Conservatives have long argued that Obamacare would depress the labor market. In February the Congressional Budget Office (CBO) concurred. The CBO estimated that by 2024 Obamacare would reduce employment by 2.5 million workers—triple its previous estimate. The agency found the Affordable Care Act (ACA) would reduce both the supply of and demand for labor.

Until the CBO report only conservatives paid much recent attention to the supply side of the labor market. Liberals almost universally diagnosed the labor market as suffering from insufficient demand. Labor demand certainly did drop in the recession. But several economists pointed out low demand could not fully explain the economy’s prolonged weakness.

For one thing, demand substantially increased during the recovery—the stock market, corporate profits, and GDP fully recovered—yet job growth remained sluggish. Relatedly, outside the manufacturing and construction sectors businesses increased their use of nonlabor inputs (i.e. energy, capital, buildings) even as they reduced employment and hours. Why would businesses facing demand shortfalls use more energy and capital?

Additionally, the relationship between job vacancies and unemployment changed sharply during the recovery. Unsurprisingly, during the recession the number of job openings fell and the unemployment rate jumped. But in the recovery something surprising happened. The number of job vacancies improved substantially while unemployment did not. Job openings measure pure labor demand—do businesses want to hire—while unemployment reflects both supply and demand. The new disconnect between unemployment and job vacancies suggests some Americans had become less willing to work.

Casey Mulligan of the University of Chicago looked at this data in depth. He concluded the steep fall in labor-force participation after 2007 represented more than ageing baby boomers and discouraged workers giving up in a weak economy. He posited that labor supply had also fallen; some Americans had become less willing to work at available wages.

What could have reduced labor supply? Mulligan found that the expansion of welfare benefits through the stimulus substantially reduced the reward to working for many Americans. For example, Congress increased the duration of unemployment insurance (UI) benefits to ninety-nine weeks while increasing benefit levels and covering part of the healthcare premiums.

This substantially increased effective tax rates on working for many low- and moderate-income Americans. If they took a job they lose their UI benefits and premium subsidies in addition owing more in taxes. In 2009 an unemployed worker offered a job paying $600 a week and health benefits faced an effective tax rate of 84 percent. For every dollar he made his income would rise just 16 cents. The rest he would forfeit through taxes or reduced benefits.

Few Americans at any income level would work much when faced with such high tax rates. Mulligan found that, under the stimulus, eleven million Americans laid off during the Great Recession faced effective marginal tax rates exceeding 80 percent. Almost two million Americans faced tax rates over 100 percent: they would lose money if they accepted a job paying what they made before.

Tellingly, employment dropped the most among groups facing the largest effective tax increases. Mulligan concluded that while expanding welfare programs had humanitarian benefits, it had also reduced labor supply and dampened the labor market.

Liberals nearly universally ridiculed this analysis. Paul Krugman and others compared it to arguing soup kitchens had caused the Great Depression. They contended few workers would voluntarily remain unemployed—even at such higher marginal tax rates—and this amounted to blaming the victims of a bad economy. More substantively they pointed out that reduced labor supply should raise wages, but pay had only risen slowly during the recovery.

However, economists have long known that reductions in labor supply primarily manifest as reduced employment, with much smaller effects on wages. Reduced labor demand primarily reduces wages, with smaller effects on employment. The fact that, despite a substantial drop in labor demand, wages rose modestly after the recession suggests labor supply must have significantly contracted.

In 2010 Congress passed Obamacare. Mulligan also examined how the Affordable Care Act (ACA) affects the reward to working. He found that the ACA raised the effective marginal tax rate on about half the population by about ten percentage points. Post-Obamacare, a typical worker making median weekly wages faces an effective tax rate of 47 percent. Lower-income workers would face much higher rates. Even Alan Krueger, the former Chairman of President Obama’s Council of Economic Advisors, concluded that:

The individual mandate triggers the added problem of confronting low income families with relatively high losses in incremental disposable income as earned income rises…Added to federal income taxes, Social Security taxes, and the phaseout of the earned income credit, the individual mandate thus would present millions of low-income American families with total marginal tax rates in excess of 75 percent. Such high marginal tax rates may well make unemployment and welfare an attractive alternative to working. It would be the antithesis to supply-side economics.

The CBO report concurs with the negative effect of higher effective taxes on work decisions. The ACA increases the marginal tax rates on low-skill, low-wage workers. It also increases the payroll tax on highly skilled employees making over $200,000. The CBO reduced its estimates of the time these high wage workers spend at their job.

However the new benefits in the ACA—not the payroll tax increase—are the primary culprit in the reduction of work hours. The ACA expands Medicaid and offers generous premium subsidies to lower-income individuals and families. The value of these subsidies decreases with income which effectively raises marginal tax rates. A family of four’s premium’s double as their family income goes from one and a half to twice the poverty level.

As Mulligan and others have pointed out, these new benefits depress work incentives; people respond by working less or dropping out of the labor force altogether. Accounting for this effect explains most of why the CBO’s estimate tripled from its original projections. The CBO cited Mulligan’s work in arriving at its new estimates.

The White House has tried to spin reduced work as a good thing: people now no longer have to work if they do not want to and can take care of their families instead. However, the largest job losses come from not working parents but childless adults. The ACA further undercuts the principle that able bodied adults should work and take care of themselves. The decline in work hours largely comes from single adults currently working deciding they would rather stay home and enjoy more leisure.

The White House touting the benefits of less work sends a troubling signal. It clearly breaks from the bipartisan consensus reached over the last few decades that eroding work effort had become a problem with the social safety net. Two generations of Americans could make the rational decision to never work and rely on government benefits. These individuals could never climb the economic ladder or do better than their parents. Many would teach their children to imitate their mistakes. This defines the poverty trap: without work, people do not acquire the experience and skills necessary to rise out of poverty. The work requirements in the 1996 welfare reform intended to change these perverse incentives. Unfortunate by discouraging work the ACA will probably reduce economic mobility.

The CBO estimates the ACA will have a very large overall effect on work effort. The agency finds the law will reduce work hours 1.5 to 2 percent across the economy. This reduction is not freeing workers from “job lock” where people feel trapped at a particular employer to keep their health insurance. The ACA frees people from “work lock” where people need to work to get money and benefits, and instead have their benefits and subsidies often paid by other taxpayers. CBO also estimates that overall compensation will decline by 1 percent, primarily from low-wage workers working less. This will unfortunately increase the instability of the Medicare and Social Security that rely on payroll taxes for solvency. Reducing work hours reduces economic growth and wealth across the country.

These labor supply effects come on top of Obamacare’s deleterious effects on labor demand. The ACA has made hiring more costly. The Obamacare regulations have driven up the cost of employer sponsored health insurance, particularly for small businesses. Over a third of small business health plans will see premium increases of 30 percent or more. Companies must also now pay stiff penalties for full-time employees without health coverage.

Both the CBO model assumes and empirical studies find that workers will ultimately bear the cost of higher insurance premiums or mandate penalties through reduced compensation. However, many companies prefer to avoid pay cuts that could damage morale. The CBO expects most firms to slowly reduce pay by constraining raises. So in the short-term Obamacare makes hiring more expensive, because firms do not immediately pass through all cost increases to their employees.

Of course, Obamacare could reduce health costs. If it improves workers’ health and increases their productivity health costs could fall. In the long-term those benefits would also accrue to workers. But in the short term all signs point to higher costs.