Many opponents of Obamacare claim that large employers will drop employee health coverage in droves. The Wall Street Journal has made this argument a centerpiece of its opposition to the health exchanges. The argument has some face validity – employers that drop coverage can save about $10,000 per employee in insurance costs but only have to pay fines of $2000 per employee. What employer would not want to save $8000 per employee?
Supporters of Obamacare argue that if employers do not pay for insurance, they will have to increase wages. This will temper the incentives of employers to drop coverage. This follows from a classic model in labor economics that says that employers have to give workers a competitive wage/benefits bundle, and that the mix of wages and benefits is largely fungible. Thus, if benefits fall by $10,000, wages will increase by about the same amount. The theory is well accepted. While it has been difficult to construct empirical tests of this theory, the available evidence is largely supportive (though the evidence of 1:1 fungibility is less compelling than the evidence of some degree of fungibility.) This may explain why the Congressional Budget Office predicts that only a few million workers will lose their employer sponsored coverage and get pushed onto the exchange. Even so, the Wall Street Journal and others have dismissed this theory and evidence, arguing that employers who drop coverage will pocket the full savings and therefore than tens of millions of workers will be affected.
I want to propose a simple test of the naysayers’ position. The test relies on evidence that the Wall Street Journal and others should find unimpeachable –stock market valuations. This is a quick and dirty test but the results are so compelling that I think it is sufficient.
I rely on some basic labor market data. Consider that American firms pay workers about $60,000 annually in average compensation and benefits. Moreover, these labor expenses represent about two-thirds of the total cost of production. Now suppose that firms can reduce their wage/benefit package by $4000 by dropping $10,000 in insurance coverage, paying a $2000 fine, and increasing wages by just $4000. In other words, I am supposing that there is 0.5:1 fungibility between wages and benefits. (I get similar qualitative findings with any ratio that is substantially less than 1:1, as the naysayers are assuming.) This savings reduces total compensation costs by 6.7 percent and therefore reduce total production costs by 4.4 percent. In effect, this simple move would increase returns on sales by 4.4 percent! This would effectively double the average return on sales for U.S. manufacturing firms (which is currently about 5 percent) and represent an even bigger increase in retail and other industries.
If the naysayers are correct, then the passage of Obamacare should have had a remarkably powerful impact on the stock market. Share values should have doubled and then some! And the effect should have been immediate, as investors could have performed the same math as I have just done and seen the good times ahead. Of course this did not happen. I can only conclude that investors did not and do not expect large employers to reap a windfall from Obamacare. Large employers that drop coverage will have to increase compensation to workers to make up for the difference. Obamacare does not provide a free lunch to employers, and this is why the CBO and others expect such a small impact on employer-provided coverage.
What I find most remarkable is that the Wall Street Journal is a staunch believer in free markets, yet it conveniently ignores market mechanisms when it wants to make a political point.