978-0133780581 Chapter 21 Lecture Note

Unlock access to all the studying documents.

View Full Document

Chapter 21. Should Policymakers

be Restrained?

I. MOTIVATING QUESTION

Should external limits be imposed on policymakers?

The fact that the effects of policy are uncertain does not imply that limits should be imposed on

policymakers. Policymakers understand that uncertainty provides an argument for moderation in setting

policy and to the extent they are benevolent, impose their own restraint. On the other hand, the strategic

interactions between the private sector and the government and among political parties suggest that

economic performance may improve when constraints are placed on discretionary policy. Even so, the

need for discretionary policy in times of economic distress calls for care in the design of institutional

limits on policymakers.

II. WHY THE ANSWER MATTERS

Modern macroeconomics was founded on the premise that governments could take action to improve

economic performance. Toward this end, the previous 20 chapters have discussed policies available to

governments and the likely effects of such policies. Although caveats have been offered along the way,

the basic message is that governments can take action to reduce economic fluctuations and increase the

long-run level of the capital stock, but have little ability to influence growth, beyond establishing

institutions that reward innovation. This chapter examines the limits on discretionary policy. It asks

whether governments can be expected to exercise their discretion wisely and whether institutions can be

designed to encourage good policymaking.

III. KEY TOOLS, CONCEPTS, AND ASSUMPTIONS

1. Tools and Concepts

i. In economics, a game refers to strategic interactions among a set of players.

ii. A government faces a time inconsistency problem when it has an incentive to deviate from a

promised policy once private agents have made decisions based on the policy.

iii. A political business cycle occurs when policymakers try to generate expansions before elections in

hopes of securing reelection.

21-98

IV. SUMMARY OF THE MATERIAL

Despite the apparent beneficial role that macroeconomic policy can play, arguments for restraints on

policymakers are common. These arguments fall into two classes. One view is that policymakers, in

trying to do well, do more harm than good. The other view is that policymakers do what is best for them,

rather than for society as a whole.

1. Uncertainty and Policy

One reason that policymakers may do harm is that the effects of policy are uncertain. In the mid-1980s,

an exercise conducted at the Brookings Institution used twelve prominent world macroeconomic models

to predict the effects of a specified monetary expansion. There was substantial variation in the

quantitative results. Since such models (and their descendents) capture existing quantitative knowledge

about the effects of policy on the economy, the implication is that policymakers face substantial

uncertainty about such effects.

If extreme macroeconomic outcomes are very harmful, uncertainty provides an argument for moderation

in policymaking: the smaller policy changes result in a narrower range of possible outcomes. This

argument is less relevant when the economy is suffering a severe recession or hyperinflation. In this case,

the effects of relevant compensating policies—no matter how extreme—are likely to improve economic

performance. There is every reason to believe that policymakers understand that uncertainty calls for

restraint. Thus, the existence of uncertainty does not necessarily provide justification for imposing

external constraints on policymakers.

2. Expectations and Policy

The relationship between the government and private sector can be described as a game, i.e., strategic

interactions among a set of players. Private firms and households make decisions based on expectations

of future policy. The government forms policy based on the expected response of the private sector.

Viewing the economy in this perspective can provide a rationale for external constraints on policymakers.

Sometimes a government has an incentive to deviate from a promised policy once private agents have

made decisions based on the policy. In this case, the government’s optimal policy is said to be time

inconsistent. For example, as discussed in Chapter 8, central banks can temporarily reduce

unemployment below the natural rate by generating unexpected inflation. At the same time, if inflation is

costly, central banks have an incentive to announce a policy of low money growth to reduce expected

(and hence actual) inflation. Combining these objectives, the central bank’s optimal policy is to announce

a tight monetary policy to convince the private sector that inflation will be low, but to implement a more

expansionary policy to reduce unemployment (temporarily). If the central bank attempts to carry out this

program, however, it will lose credibility quickly, and the private sector will set wages and prices in

expectation of high inflation. The loss of credibility will contribute to high inflation and eventually

eliminate the ability of the central bank to reduce unemployment below the natural rate. In the long run,

the central bank’s program will have no effect on unemployment (it will return to the natural rate) but will

result in high inflation. Under these circumstances, economic performance would improve if the central

bank could commit itself credibly to maintain low money growth. In this case, the public’s expectations

21-99

would be consistent with low inflation, and the natural rate could be achieved with a lower inflation rate.

Thus, time inconsistency provides an argument for external restraints on government actions.

External restraints are needed because self-restraint may not be credible. But care should be taken to

preserve flexibility in times of economic distress. For example, one way to impose an external constraint

on the central bank is to legislate a money growth rule. Such a restraint prevents the central bank from

cheating on policy targets (a desirable outcome), but also eliminates its ability to respond to adverse

shocks (an undesirable outcome). A superior alternative is to make the central bank politically

independent of the government in power and to appoint a central banker with a known distaste for

inflation. The evidence suggests that central bank independence is associated with lower inflation.

3. Politics and Policy

Sections 21.1 and 21.2 assumed that policymakers were benevolent. In fact, politicians may act to

maximize their own reelection prospects. If voters are shortsighted, politicians have an incentive to

implement policies that generate short-run benefits, regardless of the long-run costs of these policies.

There is not much evidence in favor of this proposition in the United States. Until the 1980s and aside

from the Great Depression, the ratio of government debt to GDP tended to increase only during wars.

Thus, the evolution of deficits and debts seemed to track economic circumstances rather than the

shortsightedness of voters. The bouts of increasing deficits and debt in the 1980s and the early years of

this decade appear to have more to do with games among policymakers, as described below, than with the

short-sightedness of voters. Moreover, if voters were shortsighted, politicians could improve their

changes of reelection without much cost by generating expansions just before elections. Thus, there

would be a political business cycle, with growth highest in the final years of presidential administrations.

In the postwar period, U.S. growth has been highest in the final years of presidential administrations, but

the difference across years has been relatively small on average.

Another source of harmful policies emerges out of strategic games among policymakers. Substantial

policy disagreements occasionally result in wars of attrition between political parties that result in the

postponement of needed policies, such as deficit reduction. For example, the tax cuts of 1980s and the

early years of this decade seem to have been motivated in part by a desire to cut spending. In both

episodes, the tax cuts led to substantial increases in the deficit, and ultimately to concern about reducing

the deficit. In both cases, Republicans typically argued for spending cuts (in nondefense programs) and

Democrats for increased revenues. The deficits of the 1980s were eventually eliminated with spending

restraint, but the surplus of the last years of the Clinton administration was quickly reversed. The U.S.

budget has moved into a large deficit, which is projected to continue for some time. Republicans and

Democrats now seem poised to engage in war of attrition over deficit reduction.

Some economists believe that the only way to break the current impasse is with some sort of legislative or

constitutional limit on fiscal policy. A problem with this approach is that it may limit needed fiscal

flexibility. For example, when the economy is in recession, policymakers should have the option of fiscal

stimulus. The problem is to impose an effective limit on fiscal policy, while preserving enough flexibility

to respond to changing economic circumstances. This is a difficult task. A balanced budget amendment

to the Constitution, which has been advocated by some, would need an escape clause for recession or

other emergencies. How to define such an escape clause without eviscerating the fiscal discipline remains

a challenging problem.

21-100

V. PEDAGOGY

The discussion of time inconsistency provides an opportunity to revisit the issue of fixed exchange rates

and exchange rate crises. A credible commitment to a fixed exchange rate eliminates a government’s

ability to conduct discretionary monetary expansion. Thus, it has been argued, a credibly fixed exchange

rate will reduce inflation expectations and inflation. On the other hand, as numerous currency crises

indicate, it is difficult for governments to convince markets that they will not devalue or adopt a flexible

rate in times of distress. Thus, there is an argument that some governments would do well to completely

tie their hands with respect to monetary policy by establishing a currency board or adopting the U.S.

dollar as their currency. Argentina’s experience, however, demonstrates that even currency boards do not

eliminate expectations of devaluation.

VI. EXTENSIONS

For students who know rudimentary calculus, instructors can present a simple formalization1 of the time

inconsistency problem described in this chapter. Suppose the policymakers’ loss function can be written

L=(a/2)π2–b(π-πe).

In words, policymakers care about inflation and unemployment in excess of the natural rate. The first

term of the loss function reflects concern about inflation. The second term reflects concern about excess

unemployment, which is determined by the difference between actual and expected inflation.

If policymakers announce an intention to achieve zero inflation, and this is believed (πe = 0), they have

two options. They can follow through on the announcement and produce zero inflation, in which case, π

= 0, u = un, and L = 0. Alternatively, they can act in a discretionary fashion, renege on their commitment,

and attempt to achieve the socially optimal rate of inflation, conditional on πe=0. Minimizing L, subject

to πe = 0, generates an optimal rate of inflation of b/a and a loss of (–b2/2a), which is smaller than the loss

achieved by following the announced rule. The gain comes from the benefit of surprise inflation in

lowering unemployment below the natural rate. Note that the form of the loss function implies that the

cost of inflation is zero (at the margin) when inflation is zero, so there is always some gain to surprise

inflation.

The problem is that the private sector is aware of these incentives. Therefore, in the absence of some

credible mechanism that forces policymakers to adhere to their announced policy, the private sector will

never believe the announcement. Expecting that policymakers will act in a discretionary fashion, the

private sector will set πe = b/a. Because this constant value of πe does not alter the solution of the

policymaker’s optimization problem, the actual rate of inflation will indeed turn out to be π = b/a. But

since this inflation rate is no longer a surprise, there are no unemployment gains and the loss becomes L

= (b2/2a), a worse outcome than would have been achieved by following the rule. Since policymakers are

assumed to be benevolent—i.e., their loss function accurately reflects social preferences—society would

be better off by removing their discretion if possible.

1 The formalization is based on Barro, Robert J. and David B. Gordon (1983), “Rules, Discretion, and

Reputation in a Model of Monetary Policy,” Journal of Monetary Economics, 12:101-121.

21-101

21-102