Productive Enough for Government Work

It's difficult to measure the efficiency of federal agencies and programs.

Let's put aside the productivity of Congress.

Photographer: Joe Raedle/Getty Images

Statistics for the U.S. economy typically don’t pick up the productivity -- or lack thereof -- of government itself. The services government provides are not sold on markets, and so can’t be valued by standard methods. Nonetheless, government is a big part of any modern economy, with federal spending in the neighborhood of 20 percent of gross domestic product. So its productivity is a major issue.

Some ways of looking at government make it seem very productive indeed. In my capacity as a professor of economics, I’ve given talks at, visited or placed students at the Congressional Budget Office, the Federal Reserve, the Bureau of Labor Statistics, the Census Bureau, the National Institutes of Health and the National Science Foundation, among others. I have found each place to be a remarkable assemblage of talent, and the staff have strong work ethics, even when many of them could earn much more money in the private sector. It would seem that individual productivity within these institutions is high, or at the very least not abysmally low.

Another way of thinking about government productivity is to look at the biggest programs, namely Social Security, Medicare and Medicaid. They send large amounts of money across the country at very low administrative costs. For instance, Medicare expenditures exceed $500 billion a year, with administrative costs estimated at about 2 percent.

That too, on the surface, might appear to be an example of highly efficient government, but it also shows why the productivity of government is so difficult to measure. The main costs of Medicare are not on the books of the government, but rather are the opportunity costs of diverting resources from the private sector. Medicare has shaped how much Americans pay for treatment, pushed many doctors into serving the elderly rather than the young, and pulled more workers into the health-care sector and away from other endeavors. Moreover, the taxes used to fund Medicare discourage labor supply. However we might assess this bargain, the administrative burden of Medicare doesn’t come close to reflecting the full costs of the program.

It’s easy to run a cost-benefit study showing that Medicare is better than doing nothing at all, but much harder to show that Medicare is superior to an alternative -- such as the greater use of competitive bidding, better monitoring of spending and tougher restrictions on fraud, more means-testing, less emphasis on end-of-life care, or changing the system so older Americans who can afford it continue to pay some fraction of their treatment costs instead of covering the "donut hole" through private insurance.

A worrying aspect of Medicare is how popular the program is with its beneficiaries. According to one calculation, the average man on Medicare receives $119,999 more in benefits from the system than he puts in through taxes, while the average woman receives $146,000 in net benefit. Any program run on that basis is going to be loved, and those numbers raise the question of whether Medicare reform really will be possible when it proves necessary, as inevitably it must.

We’re now getting closer to the most significant source of inefficiency in government, namely how policy interacts with what the electorate wants. If you re-examine the list of institutions in the second paragraph, it is noteworthy how many of them are buried relatively deep in the bureaucracy and are not hot button issues with most voters (the Fed’s interest rate policy is a partial exception). When it comes to running the census, for instance, it is necessary not to unduly alienate voters, but voters are not clamoring for “census benefits.” And so the program can proceed relatively efficiently.

What then are the biggest and most publicly visible programs most efficient at? I would suggest they do the very best job at making their benefits visible and their costs relatively nontransparent and hard to identify. Now, if you visited a corporation and found the company had that very same specialty, what would you likely infer about the path for the productivity of that company? What if there were no standard measures available for measuring the productivity of that company and no possibility that competition could knock it out of the marketplace?

I know what I would think, yet many people are unwilling to apply that logic when assessing the productivity of the U.S. federal government.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Tyler Cowen at

    To contact the editor responsible for this story:
    Stacey Shick at

    Before it's here, it's on the Bloomberg Terminal.