• Mindless federal spending decisions: The case of the NIH

    The US fiscal year ends on September 30 and the government reaches its borrowing limit a few weeks later. Because there will be drama about whether the government will, like, function, there won’t be much discussion about whether we are spending on the right things.

    During the 2011 debt ceiling crisis, the President and Congress passed a sequester that forced the National Institutes of Health (NIH) to spend 5% less on health research. From Science Magazine,

    Including ongoing (already awarded) grants that are ending, the total number of research grants will drop by 1357 to 34,902 awards.

    But there are still 35,000 active research projects. What’s the harm?

    Set aside the consequences to researchers (including me) who compete for these grants. The NIH should be evaluated on whether it saves lives, not jobs.

    It does save lives. David Cutler and Srikanth Kadiyala estimated the returns to investment in medical research by looking at improvements in survival from heart disease, the most common killer of Americans. (Austin has written on Cutler’s work on the value of health care here, here, and here.)

    The value of medical research is longer, healthier lives. …the average 45 year old will live an additional 4½ years today over what he or she did in 1950, solely because cardiovascular disease mortality has decreased. About two-thirds of this change, or 3 years in total, results from improved medical treatment and one-third, or 1½ years, results from behavioral change… Using [standard methods for estimating the values of years of additional life], the value of medical treatments [per American citizen] is about $120,000. Similarly, the value of behavioral changes is $30,000.

    OK, improved medical treatment of heart disease gave each American $120,000 worth of future life. But is this still a good deal after you subtract what that American had to pay for living longer?

    The typical 45 year-old can expect to spend about $30,000 in present value on cardiovascular disease care over their remaining lifetime. There was some spending on cardiovascular disease in 1950, but this was small. The increase is therefore about $30,000. Compared to the $120,000 of benefits, it is clear that this spending increase is worth it. We estimate a rough rate of return of 4 to 1 for medical treatment changes. Typically, society is happy with rates of return of about 1.1 to 1. The rate of return we estimate is an order of magnitude higher.

    NIH research on health behavior and cardiovascular disease was an even better deal.

    The cost of behavioral change is the money spent on researching these behaviors and conveying that research to the public. From 1953 to 1997, NIH spending on all factors related to cardiovascular disease—-including behavioral knowledge and other research—- amounted to around $3 billion dollars per year or $10 per person per year. For a person with a 50 year taxpaying span, this is about $500 over their lifetime (ignoring discounting). There are also costs of disseminating this information to people… another $500 of expense. Total costs of behavioral change are therefore about $1,000 total. [Hence] Research and dissemination costs of about $1,000 per person brought benefits of about $30,000 per person. The return here is 30 to 1.

    Of course, the NIH spends money on many other disorders. Our research investments on some disorders were probably less successful than cardiovascular research. However, there are likely also some disorders — such as AIDS — where we achieved an even better return. If you accept Cutler and Kadiyala’s analysis, we get a far above market return on investments in health research. Reducing the deficit by spending 5% less on the NIH isn’t responsible. It’s stupid.

    However, Molly Ball believes that the sequester was a good idea.

    by virtue of its very do-nothingness, the do-nothing Congress got a big thing done. First, in the fiscal-cliff deal struck around the new year, wealthy Americans’ income-tax rates went up, a policy change long sought by the president and his party. Then, in March, the budget ax known as sequestration fell, chopping $1 trillion from federal spending over the next decade—a cherished goal for fiscal conservatives. [And] More revenue plus less spending equals a lower deficit.

    We cut the federal budget deficit and that’s great. But the across-the-board spending cut in the sequester was mindless. And mindlessness is the rule, not the exception. Peter Orszag (former director of the Congressional Budget Office under Barack Obama) and John Bridgeland (director of the White House Domestic Policy Council under President George W. Bush) wrote that

    Based on our rough calculations, less than $1 out of every $100 of government spending is backed by even the most basic evidence that the money is being spent wisely.

    The sequester is Exhibit A among stupid spending decisions. In the coming weeks, can we perhaps think about how to reduce government spending?


    • I completely agree. However, it appears Cutler did not actually compare the scenario with NIH grants to the scenario without them. There are selection effects going on where NIH ends up funding the best projects while other sources get stuck funding less promising projects. So arguably, much of that high-value research would have happened anyway, and the actual unique returns to NIH funding is much lower. You could even weave a story, however unlikely, where the actual return on NIH is negative since tax-and-spend reduces GDP and therefore the available funds to invest in research, which would still be consistent with Cutler’s results.

      • Thanks for the thoughtful comment.

        You raise a good point. We can’t run an experiment to tease out the effect of funding the NIH.

        Having said that, I would dispute your assumption that someone else will pick up the best projects. I don’t think there is any foundation that can carry out the quality of peer review provided by the NIH. So my guess is that diminishing the NIH role _lowers_ the quality of the science that gets funded.

        It’s not clear to me that reducing NIH funding would increase measured GDP (and I gather you are not committed to that view). But let’s assume that is true. The small increase in GDP would have a quite small positive effect on health. But remember that Cutler’s argument is that health care has a large unmeasured effect on life expectancy which would massively increase GDP if it were included in measured GDP. So even if you are right about the effect of “tax and spend”, Cutler would likely argue that the effect on overall well-being would be negative.

    • Very good point but unfortunately, our government is hostage to crazies who don’t think rationally, just ideologically.
      As far as reducing government spending… is that even the right question? As you point out, we should be looking to see if the money we spend is spent wisely. What is the “return” on that money. It may be (as some have argued) that we should have more spending on social welfare and to improve the economy. One could make a good case that our large “defense” and “security” spending is largely wasted and in some cases counterproductive.
      However, I really don’t think we are going to have any rational “fact based” discussion of these issues in the next few months… it will all be Sturm und Drang.

    • Pretty sure that Orzag was OMB Chief not CBO chief.

      • Thanks for catching the error. Orszag was director of OMB under Obama. He was director of the CBO during the last two years of GW Bush.