All Commentary
Tuesday, January 24, 2012

Two Kinds of Government Failure

Incentives and knowledge.

We use the term “government failure” to refer to interventions that systematically make things worse than they were before; where the cure is somehow worse than the disease.  To those of us who study interventionism, however, there are actually two meanings of government failure.  Two quotations illustrate my point.

The first is from the textbook I’ve used for many years in my introductory microeconomics class by Heyne, Boettke, and Prychitko:

(1) …the demonstration by modern political economists of the tendency within democratic governments to concentrate benefits and disperse costs within policy-making has been one of the most important contributions to our intellectual understanding of why good politics is not necessarily good economics (331).

Political Manipulation

For example, it would be grossly inefficient to implement a government project that costs $3 million but only provides a total benefit of $1 million.  But if through political manipulation you could make everyone in a community of one million pay an equal share of the cost and at the same time hand over the benefit of the project to say 1,000 people, it may well be enacted.  That’s because the no one would pay more than $3, while beneficiaries would be getting $1,000 each.  The gainers would fight a lot harder to pass the project than the losers would to fight it.

Now, this from Friedrich Hayek’s The Road to Serfdom:

(2) Against the innumerable interests which could show that particular measures would confer immediate and obvious benefits on some, while the harm they caused was much more indirect and difficult to see, nothing short of some hard-and-fast rule would have been effective (71).

Hayek is talking here about why sometimes a dogmatic resistance to all government intervention has been helpful in preserving liberty.  But I would like to point out that the kinds of measures he has in mind are things such as minimum-wage laws that appear to benefit workers at the bottom of the wage scale but, as is well established in the economics literature, actually price lower-skilled workers out of the market over time.  Currently in the state of New York it is illegal for anyone to work in most jobs for less than $7.25 an hour, which means if your labor is worth less than that to employers, as is likely the case if you are young, you will not be hired.

At first glance these two statements appear to be saying more or less the same thing.  Both identify a source of systematic government failure: concentrated benefits versus dispersed costs; or immediate/obvious benefits versus harm that is indirect/difficult to see.  Both identify interventions that benefit only some while the consequences fall elsewhere or elsewhen.  And both imply that the outcome of these interventions will tend to be undesirable to someone.


The Heyne et al. statement (1) is in terms of “benefits” and “costs” while the Hayek (2) talks about benefits and harm.  Perhaps you could interpret “harm” as the “cost” that is dispersed in the first statement, although “harm” has a more general meaning that may not be easily quantifiable, such as undermining the norm of preserving the free market.

Furthermore, (1) focuses on concentrated benefits, while (2) on the immediate and obvious benefits.  Beneficiaries of the government project get $1,000 while only some under the minimum-wage benefit at all.  Now beneficiaries in both cases can see immediately what they are getting, but there is a certain sense in which those proposing the minimum-wage law (mistakenly) believe it will benefit everyone, not just a few, while people in the concentrated benefits approach are fully aware that only a few will indeed benefit from the intervention.

The costs in (1) are dispersed while the harm in (2) is indirect and difficult to see.  The $3-per-person cost of the government project is certainly not indirect or difficult to see; rather everyone, both gainers and losers, do see it but it’s too small an amount for most to bother with.  Few have an incentive to do anything about it.  However, the harm of the minimum-wage law is indeed indirect and difficult to see.  Acquiring that knowledge is hard.

Time is not necessarily an element in (1) but it is in (2), in fact it’s because some time has to pass before you can see the full consequences of an intervention that they are often difficult to see.

Inefficiency and Error

Finally, the nature of government failure in (1) is based on inefficiency: the total costs exceed the total benefits.  Government failure in (2), on the other hand, is based on error.  (1) implies that inefficiency is systematic because the political rules of the game make people follow perverse incentives, while (2) implies that error is systematic because of a failure to learn.  And (1) suggests that people can manipulate the system with full knowledge of all the consequences so that they can effectively concentrate benefits on themselves or their cronies and disperse the costs among taxpayers; while (2) suggests that the harm that a policy causes is largely unintentional and the result of ignorance.

The first approach emphasizes incentive problems, the second knowledge problems.  Neither is necessarily better than the other, but there are occasions when one or the other might be more appropriate.  Often both are applicable to the same intervention.  For example, some political economists argue that labor unions say they favor raising the minimum wage to express solidarity with their non-union brothers and sisters but really support it because it insulates its members a little more from unskilled workers.  One author has even characterized the approach to political economy that emphasizes “concentrated benefits, dispersed costs” as “a proposition about inferring intentions from outcomes” (47).  I would not go that far.

(If you’d like to read a lengthier treatment of this topic check here.)

  • Sanford Ikeda is a Professor and the Coordinator of the Economics Program at Purchase College of the State University of New York and a Visiting Scholar and Research Associate at New York University. He is a member of the FEE Faculty Network.