• Simply put: Employer-sponsored insurance tax subsidy

    Simply put” is an ongoing series. See the introduction for an explanation of the series and the full list of topics that have been or may be covered. Feel free to suggest other topics in that post.

    Every dollar an employer pays directly toward health insurance is not taxed. Many, but not all, of the dollars employees contribute toward their premiums are not taxed either (it depends on how their health benefits programs are structured). It’s been shown in many studies that what an employer doesn’t pay in health care premiums, it does in wages. That is, if an employer pays a dollar less in premium, it’s employees earn a dollar more in wage, on average.

    But, unlike employer-paid premiums, wages are taxed. That difference makes health insurance relatively more attractive than it might otherwise be. Employees get a discount through the tax system. That gift from Uncle Sam costs the government $250 billion per year, no chump change!

    What does Uncle Sam, through our individual actions, buy with that tax expenditure? Two things: (1) more firms offering employer-based health insurance plans and (2) the offering of more generous plans.

    Economists have estimated the effect of the tax subsidy on these two phenomena. The classic result, from Gruber and Lettau is that a one percentage point increase in the tax subsidy leads to a 0.25 percentage point increase in employer offers, but a much larger proportional increase in the generosity of plans offered. That is, the tax subsidy has a bigger effect on plan generosity than on whether a plan is offered.

    Recently, in work with Steve Pizer and Lisa Iezzoni, I re-estimated the effect of the tax subsidy on employer offers and found it to be twice as large as what Gruber and Lettau found.

    Either way you slice it, the employer-sponsored health insurance tax subsidy supports the employer-based insurance system and encourages more generous health plans than would otherwise exist. The consequence is more people are covered, but they also spend more through their insurance plans than they would otherwise.

    Further Reading

    Pizer S, Frakt A, Iezzoni L. (2011). The Effect of Health Reform on Public and Private Insurance in the Long Run. HCFE Working Paper.

    • I have to take exception to the claim that more people are covered because of the tax subsidy. Although I’m sure there would be people who would choose to not buy insurance if given the same amount of money in wages, when we tie insurance to employment, it means the unemployed are necessarily uninsured. As I recall, it was estimated that some 45% of the famous “47 million uninsured” were only uninsured temporarily because they were temporarily unemployed.

      • @Scott – And yet, within the system that exists, the more employer-based insurance is subsidized, the greater the proportion insured. That’s what the evidence shows and I’m sticking with it.

        (That does not mean I support the employer-based system. I do not. We can do better.)

    • I think that employers providing health insurance made sense when spending was a much smaller part of GDP but now it seems a bad way to fund health spending. The deductibles in employer plans are lower than what most people opt for when they have to buy their own health insurance. With Medical care are running at 17% of GDP we need individuals to help making choices with case by case individual information.

      SO how to hasten the end of employer provided health insurance. Maybe we could sell a change by increasing the standard deduction buy an amount that would make up for the current tax savings on an average employer based account.

    • @Scott
      also, the cycle of in and out of coverage means that far more people are exposed to a spell of being uninsured than the number at a point of time estimates….if they get ill or hurt then, could be bad. So, you could argue the cyclical nature means problem less severe or more severe. In fact, if it was always the same folks, it would be easier to address.