There has emerged substantial
criticism
of the defense
budget recently proposed by the Biden administration — the critics include
several of my AEI
colleagues — as insufficient on an inflation-adjusted basis in the context
of the Ukraine crisis and the defense of US interests around the globe.

That view of whether
the proposed budget is sufficient to carry out the National
Defense Strategy may be correct. (It is reasonable to observe, however,
that most such discussions fail to distinguish clearly among US interests
vital, important, desirable, and marginal.) Correct or not, the critics assume
that the budget is an unbiased estimate of the cost of defending whatever set
of US interests are deemed sufficiently high as priorities.

That is not correct.
Begin with the observation that national defense to a significant degree is an
example of what economists classify as “collective” (or “public”) goods:
Provision of a unit of a collective good (e.g. protection from foreign threats)
to one citizen does not reduce the amount of that good available to others
(although different citizens may place different values on the collective good).
Another central characteristic is the difficulty of excluding nonpayers from
receiving the benefits of the collective good; people who fail to pay their
taxes cannot be excluded from the benefits of defense services.

Because nonpayers
cannot be excluded, the standard textbook analysis predicts that the private
sector will produce too few units of collective good because there exists a
powerful incentive to obtain a free ride on the defense benefits funded by
others. (Everyone prefers a trash-free highway, but very few stop voluntarily along
the shoulder to pick up used fast food containers.) And so the standard
textbook analysis concludes with the proposition that government provide the
collective good because it can enforce payment from those who otherwise would
attempt to obtain the free ride, yielding an “efficient” output level for the
collective good.

That standard
textbook argument is not correct. Consider a democracy with a simple majority
decision rule, and suppose that the efficient defense budget — the budget that
equates marginal defense benefits and costs — is D*. Suppose that the
government produces only defense services (a collective good) and transfer
payments (T, a private good).

Because budgets are
determined by a simple majority vote, the median voter — the voter with budget
preferences precisely in the middle of the range of all voter preferences — is
the marginal member of the majority. For each dollar per voter cut from the
defense budget, two dollars of transfer payments can be given to each member of
the majority coalition. (In a large polity (political community), the majority
is 50 percent of the electorate plus one.)

So the median voter
decides the amount of spending on defense and on transfers. That voter — the
majority — has an incentive to reduce spending on defense below D* and to
transfer those resources to the majority until it views $1 per voter of defense
as equal in value to $2 of transfers per member of the majority.
Notwithstanding the standard textbook analysis, democracies too have
powerful incentives to produce too few defense and other collective goods.

This prediction is
robust in the public
choice
literature, even in analyses (models of democratic decision making)
far more complex than that just described. One obvious way to reach a budget
equilibrium in which provision of the collective good is efficient is to offer
more “pork” — wealth transfers — to various interest groups so as to forge a
majority coalition supporting both the efficient defense budget and an increase
in transfer payments. Too some significant degree, the US political system does
just that even within the defense budget itself: Military bases and other
facilities are spread out geographically across more congressional districts
than a narrow “efficiency” criterion might suggest, although the Base
Realignment and Closure program eroded that effect. Major acquisition
programs always spread the attendant contracts and subcontracts across as many
congressional districts and industries as possible.

The point here is that the defense budget itself is an underestimate of the cost of delivering the optimal amount of national security services. The true cost includes the costs of the “efficient pork” needed to forge a coalition in support of the efficient defense budget. It very well may be the case that the national security services funded in that efficient defense budget are sufficiently valuable that the “true cost” is worth it. But defense analysts and interested observers should bear in mind that the true cost is greater than the notional defense budget, and perhaps by an amount that is significant. Those favoring an increase in the defense budget combined with a reduction in domestic spending are very likely to be disappointed.

The post A note on defense spending and budget pork appeared first on American Enterprise Institute – AEI.