Long-Term Objectives for Government Debt

Long-Term Objectives for Government Debt

Alan J. AuerbachUniversity of California, Berkeley

January 29, 2008

Outline

I. What are the underlying objectives that motivate concern about the public debt?

II. Why simple deficit targets and budget rules are inadequate

III. Setting long-run targets for fiscal policy

I. Underlying Objectives

1. Intergenerational equity• Debt that is too large shifts too much of a burden

onto future generations• But not clear what “too much” is when future

generations will be better off• Not a question for economists alone to answer

I. Underlying Objectives

2. Economic performance• Debt accumulation may “crowd out” domestic

capital formation or trade surpluses• The need to service the debt may force

substantial increases in future marginal tax rates, leading to increased distortions of economic decisions

I. Underlying Objectives

3. Fiscal sustainability• If we rule out a “bubble” in which the

government can forever service debt simply by issuing more debt, higher debt requires higher future taxes and/or lower future spending

• But existing policy trajectory may be quite inconsistent with this requirement

• If policy is not sustainable, sudden involuntary adjustments may follow, causing greater economic disruption than planned policy changes

Weighing the Objectives

• Objectives point to similar, but not identical policy paths– For example, should tax rates be rising (equity) or

constant (efficiency)?• Relative importance depends on empirical

aspects of economic behavior– If bequest motives are strong, then

intergenerational equity may be less important– If responses to tax rates are small, then keeping tax

rates smooth and low may be less important

Key Point

• It makes no sense to establish a target deficit path without first specifying underlying objectives

Targets or Restrictions?

• Is it enough to determine targets, or are rules needed to enforce them?

• Answer depends on the alignment of short-term government incentives with long-term objectives– If alignment strong, targets provide more

flexibility for short-term action– If alignment weak, then simple and less flexible

rules may work better

II. Inadequacy of Simple Targets

• Several factors weaken and make more ambiguous the relationship between debt path and underlying objectives

Government Assets

• Simple answer: subtract from government debt• But a lack of symmetry because assets may not

provide a revenue stream, so don’t offset need for tax increases

• Also, what is an asset? Many expenditures (e.g., health, education, etc.) provide future benefits

Implicit Liabilities

• Isn’t the obligation to pay future health and retirement benefits similar to a debt obligation?– Intergenerational redistribution ()– Need for future tax increases ()– Strength of obligation (?)

• May not present the same problem of sustainability if commitment weak, but otherwise like debt

Deferred Tax Assets

• Tax systems that are otherwise equivalent differ in their timing of tax collections, and this can give rise to “deferred tax assets”

• Example: to encourage saving, can exempt earnings on saving or defer initial tax on amount saved and tax ultimate withdrawal– the two approaches have same economic effects

but differences in timing of tax payments– deferral approach appears to collect fewer taxes,

but not if we take account of deferred tax assets

A Difficulty

• Both implicit liabilities and deferred tax assets (or liabilities) must be measured relative to some benchmark tax and transfer system

Variations Along Desired Path

• Targets may vary from year to year because of variations in the social value of deficits– short-term stabilization policy– other reasons why surges in deficits may be

valuable• Wars are an obvious illustration, but also

higher share of the population “in need”– Could justify higher spending on transfers as

population ages, but only for temporary surges

Deficits versus Spending

• Given deficit targets, should there also be spending targets?

• Logic:– Spending exhaustive, tax cuts aren’t– Controlling spending limits government power

• Big logical problem: what distinguishes direct expenditures from “tax expenditures”?

• So, spending targets can make sense only to the extent that spending is well-defined

Deficits and Generational Distribution

• Regardless of the adjustments that one makes, policies with the same deficit paths can be associated with differences in how burden is shared among generations

• For example, accounting for implicit liabilities may provide a more accurate measure of burden on future generations, but won’t tell us to whom these liabilities are owed or how policy changes influence this

Table 1. The Impact of Changes in a Public Pension System

Impact on: Experiment Current

Deficit S-Term Deficit

Implicit Liabs.

Older Gens.

Younger Gens.

Future Gens.

1. Increase taxes & bens. by same PV, in five years

0 0 0 0 0

2. Reduce bens following same pattern as tax increases in #1

0

3. Reduce future bens. by the same PV as in #2, in 20 years

0 0 0

Table 1. The Impact of Changes in a Public Pension System

Impact on: Experiment Current

Deficit S-Term Deficit

Implicit Liabs.

Older Gens.

Younger Gens.

Future Gens.

1. Increase taxes & bens. by same PV, in five years

0 0 0 0 0

2. Reduce bens following same pattern as tax increases in #1

0

3. Reduce future bens. by the same PV as in #2, in 20 years

0 0 0

Table 1. The Impact of Changes in a Public Pension System

Impact on: Experiment Current

Deficit S-Term Deficit

Implicit Liabs.

Older Gens.

Younger Gens.

Future Gens.

1. Increase taxes & bens. by same PV, in five years

0 0 0 0 0

2. Reduce bens following same pattern as tax increases in #1

0

3. Reduce future bens. by the same PV as in #2, in 20 years

0 0 0

Table 1. The Impact of Changes in a Public Pension System

Impact on: Experiment Current

Deficit S-Term Deficit

Implicit Liabs.

Older Gens.

Younger Gens.

Future Gens.

1. Increase taxes & bens. by same PV, in five years

0 0 0 0 0

2. Reduce bens following same pattern as tax increases in #1

0

3. Reduce future bens. by the same PV as in #2, in 20 years

0 0 0

III. Setting Long-Run Targets

• Start with simple assumptions and see how complications change prescription

A Path for Deficits and Spending

• Suppose– revenue and spending well-defined– composition of revenue and spending fixed, so that

only annual levels of each to be determined• Then each path for deficits and spending will

correspond to a unique outcome, and we can choose from among fiscally stable policies the one that best balances equity and efficiency objectives

A Path for Deficits and Spending

• This path would exhibit variation over time to reflect time variations in the value of resources– for example, we might see large deficits

accumulate to provide benefits for large elderly cohorts, with surpluses both before and after to help spread the burden and tax distortions

– this is one interpretation of current spending forecasts, but current tax policy would need to respond to generate large short-run surpluses

Notes

• Uncertainty about the need for future spending generally points to higher surpluses now

• High current surpluses, even if consistent with policy objectives, present additional problems– political: can the surplus be sustained?– economic: how should the surplus be invested?

• Usefulness of spending targets may be limited to certain categories

Adjusting Deficit Targets

• But composition of spending and revenues changes, so given path of spending and deficits can correspond to many patterns of burdens and distortions

• Makes adjustments for government assets, implicit liabilities, and deferred tax assets useful– adjust taxes and spending to neutralize differences in the

timing of burdens and tax distortions– this gives debt a more consistent meaning as the

composition of revenues and spending changes– also, may make deficit targets look more “normal” when

there are demographic fluctuations

Adjusting Deficit Targets

• With adjustments comes subjectivity– when is a deferred liability more a commitment

than a plan?– what is “current policy” for the future if that policy

is not explicit?• Simple procedures for how to draw lines may

be difficult to define ex ante

Further Measures

• Even adjusted deficits, as explained before, do not provide all available information about distortions and distribution of burdens

• Thus, augment using “generational accounts,” which allocate fiscal burdens across cohorts– more information and calculation required, so may

not be feasible for very frequent calculations– but little argument against using in conjunction

with other measures

Further Measures

• In principle, only way to determine optimal policy path is within the context of a general equilibrium model of the economy, which traces through all interactions of policy and behavior of households and firms

• This will not be feasible without considerable simplifying assumptions, but even with such simplifications can provide a useful check that chosen targets make sense given underlying objectives

Conclusions

• Deficit targets are not an ultimate objective, but need to be derived from a weighing of underlying economic and social objectives

• Rules may need to complement or supplant targets if short-run government incentives diverge from longer-run goals

• Targets based on simple measures of debt are inadequate to deal with changing composition of spending and revenues

Conclusions

• Adjusted debt measures make more sense but adjustments involve subjectivity that must be confronted

• More detailed measures, notably generational accounts, provide additional information useful to evaluate policy

• A full general-equilibrium evaluation to determine optimal policy path is not feasible without many simplifying assumptions, but is still useful