Beginning in 2014, essentially everyone without insurance coverage from an employer (or the employer of a family member) will be eligible for significant subsidies for health insurance expenditures, at least if their family income is below 400% of the poverty line (a growing majority of America fits in that category). These ACA subsidies are so large that most of the time they outweigh the penalties employers will pay for not providing coverage.
As compared to 2013, there will be less demand for employer health insurance coverage in 2014 because other means of health insurance coverage are getting subsidized. Naturally, this should reduce the share of employers who offer coverage and reduce the number of people receiving coverage through an employer.
A number of economists are predicted this. However, I am afraid that they are under-estimating the margins on which the labor market can respond (sound familiar?) and thereby under-estimating the negative impact of the ACA on the fraction of people receiving health insurance coverage from an employer.
As far as I can tell, the prevailing approach to quantifying the effect is to look at econometric studies measuring the effect of health insurance costs (usually costs associated with income tax exclusion rules) on the likelihood of an employer offering coverage. This CBO report discusses its own estimates and the estimates of others; the debate seems to be about which of the econometric estimates to use.
To see why this approach would, barring another offsetting mistake, lead to an underestimate, suppose the employer insurance offerings were completely insensitive to costs. In other words, we are assuming for the moment that employers continue to offer or not offer insurance in 2014 as they did in 2013. Still, the ACA would reduce the number of people getting health insurance coverage through an employer.
First, the ACA significantly raises marginal tax rates (the CBO was the first to notice this, see their most recent discussion). That will cause fewer people to work for any type of employer (with or without coverage). (To the extent this is a significant factor pushing employer coverage rates below forecast, you can expect that the forecasters will blame their miss “on the bad economy” but the bad economy itself would be the result of the ACA).
Second, the ACA will make firms offering insurance to their employees less competitive (in the markets in which they sell their output) with other firms, because the subsidies go only to the latter firms. In other words, some of those subsidies will be passed on to consumers and investors, who adjust their spending patterns away from the items made by the former firms and toward the items made by the latter.
Third, the ACA will change factor market comparative advantage (I have been learning about this as part of a project with Chicago grad student Trevor Gallen). Even if the total amount of workers at employers with and without coverage were constant and there were no change in the allocation of production between the two types of firms, employers without coverage would, in effect, give up some of their skilled workers in order to get unskilled workers. These two types of workers are not traded one-for-one because, by definition, a skilled worker is more productive than an unskilled worker. In other words, comparative advantage by itself says that employers without coverage will get more workers but of lesser average “quality” while employers offering coverage will get fewer workers but of greater average quality.
By my count, there are a total of four margins by which, in response to the ACA, the labor market would reduce the fraction of people receiving health insurance through an employer. As far as I can tell, the literature about the ACA and insurance coverage has so far considered only one.
(This doesn’t mean that a shift away from employer coverage is a bad thing. But I don’t see how we can get the welfare economics right without considering more than one of four the major economic incentives.)