The most important question in health reform is this: How should the government subsidize private health insurance? And both parties are in broad agreement on the answer.

The subsidy should be in the form of a fixed sum tax credit.

Consider the traditional way of encouraging health insurance. For employer-provided insurance, the subsidy is in the form of an exclusion. Unlike wages, employer premium payments are not included in the taxable income of the employee.

For the self-employed, the subsidy is in the form of a deduction. And for other individuals, health insurance premiums and medical expenses can be deducted to the extent that they exceed a certain percent of income (currently 10%).

In all three cases, people face the following perverse incentive:

• If they purchase more insurance their taxes will go down and if they purchase less health insurance their taxes will go up. In this way, the tax system encourages all of us to choose more generous coverage than we would otherwise select.

• If we combine a 15% (FICA) payroll tax with a 25% federal income tax and a 5% state and local income tax, a middle income family is facing a 45% marginal tax rate. In high tax states, the rate can exceed 50%—even though the family is far from wealthy.

• At a 50% marginal tax rate, government at all levels is paying for half the cost of any additional insurance the family chooses to buy. Insurance that costs $1 will be viewed as worthwhile, even if it is worth only 51 cents to the buyer.

In other words, insurance can be extremely wasteful and still be attractive to subsidized purchasers.

Alternatively, if the buyer saves a dollar by choosing less generous coverage, that dollar will become taxable income. The government will seize one-half of it.

No wonder our health care system has so much waste in it.

With a fixed sum tax credit those incentives change remarkably. Here is how it works in the ObamaCare exchange:

Subsidy = Premium* (1 – hY)

where Premium* is the second lowest premium charged for Silver plans, Y is the buyer’s income and h is the maximum percent of income people have to pay for such a plan. Notice that this sets the subsidy at a fixed dollar amount.

The subsidy is refundable: the buyer gets the credit even if he doesn’t owe any income taxes. It is also advanceable: within the exchange, the subsidy goes directly to the insurer, bypassing the buyer altogether. But most important, the subsidy is the same whether the individual chooses another Silver plan or a Bronze, Gold or Platinum plan. That means that buyers who choose more expensive plans pay 100% of the extra premium out of their own pockets. Alternatively, buyers who choose a less expensive Bronze plan get to keep 100% of the savings from making that choice.

Any extra expense is paid with after-tax dollars. Any reduction in expense increases the buyer’s after-tax resources. Since most other consumption is also paid for with after-tax dollars, this puts health insurance premiums and other goods and services on a level playing field.

With a fixed sum tax credit, buyers are not encouraged to over-insure or under-insure. Every costly feature of health insurance (lower deductibles and copayments, wider networks, more generous benefits) will be at the expense of all other ways of spending the consumer’s dollars.

ObamaCare is, of course, a Democratic health reform. But all the Republican health reform proposals make use of this idea as well.

If there is a fault in ObamaCare in this regard it is that it doesn’t go far enough. The credit vanishes after family income reaches 400% of poverty and the traditional subsidies (described above) kick in. Also, it leaves the traditional tax treatment of employer-provided coverage fully in place.

In the 2008 election, John McCain proposed a more radical approach: a fixed sum tax credit to replace all existing tax subsidies for health insurance. The credit would be the same regardless of where the insurance is obtained—at work, in the marketplace or in an exchange. The legislative version of the McCain plan was the Coburn/Burr/Ryan/Nunes bill.

Although President Obama and Democrats in Congress often say the Republicans have no alternative to ObamaCare, they appear to have very short memories. The Obama campaign spent millions of dollars attacking the McCain health plan (demagoguing it, actually). Judging by the number of Obama TV ads, it was the principal issue of the entire election. And when ObamaCare finally came to a vote in the Senate, Harry Reid refused to allow a vote on the Coburn/Burr alternative.

Other Republican approaches have not gone as far as the McCain approach, but they too adopt the fixed tax credit as the vehicle for subsidy:

• The House Republican Study Committee plan (Roe bill) replaces existing tax subsidies with a uniform tax credit, but since the credit is not refundable, it provides very little benefit to the bottom half of the income distribution.

• Rep. Tom Price has a bill that creates credits for those who buy their own insurance, but leave the employer-based system in place.

• A new Coburn/Burr/Hatch bill mimics the ObamaCare approach (tax credits that phase out for the individual market, but less generous than under the ACA), but it limits the amount of health spending that can be excluded from income at work.

• A 2017 Project proposal would make the new Coburn tax credit independent of income and would even create a special tax credit for Health Savings Accounts, but it would leave the current employer system intact.

Here’s the bottom line: it is now well established in both political parties that the credit approach is better than all others. For example, prior to becoming President Obama’s economic adviser, Jason Furman endorsed a health reform that looked very much like the John McCain proposal. So in thinking about how to reform ObamaCare (or replace it) we all should be thinking about how to extend tax credits to everyone.

All of the approaches above are primarily focused on how to subsidize third-party insurance. But presumably, we don’t want third-parties (insurers, employers and government) to pay every medical bill. Significant cost-sharing means that individuals must self-insure for their portion of the costs and one way to formally do that is through a Health Savings Account (HSA). How does that fit into the tax credit scheme?

Mark Pauly and I answered this question almost 20 years ago.

Consider again how the tax credit works. Taxpayers get a dollar for dollar subsidy for a certain amount of health insurance. Presumably this is the core insurance that we want everyone to have. The current system subsidizes the last dollar just as much as the first dollar. The tax credit only subsidizes the first dollars. Marginal purchases are all made with unsubsidized, after-tax dollars.

So in order to put third-party insurance and individual self-insurance on a level playing field, we need deposits to HSA accounts to also be made with after-tax dollars. That means that the appropriate account must be a Roth-type account, with after-tax deposits and tax-free withdrawals.

A Roth HSA could replace the plethora of existing accounts: Regular HSAs, Medical Savings Accounts (MSAs), Health Reimbursement Arrangements (HRAs) and Flexible Spending Accounts (FSAs). (See my survey article in Health Affairs.) It should be completely flexible—wrapping around any third-party insurance plan and permitting maximum experimentation and innovation to find new and better ways of managing health care costs, especially the cost of chronic illness.

A model to follow is the Cash and Counseling program in Medicaid, where homebound disabled patients manage their own budgets. (See this international survey of similar accounts.)

From their inception HSAs were criticized by many Democrats as being a sop for the wealthy and the healthy. Yet Jay Rockefeller was one of a number of Democrats who greatly admired the Cash and Counseling program. Given this history, perhaps we should engage in some relabeling.

To get bipartisan harmony, we might consider calling the Roth HSA accounts “Rockefeller Accounts.”