Report | Inequality and Poverty

Fundamental health reform like ‘Medicare for All’ would help the labor market: Job loss claims are misleading, and substantial boosts to job quality are often overlooked

Download PDF

Press release

Fundamental health reform like “Medicare for All” would be a hugely ambitious policy undertaking with profound effects on the economy and the economic security of households in America. But despite oft-repeated claims of large-scale job losses, a national program that would guarantee health insurance for every American would not profoundly affect the total number of jobs in the U.S. economy. In fact, such reform could boost wages and jobs and lead to more efficient labor markets that better match jobs and workers. Specifically, it could:

  • Boost wages and salaries by allowing employers to redirect money they are spending on health care costs to their workers’ wages.
  • Increase job quality by ensuring that every job now comes bundled with a guarantee of health care—with the boost to job quality even greater among women workers, who are less likely to have employer-sponsored health care.
  • Lessen the stress and economic shock of losing a job or moving between jobs by eliminating the loss of health care that now accompanies job losses and transitions.
  • Support self-employment and small business development—which is currently super low in the U.S. relative to other rich countries—by eliminating the daunting loss of/cost of health care from startup costs.
  • Inject new dynamism and adaptability into the overall economy by reducing “job lock”—with workers going where their skills and preferences best fit the job, not just to workplaces (usually large ones) that have affordable health plans.
  • Produce a net increase in jobs as public spending boosts aggregate demand, with job losses in health insurance and billing administration being outweighed by job gains in provision of health care, including the expansion of long-term care.

While the overall effect of fundamental health reform on the labor market would be unambiguously positive, this does not mean policymakers should ignore the distress caused by job transitions forced by this reform. Specifically, policy support should be provided to help displaced health insurance and billing administration workers move into new positions. But we should not let critics of Medicare for All inflate the scale of this transition challenge or falsely present the number of jobs displaced in individual sectors as the net effect of reform on labor markets. The number of health insurance and billing administration workers who would need to transition implies an increase in the rate of overall job market churn that is relatively small: Job losses for these workers would be equivalent to one-twelfth the size of economywide layoffs in 2018.

Background: The need for fundamental health reform

Currently, despite the significant gains in health care coverage spurred by the passage of the Affordable Care Act (ACA) in 2010, roughly 23 million Americans between the ages of 19 and 64 are uninsured, and another 64 million are underinsured (Collins, Bhupal, and Doty 2019).1 In addition to problems with access, the American health care system also suffers from excess costs.2 While excess health care cost growth has slowed notably in the last decade, it would be prudent for policymakers to try to keep this cost growth in check with significant policy reforms rather than simply hoping for the best going forward. Some highly important health-related prices have begun rising rapidly in the very recent past. Insurance premiums, for example, rose 20% in 2019.3 Overall spending on prescription drugs rose more than 9% between the fourth quarter of 2018 and the fourth quarter of 2019—the largest year-over-year change since 2015.4

Bivens 2018b provides data demonstrating that health spending in the U.S. is higher than in advanced peer countries and has risen faster over time—and yet continues to buy worse health outcomes. The higher and faster-growing spending of the United States is driven by faster growth of prices, not by growth in the volume of health care goods and services consumed. Further, international evidence shows that a key component of controlling cost growth is a strong public role in setting and negotiating the prices of health care goods and services.

A fundamental reform like Medicare for All (M4A) would make coverage universal. Further, by providing a counterweight to (or outright eliminating) the substantial market power that keeps prices high and that is currently wielded by many key players in the health care sector (e.g., insurance companies, drug companies, specialty physicians, and device makers), such a reform could also have great success in containing health care cost growth. This could in turn provide relief from many of the ways that rising health costs squeeze family incomes.

An underappreciated benefit of such a reform is that it would also lead to a much better functioning labor market in many areas. Job quality would increase, job switching would become less stressful, better “matches” between workers and employers would boost productivity, and small businesses would be much easier to launch.

Despite the fact that M4A could deliver these large benefits to efficient labor market functioning, the policy often comes under fire from critics making highly exaggerated claims about the potential job loss that could occur under such a reform. The grain of truth in some of the claims is that, like any productivity improvement, the adoption of a reform like M4A would require the redeployment of workers from one sector (the health insurance and medical billing complex) to other sectors (mostly the delivery of health care). But there is little in the M4A-induced redeployment of workers that would greatly stress the American labor market over and above the uncertainty and churn that characterizes this labor market every year. Smart policy could make this redeployment eminently manageable for those workers who would be required to make the transition.

This brief highlights some labor market implications of M4A and critically examines claims that large job losses in the health insurance and billing administration sectors would make M4A an undesirable policy.

Health reform as labor market policy: Key effects for workers

Fundamental reforms like M4A could greatly aid labor market outcomes for U.S. workers. The most obvious benefits would be higher wages and salaries, increased availability of good jobs, reduced stress during spells of job loss, better “matches” between workers and employers, and greater opportunity to start small businesses.

Higher cash wages and salaries

Medicare for All could increase wages and salaries for U.S. workers by reducing employers’ costs for health insurance—freeing up fiscal space to invest in wages instead. The share of total annual compensation paid to American employees in the form of health insurance premiums rather than wages and salaries rose from 1.1% in 1960 to 4.2% in 1979 to 8.4% in 2018.5 If this post-1960 increase had been only half as large—and employers had spent the health cost savings on wages and salaries—the take-home wages of American workers would have been almost $400 billion higher in 2018.6 Given that the share of total compensation spoken for by health insurance premiums is starting from a high base today, any reform that managed to slow the excess growth of health spending going forward would go a long way in making space for faster growth of cash compensation.7

Increased availability of ‘good jobs’

Medicare for All could increase job quality substantially by making all jobs “good” jobs in terms of health insurance coverage and by increasing the potential for higher wages. While the definition of a “good job” is always going to be a bit imprecise, the vast majority of U.S. workers would say that a good job is one that pays decent wages and that also provides the health insurance coverage and retirement income benefits that most of today’s workers can only reliably access through employment. Nearly half of jobs fail this test on account of health care coverage alone: In 2016, 46.9% of workers held jobs in which their employer made no contributions to the workers’ health care; for workers in the middle fifth of the wage distribution, 42.9% held jobs in which the employer made no contribution to their health care (EPI 2017).

By making health coverage universal and delinking from employment, M4A would make it far easier for employers to offer good jobs in this regard, as every job would now be accompanied by guaranteed health care coverage. Further, as noted above, wages and salaries would have substantial room to grow if health care costs were taken off of the backs of employers. Schmitt and Jones (2013) estimate the share of good jobs—jobs that clear a specified wage floor8 and provide health and retirement coverage—in overall employment each year between 1979 and 2011. They then look at various policy changes that would boost this share. They find that providing universal health coverage would boost the probability that any given job in the economy is a good job by almost 20%—and that’s even before any potential boost to the share of jobs that are good jobs coming from cash wage increases provided as employers shed health care costs.9 The boost to job quality from making health coverage universal would be even greater for women workers, as women are currently less likely to receive employer-sponsored health insurance benefits from their own employers.10

Less damaging spells of joblessness

Medicare for All could make job losses and transitions less stressful by delinking employment and access to health insurance, emulating the universal access to health care offered by our rich country peers. The U.S. is unique among the rich countries of the world in how much it ties crucial social benefits—like health insurance and retirement income—to specific jobs. Hacker (2002) has referred to this arrangement as the “divided welfare state,” with some Americans having relatively full access to health and retirement security while others have access to virtually none, all based on the specific jobs they have. This makes some jobs in the U.S. economy especially valuable, and hence especially damaging to lose. Manufacturing workers without a college degree, for example, likely incur enormous income and social benefits losses in the event of job loss stemming from either automation or trade. The ability of universal, public social benefits to make individual job losses less damaging has been long recognized by social scientists (see, for example, Estevez-Abe, Iversen, and Soskice 2001).

Smooth job transitions contribute to economic dynamism by helping ensure that vacancies are filled quickly by appropriate workers and that unemployed workers can quickly find new jobs that make good use of their skills. Smooth job transitions will also be an important components of meeting crucial policy goals such as mitigating greenhouse gas emissions with wholesale changes in how energy is created. Policies that make job transitions easier and inspire less resistance from workers should be encouraged. Fundamental health reform that, like M4A, guarantees access to insurance regardless of one’s current job status is a key part of making such transitions easier.

Better labor market matches between workers and employers

Medicare for All could decrease inefficient “job lock” and boost small business creation and voluntary self-employment. Making health insurance universal and delinked from employment widens the range of economic options for workers and leads to better matches between workers’ skills and interests and their jobs. The boost to small business creation and self-employment would be particularly useful, as the United States is a laggard in both relative to advanced economy peers.

Substantial evidence indicates that our current system of employer-sponsored insurance (ESI) creates significant “job lock”—a condition in which workers who don’t want to lose their current ESI stay in their current jobs rather than make transitions that would better meet their needs. In a comprehensive review of this literature, Baker (2015) finds:

The likely range of a job-lock effect is a reduction in turnover—the rate at which people leave jobs—of 15–25 percent among workers with EPHI [employer-provided health insurance, or ESI]. With normal turnover for prime-age workers (people ages 25–54) in the range of 15–20 percent per year, this job-lock effect implies a reduction in annual turnover of around 4 percentage points among prime-age workers with [employer-provided health insurance, or ESI].

Making employment decisions based on access to ESI rather than on other criteria—such as work–life balance, cash wages, and commuting distance—can lead to employment “matches” that are less productive and that decrease overall worker welfare relative to job choices that are not constrained by the availability of health insurance.

More small-business formation

Despite policymakers’ frequent claims that they seek to support small businesses in the U.S. economy, the United States has a notably small share of small-business employment relative to our rich country peers. In 2018, for example, the U.S. was dead-last among the members of the Organisation for Economic Co-operation and Development (OECD) in its share of self-employment, at just 6.3% of employment. Countries that are frequently portrayed in U.S. business reporting as being choked by regulation—like Spain, France, and Germany—have far higher shares of self-employment, at 16.0%, 11.7%, and 9.9%, respectively (OECD 2020).

Besides a low share of self-employment, the U.S. also had significantly lower shares of overall employment in small businesses, across nearly all industrial sectors. The latest OECD data show that the U.S. share of employment in enterprises with fewer than 50 employees is lower than in any other country except for Russia (OECD 2018, Figure 7). In an earlier overview of trends in employment by firm size, Schmitt and Lane (2009) highlight how health care policy plays two key roles in potentially explaining cross-country trends. First, because health care is nearly universally provided in other rich countries, workers choosing to start their own businesses in those countries do not face a cost confronting would-be entrepreneurs in the U.S.: the loss of ESI. Second, small businesses in the U.S. are at a distinct disadvantage in recruiting employees because the cost of providing health care coverage is significantly higher for small companies.11

Employment effects of fundamental health reform: gains in health care, losses in insurance and billing—with likely economywide net job gains from rising economic demand

Like all positive productivity gains, Medicare for All would be more likely to increase the total number of jobs in the U.S. economy, even as health reform leads to the redeployment of workers from some sectors and into others.

Despite the many labor market benefits of fundamental health reform like M4A, many critics have claimed that such reform would lead to a loss of jobs. This claim is misleading. One small grain of truth to it is that the universal provision of health insurance would allow people who would strongly prefer not to work (or not to work full time), but who have remained in their current jobs in order to retain health insurance, to be free to quit. This type of voluntary reduction in labor supply following a health reform would be strongly welfare-improving. For example, the ACA was clearly associated with a large increase in parents with young children transitioning to part-time work (see Jørgensen and Baker 2014). To the degree this occurred because these parents no longer needed to work full time to obtain ESI, and they preferred spending more time with their children for reasons of work–life balance, it should be seen as a clear win for the policy.

Generally, people expressing concern about job loss stemming from a policy are concerned about involuntary job loss that leads to a higher level of unemployment in the economy. Unemployment is almost entirely a function of the level of aggregate demand: spending by households, businesses, and governments.12 The effect of fundamental health reform on the level of aggregate demand depends in turn on the balance of increased public spending and the means of financing this spending. All else equal, more public spending will boost aggregate demand and create jobs, while higher taxes will reduce aggregate demand and restrain job growth. Further, the progressivity of taxes used to finance fundamental health reform will also condition its effect on aggregate demand. The more progressive the taxes that finance health reform, the less they will drag on job growth. Increased public spending combined with progressive tax increases would almost certainly boost the level of aggregate demand and lead to lower unemployment, all else equal.

While the overall number of jobs and the level of unemployment in the economy is largely a macroeconomic issue determined by aggregate demand, claims that fundamental health reform like M4A will lead to job loss sometimes sound plausible because it is easy to envision the specific jobs that might be displaced: jobs in the health insurance and billing administration sectors. But these job displacements would be balanced by likely job gains in other sectors—most particularly in health care delivery. The health insurance coverage expansions of M4A will boost demand for health care goods and services, and workers will need to be hired to meet this demand.

Job losses in the health insurance and billing administration sectors

A recent analysis of the economic effects of M4A (Pollin et al. 2018) includes the projection that up to 1.8 million jobs in the health insurance and billing administration sector (the divisions of hospitals and doctors’ offices dedicated to administrative processing of bills and payments) could be made redundant. These potential 1.8 million lost jobs are frequently presented as if they constitute the net employment effect of M4A.13 This is a deeply flawed misrepresentation of Pollin and his colleagues’ work. In fact, their estimates are a gross (not net) measure of job displacement or “churn”—the regular process of workers starting and leaving jobs during the course of their work lives. Relative to the scale of other gross measures of job churn, the churn associated with M4A is not large.

It is true that one source of cost savings from the introduction of M4A is the reduced demand for insurance and billing administration. In turn, this reduced demand would shift employment out of these sectors. This could certainly cause challenges and economic distress for the workers within these sectors who are directly affected. But for some perspective, it is worth noting that 21.5 million workers were laid off in 2018 (BLS 2020b). If the 1.8 million workers that Pollin et al. (2018) identify as potentially being displaced by M4A were forced to transition over the four-year phase-in commonly identified with M4A plans, this would increase the national rate of layoffs by about 2%. It is also worth noting that even within just the finance and insurance sectors, there have been 1.7 million layoffs in the past four years (BLS 2020b). And yet it’s safe to say that very few people even in the business press have made any note of this. This is not a shock: Our economy generates a huge amount of job churn every year. This churn is the hallmark of growth in productivity—getting more economic output with fewer inputs. While productivity growth can indeed put downward pressure on jobs in the sector experiencing it directly, Autor and Salomons (2018) demonstrate that productivity gains within a given sector strongly boost job growth in other sectors, as the savings to households and businesses stemming from enhanced productivity increase purchasing power that supports demand for these other sectors’ outputs.

If workers in the insurance or billing administration sectors were particularly hard-pressed for reemployment prospects because of geographic isolation or low average levels of educational credentials, their displacement might pose particular concern to policymakers. But employment in the health insurance and billing administration sectors is not particularly geographically concentrated,14 and Pollin et al. (2018) show that 56.5% of workers in these sectors have a four-year college degree or more education, a far greater share than the overall labor force (in 2018, 37.6% of workers had a four-year degree or more education, according to EPI 2020b).

Substantial likely job gains in the health care sector

While it may seem counterintuitive, fundamental health reform like M4A is almost guaranteed to substantially expand employment in the health care sector overall, even taking reduced billing administration employment into account. Often people hear that fundamental reform is aimed at cost containment and then imagine that part of this cost containment will take the form of fewer jobs providing health care, but this is not necessarily the case. As noted before, the U.S. is an outlier in terms of how much it spends on health care, but its health care workforce as a share of the total workforce is not out of line with shares in other countries. For example, in 2017 the health care workforce in the U.S. was equal to 13.4% of the overall workforce, while the share averaged 12.9% in the 20 other richest OECD countries.15 Additionally, seven of these other countries had health care workforce shares equal to or higher than the U.S.’s 13.4%.16

Pollin et al. (2018) estimate that expanded access to health care could increase demand for health services by up to $300 billion annually. Given the current level of health spending and employment, this would translate into increased demand for 2.3 million full-time-equivalent workers in providing healthcare.17 Obviously all of the workers displaced from the health insurance and billing administration sectors could not necessarily transition into these jobs seamlessly, but well over 10% of workers in the health insurance sector, for example, are actually in health care occupations (e.g., they are doctors or nurses).18

Further, several M4A plans have provisions to pay for long-term care services. Reinhard et al. (2019) have estimated that in 2018, Americans provided roughly 34 billion hours in unpaid long-term care. If this care was divided up among full-time paid workers, it would require 17 million new positions. Of course, not all of this currently unpaid care would be converted into paid positions in the job market. But if even 10% of unpaid care translated into new jobs, it would create enough new demand for workers to essentially offset the displacement of workers in the health insurance and billing administration sectors.

The upshot: M4A creates a small amount of manageable churn but increases the overall demand for labor and boosts job quality

The job challenge relating to a fundamental health reform is managing a relatively small increase in job churn during an initial phase-in period. Most Medicare for All plans explicitly recognize and account for the costs of providing these workers the elements of a just transition. As noted previously, this sort of just transition is far easier when health care is universally provided.

Besides this challenge, the effect of fundamental reform like M4A on the labor market would be nearly uniformly positive. The effect of a fundamental reform like M4A on aggregate demand is almost certainly positive and will therefore boost the demand for labor. The number of jobs spurred by increased demand for new health care spending (including long-term care) will certainly be larger than the number displaced by realizing efficiencies in the health insurance and billing administration sectors.

Finally, the introduction of fundamental health reform like M4A—particularly reform that substantially delinks health care provision from specific jobs—would greatly aid how the labor market functions for typical working Americans. Take-home cash pay would increase, job quality would improve, labor market transitions could be eased for employers and made less damaging to workers, and a greater range of job opportunities could be considered by workers. The increased flexibility to leave jobs should lead to more productive “matches” between workers and employers, and small businesses and self-employment could increase.

Fundamental health reform would benefit typical American families in all sorts of ways. Importantly, contrary to claims that such reform might be bad for jobs, this reform could substantially improve how labor markets function for these families.


About the author

Josh Bivens joined the Economic Policy Institute in 2002 and is currently EPI’s director of research. His primary areas of research include mac­roeconomics, social insurance, and globalization. He has authored or co-authored three books (including The State of Working America, 12th Edition) while working at EPI, has edited another, and has written numerous research papers, including many for academic journals. He appears often in media outlets to offer eco­nomic commentary and has testified several times before the U.S. Congress. He earned his Ph.D. from The New School for Social Research.


1. Underinsurance includes coverage gaps throughout a year.

2. “Excess costs” typically refers to health care costs that are rising faster than other economic benchmarks, such as overall gross domestic product. The Congressional Budget Office, for example, defines excess costs as the percentage change in health care costs per beneficiary minus the percentage change in per capita gross domestic product (Banthin 2017).

3. Author’s analysis of data from BLS 2020a.

4. Author’s analysis of data from BEA 2020, Table 2.4.5U, line 120.

5. Data from BEA NIPA Table 7.8, line 17, divided by data from BEA NIPA Table 2.1, line 2 (BEA 2020).

6. If the share of total compensation directed toward premiums had grown by just 3.65 percentage points rather than the actual increase of 7.3 percentage points between 1960 and 2018, and the difference had all been directed toward increased wages and salaries, then wages and salaries in 2018 would have been higher by an amount equal to 3.65% of total compensation of employees in that year ($10.93 trillion), or $399 billion (BEA 2020).

7. A quick example may help make the point. If health insurance premiums rose by 7% per year, they would double in 10 years. If growth of other forms of compensation remained flat, this would lead to the share of health insurance premiums in total compensation doubling in 10 years. If premiums started from 1% of total compensation in the base year (as in 1960), this doubling would only “crowd out” 1% of total compensation that could be taken in the form of cash wages and salaries. If instead premiums started from a base of 8.4% of total compensation (as in 2018), this doubling would “crowd out” 8.4% of total compensation that could be taken in the form of cash wages and salaries.

8. The wage floor Schmitt and Jones specify for a good job is the median wage of men in 1979, adjusted for inflation. When adjusted for inflation into 2019 dollars using the CPI-U-RS, the 1979 men’s median hourly wage is $21.07 (EPI 2020a).

9. Specifically, in Table 1 Schmitt and Jones report that the share of good jobs would rise from 24.1% to 28.8% with the introduction of universal health coverage. Dividing this 4.7-percentage-point increase by 24.1% yields a 20% increase.

10. See Schmitt and Jones 2013, Table 2.

11. For evidence on the cost of providing health insurance by employer size, see Hertel-Fernandez, Gould, and Bivens 2009, Figure C.

12. For an explanation of how aggregate demand determines the level of unemployment, see Bivens 2018a.

13. See Pradhan 2019 for an example of these numbers being presented as the overall effect of M4A on jobs.

14. For example, if one sums the share of employment in health insurance, hospitals, and medical offices for each U.S. state, this share is essentially perfectly predicted by the state’s population share (correlation coefficient of 1.0). This is notably not true, for example, if one does the same exercise for manufacturing employment and state population shares (correlation coefficient of 0.8). Author’s analysis of data from BLS 2020c.

15. Data are from the OECD Health Statistics program (OECD 2019). The 20 OECD countries compared with the U.S. are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Iceland, Ireland, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Spain, Sweden, Switzerland, and the United Kingdom. Data for Canada and Japan are from earlier years (2016 and 2015, respectively) because data for 2017 are unavailable. Without Italy and Spain (both of which have very low health care workforce shares, below 8%), the U.S. would be very slightly below the OECD average for its share of the health care workforce.

16. The seven countries are Denmark, Finland, France, Netherlands, Norway, Sweden, and Switzerland.

17. In 2018, spending on health care services was $2.35 trillion (data from BEA 2020, NIPA Table 2.3.5), while full-time-equivalent employment was 18.25 million (BEA 2020, NIPA Table 6.5D). This translates into $129,000 in health spending per full-time-equivalent job. Dividing $300 billion by $129,000 yields the 2.3 million new full-time-equivalent workers needed to satisfy this new demand.

18. Author’s analysis of BLS 2019.


Autor, David, and Anna Salomons. 2018. “Is Automation Labor-Displacing? Productivity Growth, Employment, and the Labor Share.” Brookings Papers on Economic Activity Conference Drafts, March 2018.

Baker, Dean. 2015. Job Lock and Employer-Provided Health Insurance: Evidence from the Literature. AARP Public Policy Institute, March 2015.

Banthin, Jessica. 2017. “Health Care Spending Today and in the Future: Impacts on Federal Deficits and Debt.” Congressional Budget Office briefing, presented to a conference organized by the Center for Sustainable Health Spending, Washington, D.C.

Bivens, Josh. 2018a. Recommendations for Creating Jobs and Economic Security in the U.S.: Making Sense of Debates About Full Employment, Public Investment, and Public Job Creation. Economic Policy Institute, March 2018.

Bivens, Josh. 2018b. The Unfinished Business of Health Reform: Reining in Market Power to Restrain Costs Without Sacrificing Quality or Access. Economic Policy Institute, October 2018.

Bureau of Economic Analysis (BEA). 2020. National Income and Product Accounts, NIPA Tables (online database), Tables 2.1, 2.3.5, 2.4.5U, 6.5D, and 7.8. Accessed February 2020.

Bureau of Labor Statistics (BLS). 2019. “National Industry-Specific Occupational Employment and Wage Estimates.” NAICS 524114 – Direct Health and Medical Insurance Carriers. Last modified April 2, 2019.

Bureau of Labor Statistics (BLS). 2020a. Consumer Price Index program, “Databases, Tables & Calculators by Subject” (interactive data retrieval portal), Series ID CUUR0000SEME. Accessed February 2020.

Bureau of Labor Statistics (BLS). 2020b. Job Openings and Labor Turnover Survey (JOLTS), “Databases, Tables & Calculators by Subject” (interactive data retrieval portal), Series ID JTU00000000LDL and JTU52000000LDL. Accessed February 2020.

Bureau of Labor Statistics (BLS). 2020c. Quarterly Census of Employment and Wages Data (interactive data retrieval portal).

Collins, Sara R., Herman K. Bhupal, and Michelle M. Doty. 2019. Health Insurance Coverage Eight Years After the ACA: Fewer Uninsured Americans and Shorter Coverage Gaps, but More Underinsured. The Commonwealth Fund, February 2019.

Economic Policy Institute (EPI). 2017. State of Working America Data Library, “Health Insurance Coverage.” Last updated February 13, 2017.

Economic Policy Institute (EPI). 2020a. State of Working America Data Library, “Median/Average Hourly Wages.” Last updated February 20, 2020.

Economic Policy Institute (EPI). 2020b. State of Working America Data Library, “Wages by Education.” Last updated February 20, 2020.

Estevez-Abe, Margarita, Torben Iversen, and David Soskice. 2001. “Social Protection and the Formation of Skills: A New Interpretation of the Welfare State.” In Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, edited by Peter A. Hall and David Soskice. New York: Oxford Univ. Press. For a publicly available earlier version of the paper, see the 1999 paper prepared for presentation at the 95th American Political Association Meeting in Atlanta, Georgia:

Hacker, Jacob. 2002. The Divided Welfare State: The Battle over Public and Private Social Benefits in the United States. Cambridge Univ. Press.

Hertel-Fernandez, Alexander, Elise Gould, and Josh Bivens. 2009. Health Care Reform—Big Benefits for Small Businesses. Economic Policy Institute, July 2009.

Jørgensen, Helene, and Dean Baker. 2014. The Affordable Care Act: A Family-Friendly Policy. Center for Economic and Policy Research, September 2014.

Organisation for Economic Co-operation and Development (OECD). 2018. Entrepreneurship at a Glance: 2018 Highlights. October 2018.

Organisation for Economic Co-operation and Development (OECD). 2019. OECD Health Statistics 2019 (online database). Last updated November 15, 2019 (accessed February 2020).

Organisation for Economic Co-operation and Development (OECD). 2020. “Self-Employment Rate” (interactive online data table). Accessed February 2020.

Pollin, Robert, James Heintz, Peter Arno, Jeannette Wicks-Lim, and Michael Ash. 2018. Economic Analysis of Medicare for All. Political Economy Research Institute, November 2018.

Pradhan, Rachana. 2019. “Medicare for All’s Jobs Problem.” Politico, November 25, 2019.

Reinhard, Susan, Lynn Friss Feinberg, Ari Houser, Rita Choula, and Molly Evans. 2019. Valuing the Invaluable 2019 Update: Charting a Path Forward. AARP Public Policy Institute, November 2019.

Schmitt, John, and Janelle Jones. 2013. Making Jobs Good. Center for Economic and Policy Research, April 2013.

Schmitt, John, and Nathan Lane. 2009. An International Comparison of Small Business Employment. Center for Economic and Policy Research, August 2009.

See related work on Jobs | Employer coverage | Job creation | Economic mobility | Medicare | Wages

See more work by Josh Bivens