How to measure health care cost control

How to measure health care cost control

I want to propose a four-part test for measuring any particular bill on health care cost control.

short run

long run

Federal deficit



Government health care spending



Private health care spending



In each case, I will define the test so that “yes” is a good outcome:

Test 1: The bill does not increase the federal deficit in the short run.

Test 2: The bill significantly reduces the federal deficit in the long run.

Test 3: The bill significantly slows the growth of government health care spending in the long run.

Test 4: The bill significantly slows the growth of private health care spending in the long run.

I believe our Nation’s long-term fiscal problems, and the problems resulting from the growth of per capita health care spending, are higher priorities to solve than reducing the number of uninsured Americans now. I would rather solve America’s health care cost problems of the future than expand government now. This is my value choice. I expect and accept that others will disagree.

As a result of this value choice, I believe any bill that fails any one of these four tests is fiscally and economically irresponsible, and therefore worth defeating.

There does not have to be a tradeoff. A bill could go after the core policy drivers of health care cost growth, especially the tax exclusion for employer-provided health insurance, and replace it with incentives for individuals to shop for high-value health insurance and high-value health care. Such a bill could meet all of the above tests and significantly reduce the number of uninsured. I will describe such a bill in a future post. Such a bill is not going to be passed by this Congress.

I think the administration would agree with my test. They might define Test 3 to be a subset of Test 2. I think it’s important analytically to separate the two.

In practice the test gets slightly more complex. Test 1, “The bill does not increase the federal deficit in the short run,” breaks down into (1A) “over the next five years” and (1B) “over the next ten years.” The Congressional budget rules require that a bill not increase the federal deficit over the next five years. To his credit, the President and his advisors have also been emphasizing that it is important to meet the same test over the next ten years. From a formal legislative process standpoint, only the five-year window is formally binding, because Congress passed a 5-year budget plan (called a budget resolution). In particular, proponents of a bill will need 60 votes in the Senate for any bill that fails (1A). All other tests can be violated and passed with a simple majority.

I will apply this four-part test framework to each major legislative proposal considered by Congress. I want to begin today by walking briefly through each test.

Test 1: The bill does not increase the federal deficit in the short run.

I would like to make this test more stringent – my personal policy preference would be “The bill reduces the federal deficit in the short run,” especially given the path of expected budget deficits under the President’s budget. The actual test, “does not increase,” is the test in the Congressional budget resolution. It says that at a minimum, any new spending should be offset.

I would also like to make the test apply to federal spending, rather than just the federal budget deficit. I would almost certainly oppose a bill that increases government spending over the next ten years by a few hundred billion dollars, and offsets it with the same amount of tax increases. Again, I’m matching my test to the minimally binding one that Congress will apply to itself. This means that this test for me is one-way: any bill that fails it should be opposed, and some bills that pass it should still be opposed, because they dramatically increase the size of government. Still, for the purpose of this exercise I am applying the looser deficit-based short-term test.

By choosing a looser short-term test than I would prefer, I believe I accomplish two goals:

  1. This test conforms with the formal budget rules that will govern this bill (measured over a five year period).
  2. This test fits the “Blue Dog” / conservative Democrat / moderate Republican view of the world. I think I’m taking away an excuse for them to object to my four-part test.

Test 2: The bill reduces the federal deficit in the long run.

For each of these tests, I’m defining “long run” as more than ten years. That’s an arbitrary breakpoint.

While I’m willing to say I could swallow some bills that do not increase the short-term budget deficit, a bill must significantly reduce the long-run federal deficit to be fiscally responsible. Given that our long-term federal deficit path is unsustainable to the point of national economic collapse, and given that health care cost growth is one of the primary drivers of that deficit path, not making the problem worse is insufficient. A bill must result in dramatic reductions in future budget deficits to be fiscally responsible.

This test interacts with Test 3 in a somewhat subtle way. While Test 1 is a deficit test, the facts of our long-term budget problem mean that Test 2 is driven by Test 3, which is about government spending on health care.

Test 3: The bill significantly slows the growth of government health care spending in the long run.

Test 2 is about the long-term budget deficit. Test 3 is about long-term government spending on health care. The President and his budget director are correct when they identify unsustainable per capita health care spending as a primary driver of long-term deficits. (They are incorrect when they identify it as the primary driver of long-term deficits, and dismiss the importance of Social Security spending and aging of the population, as the President did yesterday. I will return to this point in a future post.)

In mid-April I explained that America’s long-term budget problems are driven by unsustainable spending growth, and not by the level of taxation. I think it’s one of my most important posts. I hope you will find time to read it if you have not done so already. Here is the key graph:

taxes and spending long term trends

On the above graph, the white line is federal spending, and the dotted lines are various tax policies. The expanding gap between the white line and the dotted lines is the federal budget deficit. You can see the gap (deficit) explodes as the spending line pulls away from the tax lines.

America’s long-term budget problems are driven entirely by the difference between the slope of the white spending line and the dotted tax lines. Over the long run, a constant tax policy always grows as fast as the economy, and so it remains flat on a graph that measures quantities as a share of the economy. So while we can and do debate about the level of the dotted tax lines, they’re always going to be flat.

You can see that any flat tax line cannot keep up with a rapidly growing upwardly sloped white spending line. Even if you were to rais the flat dotted line to 25% of GDP, you would still have a long-term deficit problem because of the slope of the spending line. The key to success is not just lowering federal spending, it’s tilting that white line dramatically downward. This is what the President and his budget director correctly mean when they say we need to bend the (government) cost curve downward. And they deserve praise for identifying federal health care spending as a major driver of that white line’s slope.

It is therefore odd and self-contradictory that they have proposed raising taxes to offset the higher spending of a new health care entitlement for the uninsured. While you can technically meet my short-term Test 1 by doing so (in a Blue Dog / centrist way that I would oppose, but you’d meet it), it is mathematically impossible in the long run to offset a new health care entitlement with higher taxes, unless your bill also slows the growth of health care spending in other ways.

To put it graphically:

  • The new health care entitlement for those who are now uninsured would raise the level of the long-term white federal spending line.
  • Even if you increase taxes, raising the dotted federal tax lines so that the deficit gap between spending and taxes over the next five or ten years does not increase (thus meeting my Test 1), in the long run the new health care spending will grow faster than the economy, while the new tax revenue stream will grow at the same rate as the economy. You will therefore be exacerbating the long-term deficit problem caused by the white spending line above.
  • The only way to solve this is if you make other changes in the bill that bend downward the slope of the white line.

This last bullet is the President’s stated solution. In effect, he is saying, I’m OK raising long-term federal spending on a new health care entitlement, and thus raising the level of the white line in the long run, as long as we raise the dotted tax lines to offset it in the short-run, and as long as we make other changes to tilt that white line downward (or at least not upward so much.)

The President and his allies have a problem, in that their specific policy of expanding pre-paid health insurance to tens of millions of uninsured Americans will instead increase the slope of the white spending line. The academic evidence is clear that as third-party payment for health care increases, sensitivity to cost decreases and health care spending (total and governmental) increases. Creating a new entitlement for the uninsured helps the uninsured. But it worsens our long-term budget problem in two ways: it raises the level of the long-term spending line, and it increases its slope. Both exacerbate an already-devastating long-term federal budget picture.

So for the President to meet his stated goal, and to make any significant progress on our long-term budgetary problems, the rest of the bill must not only bend the spending line downward, it must do so by more than these two factors that raise the white line by creating a new health care entitlement. I think it’s a mistake to make your most serious problem worse before trying to solve it.

Test 4: The bill significantly slows the growth of private health care spending in the long run.

This is closely related to but separate from Test 3. I praise the President for correctly identifying society-wide health care cost growth as the problem to be solved, rather than just government health care cost growth or the number of uninsured. Private sector health care cost growth is what keeps the number of uninsured high, and it is what squeezes the wages and budgets of more than 200 million Americans with private health insurance. We must make policy changes that stop distorting behavior to encourage unsustainable cost growth in private sector health care.

As I said above, the expansion of third-party payment for the uninsured exacerbates this problem, as would any policy changes that might discourage people from moving to high-deductible plans, or discourage people from shopping for health insurance or medical care based on quality and price.

The President correctly identifies this problem. He admirably says it is a condition that must be met by health care legislation. Unfortunately, he has made no specific policy proposals that would achieve this goal. The President and his budget director emphasize policies that would provide private sector consumers with better information about the health care they use. They have proposed policies that would change government spending policies. They have proposed no policies that would change incentives for private consumers of health care. (I wrote about this in April.) Without such policies, you cannot meet Test 3 or Test 4. And without such policies, expanding government entitlement spending is horribly irresponsible in the long run.

18 responses

  1. You should try reading the Republican’s alternative: The Patients’ Choice Act of 2009 (link to short summary: to long summary: This is, in my opinion, the best and most comprehensive piece of healthcare reform legislation addressing all of the issues: for example…affordable healthcare coverage, coverage for ALL American’s, maintaining and incentivising research and development, incentivising competition within the marketplace to lower the cost of insurance, and addresses entitlement reform doing away with Medicaid! This extremely well-designed alternative would NOT COST AMERICAN’S ONE DIME…which is certainly not something that can be said for the Democrat proposal! Socialized medicince has been proven to fail everywhere it has been tried. American’s currently have the best healthcare in the world, and the Democrat proposal would destroy that…yet another attempt for a massive power grab by the left (not surprising)!!!

  2. The question is not the fiscal aspect of it but where in the Constitution is there the authority to even provide it in the first place?
    Everyone has the right to seek medical care, everyone has access to medical care but no one has a right to have the care paid for by someone else.
    I suppose up to now the government has been able to skirt by since doctors are not obligated directly to accept medicare and medicaid patients. However when that changes and all compensation is directly set by the government then the doctors have become involuntary employees of the government. Marvelous! Then they might as well unionize and strike or threaten to strike like other government employees.

  3. Mr Hennessey-

    You say: The academic evidence is clear that as third-party payment for health care increases, sensitivity to cost decreases and health care spending (total and governmental) increases.

    This intuitively sounds correct, and I have argued this on other forums from a theoretical standpoint. But can you point me to some actual research purporting to show this in the real world? Because most of the people I meet on the other side of this issue do *not* concede this point. They are suspicious of the Market’s ability to reduce costs and especially suspicious of the idea that exposing consumers to cost would actually HELP them by forcing costs down.

    Indeed, most of these people then point to studies of other nations with single-payer plans and show how they have kept costs under control without exposing the customer to the cost.

  4. Keith –
    On the spending graph. First, I generally agree with your analysis, but want to play devil’s advocate.
    I would imagine that you could have drawn the exact same graph, say, 10 years ago. It would look roughly the same: spending is flat (or even dipping slightly) in the short run (5 years) and then goes up like a hockey stick after that. Only “today” is 1999 not 2009.

    Yet, somehow, every year, Congress finds a way to make things work (stay within some finite bound around the long run average) every time. Basically, they “bend the curve downward” gradually over time for the next five years.

    Why can’t that pattern continue?

  5. Jesse, the fundamental difference between the upcoming years and the prior decades is the shift in demographics which will occur. The problem isn’t so much discretionary spending by the government, which could be kept in check (although this congress is showing very little inclination to do so), but rather the fact that a huge number of people from the baby boomer generation are about to retire and switch from being highly productive contributors to the system to drawing from the system. The ratio of people paying in to fund these entitlements to people pulling money out of them will drastically shift. It’s a challenge the system has not yet faced and is ill-equipped to handle.

  6. Absent from your analysis is any mention of medical quality. If costs are controlled, and medical quality sinks deep into mediocrity, then have we really cured the ailing system? Regarding the cost side of the equation, it’s bizarre that the public option proponents won’t say how much it all will cost and who will pay for it. See

  7. Pingback: Keith Hennessey: How to measure health care cost control | The Kansas Progress

  8. Innovation drives health care costs up. But it also drives up medical care quality. Conditions that were fatal 20 years, now do not significantly affect a person’s lifespan. People who were confined to wheelchairs 20 years ago can walk and even run marathons today. Innovation is a good thing. Increased medical care quality is a good thing.

    They key is that it must be voluntary spending on health care. If I choose to spend 35% of my income on health care, who are you to say otherwise? People need more exposure to more routine health care costs. Higher deductibles and having to chose what scope of coverage to purchase would help. Otherwise, how would someone know that they are paying $300 per week for allergy medicine? The individual needs to perform a cost-benefit analysis of what health care gives her returns. The government should not be making the choice. And employer insurance plans prevent employees from individually weighing options.

    I will save my comments on tort liability for doctors, hospitals, and pharmaceutical companies for another post.

  9. OK, so I may be a tiny bit confused on account of the strange way the costs of the various proposals are worded, but it seems like there should already be money to address a big part of the problem. As near as I can tell, there is an annual 350 billion cost for medicaid at the federal level. These costs are augmented by state govts as well making the true total higher. I am fuzzy here as each state seems to get some matching, but the matching is inconistent as every state does not cover the same way.

    For the last few weeks, I keep hearing from people like Obama, Pelosi, and frank, that there are 400 billion in ‘mistakes’ that can be saved. Sounds way to easy to me but hey, I get confused sometimes.

    Anyway, as near as I can tell, they should have 750 billion available in the current budget for ‘health care’ for the ‘poor’. Now as we know, the real poor, like illegals and uneducated in govt program types, show up at the hospitals for health care anyway.

    So, what if an expert community organizer, like Obama, simply chose to take this already accounted for money, and instead of trying to directly insure any of the ‘poor’, he simply used it to ‘extend’ free clinics in all the ‘appropriate’ community places? It would seem that 750 billion could go a long way. Also, he would not need to mess with any of the existing insurance structures, the ama, labor unions, or businesses.

    Obviously, services at these clinics may not reach the level of quality available to people that ‘have their own’, but it seems like it should provide many people with basic ‘get by’ services. And since the vast majority of people that do work would still have their ‘better’ private coverage, there will be incentive for people to work a bit harder and get a little more out of life for themselves.

    Notice, I did leave the long term elephant in the room, medicare, out of this discussion. But hey, they will not be able to fix all of these problems in one shot no matter how big of a gun they use!

    … seems like the government actually created this insurance gap for the ‘near poor’ by overpaying for health care for the ‘poor’, but that is another confusing story …

  10. I strongly object to simply saying the goal is to “reduce growth in health care spending” without taking a position on how/why we’d want to do that.

    What should the new incentives be? How will quality be affected? How will innovation be affected? Who will decide? Why will that be better than now?

    Health is really important, so it’s not absurd to think that we SHOULD spend increasing fractions of our income on health care, so long as we are getting something worthwhile for our dollars. On the other hand, there may be good reasons to believe that much of current spending isn’t doing anybody much good, and it might be possible to eliminate some of that while preserving the good spending. Or maybe not.

    But these are the issues you must grapple with, rather than simply assuming that “reduced growth” in spending as it is a good in itself, regardless of how it’s accomplished.

  11. “I believe our Nation’s long-term fiscal problems, and the problems resulting from the growth of per capita health care spending, are higher priorities to solve than reducing the number of uninsured Americans now.”

    Our fiscal issues, health care costs, and uninsured are all very much related. It is not an either/or issue.

    Don’t have a link but I recall a poor kid who had an infected tooth, couldn’t afford to go the dentist, infection spread to his brain, ended up having very expensive brain surgery at taxpayer expense, and then died.

  12. Here it is. I beg of you to read it. An $80 extraction would have saved his life and saved the taxpayers $250,000.

    Twelve-year-old Deamonte Driver died of a toothache Sunday.

    A routine, $80 tooth extraction might have saved him.

    Deamonte Driver, sitting next to his mother, Alyce, shows the scars from incisions for his brain surgery.
    Deamonte Driver, sitting next to his mother, Alyce, shows the scars from incisions for his brain surgery. (By Linda Davidson — The Washington Post)

    If his mother had been insured.

    If his family had not lost its Medicaid.

    If Medicaid dentists weren’t so hard to find.

    If his mother hadn’t been focused on getting a dentist for his brother, who had six rotted teeth.

    By the time Deamonte’s own aching tooth got any attention, the bacteria from the abscess had spread to his brain, doctors said. After two operations and more than six weeks of hospital care, the Prince George’s County boy died.

    Deamonte’s death and the ultimate cost of his care, which could total more than $250,000, underscore an often-overlooked concern in the debate over universal health coverage: dental care.

  13. Any solution will ultimately have to make a tradeoff between long-run benefits and short-run costs. Indeed, I have never seen any complex problem ever that does not involve making this tradeoff. Therefore, I believe your value choice is logically incorrect, especially this part

    Test 1: The bill does not increase the federal deficit in the short run.

    I would like to make this test more stringent – my personal policy preference would be “The bill reduces the federal deficit in the short run,”

    This just does not make sense. it’s like telling a high school senior to go to a community college instead of Harvard because Harvard may involve increase in debt in the short run.

  14. Michael Kirsh MD said: “Absent from your analysis is any mention of medical quality. If costs are controlled, and medical quality sinks deep into mediocrity,”

    I would ask the doctor to then explain how the Mayo Clinic consistently delivers high quality care at a lower cost…

  15. Pingback: The President’s press conference: health |

  16. Pingback: CBO calls a TKO on the House health bill – Blog Title  |

  17. Technically, its the Commerce Clause in the constitution. In reality, its just the gradual change in the size of the government and the current political norm.


Get every new post delivered to your Inbox.

Join 6,544 other followers