whether or not to expand Medicaid to residents with incomes below 138 percent of the federal poverty line (FPL) — or about $34,000 for a family of four.

As of this month, only 31 states and the District of Columbia have expanded their programs.[1] Largely because of these changes, the number of Medicaid recipients in the U.S. increased from 58 million in late 2013 to 75 million this year.

The Medicaid expansions were fully funded by the U.S. government between 2014 and 2016, with the federal share of spending slowly declining to 90 percent by 2020. In expanding states, the fraction of residents enrolled in Medicaid increased from 19.2 percent in late 2013 to 26.6 percent by May 2017. The corresponding increase was much smaller in states that did not expand, with enrollment increasing from 16.3 percent to 18.3 percent during the same period.

Table 1 lists the 10 states with the largest increases in Medicaid enrollment from late 2013 until early 2017. Each state expanded its Medicaid program and especially large enrollment increases were seen in New Mexico (a 16 percentage point gain); Kentucky (more than doubled with a 15 percentage point gain); and Arkansas (12 percentage points).

These three states have poverty rates significantly higher than the national average. States with similarly high poverty rates that did not expand Medicaid had much smaller increases in enrollment. For example, Alabama, Georgia, and Mississippi had increases of just 1 to 2 percentage points.

Table 1: Percent of State Residents Enrolled in Medicaid

State Sept 2013 May 2017 Percent Change
New Mexico 22 38 16
Kentucky 14 29 15
Arkansas 19 31 12
West Virginia 20 31 11
California 20 31 11
Colorado 15 26 11
Rhode Island 18 29 11
Montana 15 25 10
Nevada 12 22 10
Washington 16 25 10

Source: CMS State Medicaid and Chip Profiles

With full repeal of the ACA now looking unlikely, some states that chose not to expand their Medicaid programs may decide to reconsider.

The states that expanded Medicaid qualified for an additional $70 billion in federal funding in 2015 (the most recent year with complete data available). The 19 states that did not expand Medicaid received no similar incremental funding. This difference in federal funding is especially significant when one considers that state and local governments historically end up paying the medical bills for those without insurance, resulting in significant drains on their budgets (Hadley, 2008).

In essence, the 19 non-expanding states have decided to forgo a partial solution to that problem. And by not taking the extra federal payments for many of their low-income uninsured, these states have significantly lowered the federal budget deficit.

Insurance Exchanges

As of early 2017, there were 12.2 million U.S. residents receiving private insurance purchased through the state-based health insurance exchanges. This represents 4.5 percent of Americans below the age of 65, as hardly any seniors get their coverage through these exchanges.

Enrollees with incomes between 100 and 400 percent of FPL are eligible for subsidies through the exchanges.

As shown in Figure 2 for the median county in the U.S., these subsidies generally decline with income and rise with age. The majority (71 percent) of exchange enrollees receive subsidies for this coverage, with the average annual subsidy amounting to $4,452. Total federal subsidies for coverage purchased through the ACA exchanges were $38.8 billion, which represents about 1.2 percent of total U.S. health care spending.

The fraction of each state’s residents obtaining coverage through the exchanges varies even more dramatically than does the share on Medicaid.

The clear outlier among all states is Florida, with 10.8 percent of residents younger than 65 enrolled in the state’s ACA exchange as of early 2017. Other states with especially high enrollment are Maine, Utah, Idaho, and North Carolina. Interestingly, none of these five states elected to expand their Medicaid programs.

Table 2 lists the 10 states with the highest fraction of their non-elderly populations enrolled in the ACA exchanges in 2017.

Table 2: Percent of Non-Elderly Residents Enrolled in ACA Exchanges in 2017

State Percent Enrolled Expanded Medicaid
Florida 10.8 No
Maine 7.4 No
Utah 7.4 No
Idaho 7.1 No
North Carolina 6.4 No
Montana 6.3 Yes
Vermont 6.0 Yes
Virginia 5.8 No
South Carolina 5.8 No
Georgia 5.6 No

Source: Kaiser Family Foundation Total Marketplace Enrollment

In contrast, the five states with the smallest share of their residents getting insurance through the exchanges — New York, Hawaii, Kentucky, West Virginia, and Minnesota — all did expand their Medicaid programs. [2]

This is consistent with the findings of recent research that private health insurance coverage increased by substantially more in the states that elected not to expand Medicaid (Duggan, Goda, and Jackson, 2017). It also implies that any changes in the generosity of ACA exchange subsidies for coverage purchased through the exchanges would differentially affect individuals in states like Florida, Utah, and North Carolina that did not expand their Medicaid programs.

Overall Coverage

An examination of data from Gallup (2017) sheds light on which states have experienced the largest increases in overall health insurance coverage since just before the implementation of the key features of the ACA almost four years ago. Interestingly, all 10 of the states with the largest reductions in the share of adults who were uninsured from 2013 to 2016, as shown in Table 3, expanded their Medicaid programs.

Table 3: Percent of Adult State Residents Uninsured in 2013 and 2016

State 2013 2016 Percent Change
Kentucky 20.4 7.8 -12.6
Arkansas 22.5 10.2 -12.3
West Virginia 17.6 6.1 -11.5
New Mexico 20.2 9.0 -11.2
California 21.6 10.5 -11.1
Oregon 19.4 9.1 -10.3
Washington 16.8 7.2 -9.6
Arizona 20.4 11.0 -9.4
Montana 20.7 11.3 -9.4
Louisiana 21.7 12.5 -9.2

Source: Gallup-Healthways Well-Being Index

Consistent with this, all 10 states with the highest fraction of adults who were uninsured as of 2016 elected not to expand their Medicaid programs.

As shown in Table 4, the state with the highest share is Texas, with 20.5 percent of the state’s adult residents still without health insurance. Mississippi, Oklahoma, Georgia, and Florida round out the top five. As of 2016, the fraction of residents in expanding states who were uninsured was 8.2 percent (about 1 in 12), while the corresponding share in non-expanding states was almost 80 percent higher at 14.5 percent (about 1 in 7).

Table 4: Ten States with the Highest Percent Uninsured in 2016

State Percent Adults Uninsured
Texas 20.5
Mississippi 17.2
Oklahoma 16.3
Georgia 15.6
Florida 14.6
Idaho 14.0
Alabama 13.6
North Carolina 13.6
South Carolina 13.1
Wyoming 12.9

Source: Gallup-Healthways Well-Being Index

Recent research indicates that the ACA-induced increases in health insurance coverage have increased economic well-being (Brevoort, Grodzicki, Hackmann, 2017; Gallagher, Gopalan, Grinstein-Weiss, 2017) and increased access to health care (Sommers et al, 2015). Greater coverage can also provide additional resources for health care providers that differentially treat those without health insurance while simultaneously easing the pressure on state and local budgets that are often responsible for the costs of uncompensated care to the uninsured.

Reductions in Competition

Premiums for health insurance purchased in the ACA exchanges increased much more rapidly in 2017 than in the two previous years. One key driver was the planned 2016 expiration of the ACA’s reinsurance payments to insurers. These payments are essentially “insurance for insurers” and were designed so that insurers would not have a strong incentive to avoid very high-cost patients (Hall, 2010).

Between 2014 and 2016, the federal government covered most of the costs for very high-cost patients who were enrolled in the exchanges through reinsurance.[3] But the ACA included these payments for just the first three years to assist with the startup of the exchanges. Beginning in 2017, insurers no longer received these payments. This mechanical increase in insurers’ costs caused them to increase premiums by significantly more than they otherwise would have.

But the reduction in competition in the exchanges described above also contributed to the increase in premiums from 2016 to 2017.

Competition is now especially limited in markets with a relatively small population. For example and as shown in Table 5, among counties with populations of less than 25,000, the average number of insurers in 2017 was just 1.8 versus 2.9 in counties with populations of 500,000 or more.[4]

Table 5: Average Number of Private Insurers by County Population in 2015 and 2017

County Population 2015 2017
> 500k 5.8 2.9
250k - 499k 4.8 2.6
100 – 249k 4.4 2.4
50 – 99k 3.6 2.0
25 – 49k 3.4 1.9
< 25k 2.9 1.8

Source: Kaiser Family Foundation County-level Data

This represents a significant reduction from just two years earlier with substantial reductions in competition in larger markets during that time as well. Perhaps most strikingly, nearly 40 percent of counties with fewer than 25,000 residents are currently served by just one private insurer.

Reforming the ACA: Six Ideas

Following are six considerations for policymakers focused on increasing competition and choice.

Expand choice in smaller markets

One possible policy lever that could enhance competition in smaller markets would be to bundle these markets with larger urban areas. In other words, in order to serve the larger urban area, an insurer would have to also serve the smaller market.

Some states have taken this approach while others have allowed each county to be its own market. As three co-authors and I pointed out in recent research, rural counties that were bundled with nearby urban counties had more private insurers competing and lower premiums and thus better market outcomes (Dickstein et al, 2015). This would be a straightforward way to reduce the gap in competition between large and small markets.

Increase reimbursement in smaller markets

The limited interest of private insurers in smaller markets is not unique to the ACA exchanges. Similar problems existed for Medicare Advantage (MA), the program through which nearly one-third of Medicare recipients now obtain private health insurance coverage, during the 1980s and 1990s. Large counties tended to have many more options than smaller counties, and as a result virtually all rural residents were enrolled in traditional fee-for-service Medicare rather than MA. But in 1998, the federal government introduced floor payments in smaller low-cost markets that significantly increased insurer interest and led to a large increase in insurer entry (Duggan, Starc, and Vabson, 2016). Similar enhancements could be made to the reimbursement of ACA plans in smaller markets with little or no competition.

Add Medicaid Managed Care plans as an option

Another way to expand choice in the exchanges would be to allow enrollees to acquire coverage in private Medicaid managed care (MMC) plans. Currently, more than 2 in 3 Medicaid recipients are in a private MMC plan (Duggan and Hayford, 2013).

Many of today’s Medicaid recipients will later enroll in the ACA exchanges as their income rises and will then return to Medicaid plans when their income declines. In counties with limited or no participation in the ACA exchanges, states and MMC plans could expand their coverage to include exchange enrollees. Given that all but three states (Alaska, Connecticut, and Wyoming) have a significant MMC presence, this could substantially expand choice and increase competition in many underserved markets.

Increase incentives for young adults to enroll

Americans between 25 and 34 were almost twice as likely as those between 45 and 64 to be uninsured prior to the ACA (28 percent versus 15 percent in 2009). Despite this, the ACA provides much lower subsidies to younger adults as shown in Figure 2. [5]

And for those young adults not receiving subsidies, age rating regulations likely increase their premiums above what they otherwise would be. More specifically, the ACA does not allow premiums for older recipients to be more than three times greater than for younger recipients despite their expected costs often being more than three times higher. These factors serve to reduce enrollment among young adults and increase premiums in the exchanges. Revisions to the age-rating regulations or to the subsidy formula that target young adults could lead to substantial increases in coverage at relatively low cost.

Figure 2. ACA Subsidies by Income and Age in the Average Market in 2015

Fund outreach to the uninsured

During the last four years, the federal government and individual state governments have spent a great deal of time and money on informing the public about the availability of health insurance coverage through Medicaid and the insurance exchanges.

That’s led some who would otherwise be uninsured to enroll in Medicaid or to purchase subsidized coverage in their state exchange. At the federal level, funding for this outreach was scaled back considerably in early 2017.

This affects the 34 states with exchanges that are controlled entirely by the federal government.  However, many of the other 17 states expanded outreach efforts last year. Previous research has shown that outreach can make a significant difference for enrollment in government programs (Aizer, 2003; Karaca-Mandic et al, 2017). Consistent with this, in the 34 states with federally run exchanges, enrollment fell by 4.4 percent from 2016 to 2017. In contrast, in the other 17 states where outreach efforts were generally expanded, enrollment rose by 2.1 percent (Curran et al, 2017).

This divergence in enrollment trends seems likely to continue in 2018, with federal funding for outreach efforts slated to fall by 90 percent (from $100 million to $10 million) in the current fiscal year. To the extent that states with federally run or state-administered exchanges want to hold constant or increase health insurance coverage among their residents in the months ahead, expanding state and local funding for outreach is likely to have a high payoff.

Reduce uncertainty about cost-sharing subsidies

Americans with incomes between 100 and 400 percent of FPL qualify for subsidies to purchase private coverage through the ACA exchanges. The benchmark “silver” plan in the exchanges has an actuarial value (AV) of 70 percent, meaning that the insurer covers 70 percent of expected health care costs.

Plans with an AV in this range typically have relatively high deductibles and co-payment amounts. Individuals with incomes between 100 and 250 percent of the poverty line qualify for additional federal subsidies to reduce these out-of-pocket costs.

The cost-sharing subsidies help those with lower incomes cover the cost of deductibles and co-payments, which can amount to thousands of dollars following a major medical problem. These subsidies partially explain why lower income individuals are substantially more likely to enroll in the exchanges than their counterparts making more money.

For example, among individuals without another source of coverage and with incomes between 150 and 200 percent of the federal poverty line, 55 percent take up coverage in the exchanges. The corresponding take-up rate for those with incomes between 300 and 400 percent of the federal poverty line is just 14 percent (Buettgens, Kenney, and Pan, 2015).

In recent months, there has been uncertainty as to whether these cost-sharing subsidies would continue. The total amount spent on subsidies will amount to about $7 billion in 2017, which is substantially lower than the $38 billion for premium subsidies.

The Congressional Budget Office estimated that premiums in the ACA exchanges would increase by 20 percent if these subsidies were terminated (CBO, 2017). This would lead to large reductions in health insurance coverage as some individuals would be priced out of coverage. By continuing these cost-sharing subsidies, policymakers could prevent substantial price increases and insurer exits. A recent effort to forge a bipartisan compromise has focused primarily on this issue (Hackman, 2017).


While the six potential reforms outlined above will not address all of the limitations of our current health care system, they would lead to further increases in insurance coverage and — through increases in competition — prevent large price increases in the ACA’s health insurance exchanges. These reforms could lead to additional reductions in the number of uninsured and help keep the growth in overall health care spending low by historical standards.

SIEPR research assistants Nicole Chen and Valerie Scimeca contributed to this policy brief.


Aizer, Anna. “Low Take-up in Medicaid: Does Outreach Matter and For Whom?” American Economic Review Papers and Proceedings, May 2003.

Brevoort, Kenneth, Daniel Grodzicki, and Martin Hackmann. “Medicaid and Financial Health.” Unpublished mimeo, 2017.

Buettgens, Matthew, Genevieve Kenney, and Clare Pan. “Variation in Marketplace Enrollment Rates in 2015 by State and Income.” Urban Institute and Robert Wood Johnson Foundation, October 2015.

Centers for Medicare and Medicaid Services (CMS). “National Health Expenditure Data.”

Congressional Budget Office. “The Effects of Terminating Payments for Cost-Sharing Reductions.” August 15, 2017.

Cox, Cynthia, Ashley Semanskee, Gary Claxton, and Larry Levitt. “Explaining Health Care Reform: Risk Adjustment, Reinsurance, and Risk Corridors.” Kaiser Family Foundation Issue Brief, 2016.

Curran, Emily, Sabrina Corlette, Kevin Lucia, and Justin Giovannelli. “2017 Federal and State Marketplace Trends Show Value of Outreach.” To the Point, The Commonwealth Fund, May 4, 2017.

Dafny, Leemore, Mark Duggan, and Subramaniam Ramanarayanan. "Paying a Premium on Your Premium? Consolidation in the Health Insurance Industry." American Economic Review, April 2012.

Dickstein, Mike, Mark Duggan, Joe Orsini, and Pietro Tebaldi. “The Effect of Market Structure on Health Insurance Premiums: Evidence from the ACA’s Insurance Exchanges.” American Economic Review: Papers and Proceedings, May 2015.

Duggan, Mark, and Tamara Hayford. “Has the Shift to Managed Care Reduced Medicaid Spending? Evidence from State and Local-Level Mandates.” Journal of Policy Analysis and Management, Summer 2013.

Duggan, Mark, Amanda Starc, and Boris Vabson. “Who Benefits when the Government Pays More? Evidence from the Medicare Advantage Program.” Journal of Public Economics, October 2016.

Duggan, Mark, Gopi Shah Goda, and Emilie Jackson. “The Effects of the Affordable Care Act on Health Insurance Coverage and Labor Market Outcomes.” SIEPR working paper, 2017.

Gallagher, Emily, Radhakrishnan Gopalan, and Michal Grinstein-Weiss. “The Effects of Health Insurance on Home Payment Delinquency: Evidence from the ACA Marketplace Subsidies.” Unpublished mimeo, 2017.

Gallup News. “Kentucky, Arkansas Post Largest Drop in Uninsured Rates.” February 8, 2017.

Hackman, Michelle. “Senators Discuss Bipartisan Approach to Repair Obamacare.” The Wall Street Journal, September 6, 2017.

Hadley, Jack, John Holahan, Teresa Coughlin, and Dawn Miller. “Covering the Uninsured in 2008: Current Costs, Sources of Payment, and Incremental Costs.” Health Affairs, August 2008.

Hall, Mark. “The Three Types of Reinsurance Created by Federal Health Reform.” Health Affairs, June 2010.

Karaca-Mandic, Pinar, Andrew Wilcox, Laura Baum, Colleen Barry, Erika Fowler, Jeff Niederdeppe, and Sarah Gollust. “The Volume of TV Advertisements During the ACA’s First Enrollment Period Was Associated with Increased Insurance Coverage.” Health Affairs, April 2017.

Sommers, Benjamin, Munira Gunja, Kenneth Finegold, and Thomas Musco. “Changes in Self-Reported Insurance Coverage, Access to Care, and Health Under the Affordable Care Act.” Journal of the American Medical Association, July 28, 2015.

U.S. Department of Health and Human Services (HHS). “Early Release of Selected Estimates Based on Data from the National Health Interview Survey, 2016.”

[1] Some individuals with incomes below 138 percent of FPL are ineligible for Medicaid coverage. For example, undocumented immigrants or recent immigrants are typically not eligible.

[2] Enrollment in these five states ranged from 1.5 to 2.4 percent of their non-elderly residents.

[3] In 2014, the federal government covered 80 percent of an individual’s costs above $45,000 for the year. The next year this declined to 50 percent. And in 2016, reinsurance covered 50 percent of the costs above $90,000 for the year. As a result, reinsurance funding fell from 2014 to 2016 and will be zero in 2017 and beyond (Cox et al, 2016).

[4] For these statistics, I considered only the 39 states that use the platform, thus excluding California, Kentucky, and 10 other states.

[5] This is because the ACA adjusts the subsidy so that individuals with, for example, income of 200 percent of FPL would pay just 4 percent of income on their health insurance plan. This would imply about $1,000 for a single adult. But the cost of the plan is much higher for older enrollees so the subsidies are correspondingly higher too.

Mark Duggan
Publication Date
September, 2017