The University of Michigan
555 South Forest Street
Third Floor
Ann Arbor, MI 48104-2531

T 734-936-9842
F 734-998-6341



The U.S. spends far more per capita on health care than do other nations but Americans are not getting the outcomes one would expect with that higher amount of spending. Some economists fault the tax code’s overly generous employer exclusion—worth about $200 billion annually—for that. If the employer exclusion were ended, as President Bush has proposed, what changes would the health insurance marketplace face? And how would it impact the uninsured?

Katherine Baicker, Ph.D., health economics professor in the Department of Health Policy and Management at the Harvard School of Public Health.

  Harvard Economics Professor Katherine Baicker focuses her research on the financing of health insurance, public program expenditures, and fiscal federalism. Baicker served as a Member of President Bush’s Council of Economic Advisers between 2005-2007, a time when Bush first proposed replacing the employer subsidy in the tax code with a flat deduction for anyone purchasing health insurance. Baicker, who also is a research associate at the National Bureau of Economic Research, spoke with ERIU about the reasons for the disappointing return the United States gets on its health care spending, tax code reforms and other changes that might boost that return, as well as the need for policymakers to look at health care spending and universal coverage together.


What is the chief reason we are not getting the return on investment for our health care dollar?

BAICKER: The problem is that many factors contribute to not getting high value health care. Many are related to how we finance health care. The tax code, for example, subsidizes certain forms of insurance but not other forms. That pushes people towards policies that might not get as high value care as they would otherwise. On the public side, via Medicare for example, we pay for quantity of care, not quality of care. That pushes us towards care that is not of as high value as it might otherwise be.

Then there is a lot of necessary information that we lack but that is needed to take full advantage of the system. We need to know more about comparative effectiveness and best practices. That would help physicians provide high quality of care. Patients need to know more about the quality of care they consume, and the real price of the resources they use so they can make decisions right for them. A lot of things need to change.

Let’s start with tax code provisions, specifically the employer exclusion. Can you outline what the employer exclusion is and how it distorts purchases of health insurance?

BAICKER: Right now, if you get health insurance through your job, and your employer contributes to the premium, those contributions are paid with pre-tax dollars; there are no taxes paid on that benefit that you get through work, but you do have to pay taxes on your wages. If you buy insurance on your own, in the individual market, you buy that insurance with after-tax dollars. With some exceptions, the health care you buy outside of insurance – out-of-pocket spending – is also paid with after-tax dollars too.

That means any health care you consume through an employer insurance plan is discounted relative to any other kind of health care that you consume. For a lot of people you are talking about a 40 to 50 percent price difference, if you count payroll taxes and income taxes. That’s why most people want to get insurance through their job - because it is so much more affordable. The employer-provided insurance policies tend to cover a lot of routine expenses through higher premiums, such as a doctor’s office visit. If a visit is covered by an employer plan, you implicitly pay a lower pre-tax price for the visit, but if you pay for that doctor’s office visit on your own, you pay with after-tax dollars, making the visit much more expensive.

That is one of the main reasons that health insurance doesn’t look like other kinds of insurance we purchase. Insurance is fundamentally about risk, protecting ourselves against potentially very high expenses. But, that’s not what most health insurance looks like and it’s largely because of that bias in the tax code – the bias against forms of insurance that have higher co-payments.

How does the employer exclusion contribute to the gap between health care spending and outcomes?

BAICKER: First, it differentially subsidizes the kind of insurance with lower cost-sharing, minimizing what people pay for out-of-pocket. Health insurance is really valuable; everyone should have the important protections it brings. But it may not make sense to push people into policies that have lower deductibles and co-pays versus higher deductibles and higher co-payments. Subsidizing employer-provided health insurance also makes it unclear how much we pay for health care. If you ask people, “How much did your health insurance cost?” many can’t tell you because they don’t know how much their employers are contributing.

If you tell people they are consuming a $10,000 to $12,000 health insurance policy, many are shocked. If you gave them the choice: would you rather have that policy or a policy with a lower premium that covered less stuff and get to keep the difference in wages, many might rather have the wages to spend on other family priorities, but that isn’t an option because of the tax code bias that favors insurance over wages. People could only get something like 60 percent of the premium reduction back in net wages because those wages would be subject to taxation.

What impact does the employer exclusion have on wages?

BAICKER: Workers implicitly pay for their own health insurance in the form of lower wages. It’s not that employers generously give their workers health insurance policies out of kindness: they count it as part of total compensation costs. When employers try to attract employees, they offer them a package of wages and benefits, like retirement benefits and health insurance benefits and maybe parking, etc. Employers take into account the whole cost of hiring the employee. As health insurance costs and premiums go up that adds to the cost of hiring the employees, so wages won’t go up as quickly as they would otherwise. That is largely invisible to most employees because nowhere can they see written down how much their employer is contributing to their health insurance. They don’t have a choice about that trade-off. In fact, if they were to choose higher wages and a more basic health insurance policy, they would pay an income tax penalty for that.

President Bush has proposed doing away with the employer exclusion and redistributing it as a flat deduction to individuals purchasing coverage. Can you highlight a bit more of the details?

BAICKER: Sure. There are two different biases in the tax code. One is an overt bias that gives those buying insurance from their employer a tax benefit that those buying insurance on their own don’t get. Then, there is a more subtle bias against different methods of cost containment. If you get an insurance policy that keeps costs down by raising the deductible, requiring you to pay more out-of-pocket, the tax code for the most part requires you to pay those out-of-pocket costs with after-tax dollars, so you pay more. On the other hand, if you get a policy that tries to keep costs down by restricting the number of physicians that you might see or by having a more restricted formulary there is no comparable tax penalty. The tax code is biased against out-of-pocket expenses relative to other ways in which you might purchase health insurance. If you pay more in copayments in exchange for a lower premium, you are giving up a tax benefit.

President Bush’s proposal aimed to eliminate both of those biases by providing a flat tax benefit. No matter how much your insurance costs and no matter where you get it—either from an employer or on your own—you would get the same tax break. That means people buying a really expensive policy and those purchasing a very basic policy would get the same tax benefit. If you get your insurance on your own or if you get it from your job, the tax benefit is the same, a $7,500 deduction for individuals and $15,000 for couples.

Is this key in getting health insurance to look more like traditional forms of insurance?

BAICKER: It would certainly go a long way in that direction. On its own, though, it’s not a magic bullet. Many other things have to happen, too, like better information for consumers, physicians and other providers.

If the president’s proposal was adopted, what would the likely impact be on the uninsured? How many of the 46 million uninsured is it estimated would purchase coverage?

BAICKER: There is a lot of uncertainty about the exact number, but estimates range between 5 million and 12 million.

Who would these people be?

BAICKER: Many of them would be currently uninsured. They are getting no tax benefit now, but they would get a $15,000 deduction for family coverage, which would lower the cost of insurance for them. That doesn’t mean everybody would get insurance but it would be a big contribution towards the cost of the average policy.

Who would be the least likely to gain?

BAICKER: People with the lowest income might be least likely to take up the new benefit, since they would get a lower tax benefit under a flat deduction. It would still go a long way toward reducing the cost of insurance, but it might not be enough for them. Also, some people might not choose to get health insurance – young healthy people sometimes think they might not need it. The highest income people with the most expensive insurance policies would see their tax bills rise. On net, the reform would likely make the tax code more progressive, but only slightly.

What effect would the president’s proposal have on employer-sponsored insurance?

BAICKER: It would level the playing field between employer-sponsored insurance and individually purchased coverage. Right now, there is a bias toward employer-sponsored coverage, so the proposal has the potential to induce some people who get coverage through their employer to get it on their own instead. Some have worried that it would encourage employers to drop coverage. In some cases that certainly might happen. That poses important transition problems for people in firms that stop offering coverage. When they go to the existing individual insurance market, they may be underwritten based on current health status. People in poor health who are currently covered in an employer group and then go to the individual market might see their premiums go up by a lot. This group in particular would need extra help during a transition from one system to another. That can be done in a number of different ways – through state high-risk pools or risk-adjusted vouchers. Targeting extra resources towards those groups might relieve some of the difficulty.

What effect would the Bush proposal have on the individual market?

BAICKER: It would certainly increase coverage in the individual market. More uninsured people would buy insurance in the individual market. Having thicker individual markets is a good thing, but it should probably be accompanied by some individual market reforms. It would also be a big net transfer to people already purchasing insurance in the individual market.

Some people suggest the president has it half-right. Just doing away with the employer exclusion and providing the flat deduction is not efficient tax policy. What’s more, he doesn’t propose changes to the individual market. What additional changes are needed to get the biggest bang for doing away with the employer exclusion?

BAICKER: Coupling tax code reforms with individual insurance market reforms would definitely make the tax reforms more effective. I would take some issue with the statement that providing a flat deduction is not an efficient tax policy. It is much less distortionary than the tax policy we have now. An alternative would be a flat tax credit, which would have similar implications for eliminating biases in the tax code, but would have different distributional implications and would probably result in greater increases in insurance coverage. But, both, I think are more efficient tax policy and health policy than the current system. To get the most out of a policy like that, though, you would also want more information available to people and regulatory reforms in the individual market that would help make basic affordable insurance continuously available to people regardless of their health status.

Such reforms as?

BAICKER: For example, once you are insured, you shouldn’t be re-underwritten: your premiums shouldn’t go up because of changes in your health status once you get insurance. There are better protections against that in some states than in others. You certainly would want to help strengthen that. And you would want to make sure that basic, affordable policies were available to people in every market. There are some regulatory reforms that could help promote that as well. There also needs to be enough people participating that you have a good risk pool.

What would need to be done more specifically about the chronically ill who go into the individual market under this kind of new system?

BAICKER: The problem of a chronically ill person trying to get an affordable policy on the individual market is not about insurance. Once you are already sick, it is too late to insure against getting sick. It’s like buying insurance for an auto accident that has already happened. If you are insured and you get sick, your insurance is there to protect you against higher expenses now and in the future. Ideally everyone would get insurance when healthy and be protected against the costs of falling sick. Unfortunately, that’s not the case. People who are already sick and uninsured need help, but that help is a transfer, not an insurance product. That help could come in the form of subsidized high-risk pools, risk adjusted vouchers, etc., and could be based on income as well as health. But fundamentally you don’t want to design an insurance system around low premiums for people who are already sick – you want to design a system that promotes obtaining insurance coverage when healthy and broad pooling of the risk of getting sick.

What’s your view on universal coverage? Is covering everyone and trying to address cost issues superseding getting costs under control without the goal of universal coverage? Where do you come down on this? How do you view this?

BAICKER: We need to address the problem of rising costs and the problem of the uninsured together. Proposals that focus exclusively on getting people covered by insurance run the risk of not being able to afford that coverage tomorrow if costs rise. Similarly, policies that focus just on containing costs will miss opportunities to promote much better health and more efficient use of health resources by getting people insured.

The two goals are really intertwined and are more likely to be achieved when addressed together. That said, I think it is dangerous to think we can cover the uninsured without spending any more money. A lot of people focus on the inefficient way that uninsured patients get care now, which is true. People don’t consume health resources in a way that produces the best health when they are uninsured. Getting people to get good preventative care – rather than showing up in the emergency room – would also mean spending less money to achieve the same health outcome. But, that is just part of the picture. When you get people insured, they consume more health resources. They get a lot more health for it, so it may be a very good deal, but they are consuming more, not less. We don’t want to delude ourselves into thinking that we can cover the uninsured and save money simultaneously.

We should focus on policies that get higher value health care, which might then end up spending more on some people, spending less on other people, spending more on some procedures, spending less on other procedures. There will be fundamentally hard choices that have to be made, and policy makers have to think about who is going to make those choices. Somebody – individuals, the government, employers, insurers – has to decide how to allocate scarce resources. We want to set up a system where we’re devoting resources to producing as much health as we can.


Back to top

Funded by The Robert Wood Johnson Foundation, ERIU is a five-year program shedding new light on the causes and consequences of lack of coverage, and the crucial role that health insurance plays in shaping the U.S. labor market.