Download Chapter 12: Government Debt and Budget

Chapter 12: Government Debt and
Budget Balance
J. Bradford DeLong
--First Draft-1999-07-27: 4,493 words
Measuring the Debt, Measuring the Budget
Before we discuss the deficit and debt further, we need to spend a little bit of
time on measurement issues. Just what, exactly, is the budget balance? Or, rather,
which of the many possible budget balances—totals that produce one final
surplus or deficit number—that are calculated should we pay the most attention
to?
The Cash Balance
The budget bottom line that is reported in the newspapers the most often is the
so-called “unified cash” balance: the difference between the money that the
government actually spends in a year, and the money that it takes in. This
balance is called “unified” because it unifies all of the government’s accounts and
trust funds (including Social Security). This balance is called “cash” because it
doesn’t take account of changes in the value of government-owned assets or of
the future liabilities owed by the government: it is just cash in minus cash out.
(Unfortunately, the “cash” balance does make a few allowances for future
liabilities: when the government guarantees a loan, for example, the reserve set
aside in case the borrower defaults and the government actually has to make
good on its guarantee is for complicated political reasons counted as a “cash
budget” expenditure, even though it isn’t.)
It is not even completely clear what the economically-relevant budget balance is.
There are a number of adjustments that can be and frequently are made to the
budget balance to try to produce a more relevant and useful number than the
balance reported for the unified cash budget.
Adjustments
Cyclical Adjustment
The first adjustment often made to the budget balance is to take out the effects of
the fiscal automatic stabilizers. A component of the budget deficit (or surplus) is
due to the automatic reaction of the deficit to the stage of the business cycle.
When unemployment is high, taxes are low and social welfare spending high, so
the budget balance tends to swing into deficit. When unemployment is low, taxes
are high—and the budget balance tends to swing into surplus.
The effect of these automatic stabilizers on the budget balance makes it hard to
interpret the budget balance. Is such-and-such a shift in the bottom line the result
of a change in policy, or just the result of the phase of the business cycle? To allow
themselves and others to concentrate on changes in policy, the government
calculates a cyclically-adjusted or high-employment budget balance. It is what the
estimated budget balance would be if the current level of GDP were equal to
potential output.
Almost everyone who analyzes economic and budget policy prefers to make this
adjustment, and to work with cyclically-adjusted rather than raw cash budget
balances.
[Figure: Cyclically adjusted budget balance and actual cash balance, 1960present]
Inflation
A second adjustment that almost all economists believe is warranted corrects the
measured budget balance for inflation. A portion of the debt interest paid out by
the government to its bondholders is merely compensation to them for the fact
that inflation is eroding the value of their principal. At the end of the year, the
debt principal plus this inflation component of debt interest are together equal—
in their power to purchase useful goods and services—to what the debt principal
was at the start of the year.
A good measure of the deficit should be a measure of whether the government is
spending more in the way of resources than it is taking in: a measure of the
change in the real debt that the government owes.
Almost everyone who analyzes economic and budget policy prefers to work with
inflation-adjusted budget balance numbers.
[Figure: the Inflation-adjusted budget balance]
Government Investment
Yet a third adjustment corrects for an asymmetry between the treatment of
private and public assets. Private spending on long-lived capital goods is called
“investment.” A business that has total sales of $100 million, costs of goods sold
of $90 million, and spends $20 million on enlarging its capital stock reports a
profit of $10 million—not a deficit of $10 million. Standard and sensible
accounting treatment of long-lived valuable assets in the private sector is
definitely not to count their entire cost as a charge at the time of initial purpose,
but instead to spread the cost out—a process called “amortization”—over the
useful life of the asset.
Shouldn’t the government do its accounting the same way, like a business, and
amortize rather than expense its spending on long-lived assets?
There are periodic calls that the federal government budget should be reformed,
and that the federal government should do capital budgeting. But few people use
numbers based on capital budgeting.
The principal reason that capital budgeting is resisted is political. Which
government expenditures are capital expenditures? Aircraft carriers and nuclear
weapons? The interstate highway system? Improvements to trails in the national
parks? Headstart expenditures—money spent on educating poor children (after
all, it is an investment in their future)?
It is hard to see any long-run dividing line between government investment and
government consumption expenditures that would be sustainable from a
political point of view. Thus critics regard capital budgeting as simply too
difficult to implement in a helpful way.
Supporters, however, point out that not doing capital budgeting at all is, in a
sense, worse than even the least helpful implementation.
Liabilities and Generational Accounting
All of the issues surrounding capital budgeting appear again whenever the longrun future of the government’s budget is considered. Back when I worked at the
Treasury Department, some $10,000 a year was set aside for me in my Treasury
pension account. It is as if my income had been $10,000 a year higher, and I had
invested that extra $10,000 in U.S. government bonds. But bonds issued by the
government appear on the books as part of the government’s debt. But pension
fund liabilities that the government owes to ex-workers do not.
Thus there is a sense in which the right way to count the government’s debt is to
look not just at the bonds that it has issued but at all of the promises to pay
money out in the future that it has made. Indeed, a large chunk of the
government’s expenditures—those by the Medicare and Social Security Trust
Funds, for example—are presented to the public in just this way. The Social
Security deficit reported by the Trustees of the Social Security System every
spring is not the difference between social security taxes paid in and social
security benefits paid out, but is instead the long-run, seventy-five-year balance
between the estimated value of the commitments to pay benefits that the Social
Security System has made and will make, and the estimated value of the taxes
that will be paid into the Social Security Trust Fund.
But the Social Security Trustees’ Report covers just one program—albeit a big
program. And great confusion is created by the fact that the Social Security
systems expenditures and revenues are also included within the unified budget
balance. Wouldn’t it be better to bring all of taxation and spending within a longrun system like that currently used by Social Security?
Economists like Laurence Kotlikoff and Alan Auerbach say “yes!” They
propose—instead of the year-by-year budget balances—that the United States
governmetn shift to a system of “generational accounting” that would examine
the lifetime impact of taxes and spending programs on individuals born in
specific years, and that would come up with a final balance that could be used
for long-term planning.
It is hard to escape the conclusion that Auerbach and Kotlikoff have a very
strong case. Yet few analysts of the budget use their “generational accounting”
measures.
[Box: U.S. fiscal policy through the lens of alternative deficit
measures]
To be written…
Budget Balances and National Debts
Sustainability
The first question to ask about a government that is running a perennial deficit
is: “Can it go on?” Is it possible for the government to continue running its
current deficit indefinitely, or must policy change—possibly for the better,
possibly for the worse.
The thing to look at to decide whether the government’s fiscal policy is
sustainable is the time path of the ratio of the government’s total debt to GDP, the
debt-to-GDP ratio, which we will write in symbols as D/Y: “D” for debt, and “Y”
for GDP. We would like to know toward what value this debt-to-GDP ratio is
heading: what is the long-run steady-state value of D/Y given that the
government is running a constant deficit equal to d percent of GDP?
As in chapter 4, use the fact that if a ratio like D/Y at its steady state value, then
both the numerator and the denominator are growing at the same proportional
rate. We know from chapter 4 that the denominator—real GDP, Y—grows in the
long run at a proportional annual rate n + g, where n is the annual rate of laborforce growth and g is the annual rate at which the efficiency of labor is
augmented.
What is the proportional rate at which the real debt, D, is growing?
If the real debt this year is equal to Dt, then the real debt next year will be equal
to:
Dt+1 = (1-π)Dt + d x Yt
The real value of the debt shrinks by a proportional amount π as inflation erodes
away the real value of the debt principal owed by the government, and grows by
an amount equal to the deficit as a percent of GDP d times the current level of
GDP, Y.
Thus the proportional growth rate of the real debt is:
Y 
Dt 1  Dt
   d   t 
Dt
Dt 
And the debt-to-GDP ratio will be stable when these two proportional growth
rates—of real GDP and of the real debt—are equal to each other:
n + g = -π + d(Y/D)
Which happens when:
D
d

Y n g 
This is the level toward which the debt-to-GDP ratio is heading: this is the level
consistent with a constant deficit of d percent of GDP in an economy with long
run inflation rate π, and with long run real GDP growth rate n+g.
So why then do economists talk about deficit levels as being “unsustainable”? It
looks as though for any deficit as a share of GDP d, the debt-to-GDP ratio heads
for a well-defined steady-state value. But this is only half the story: the debt that
the government wants to issue goes to a stable level relative to GDP, but is this
an amount of debt that anyone wishes to hold? The higher the debt-to-GDP ratio,
the more risky and investment do financiers judge the debt of a country—and
the less willing they are to buy and hold that debt.
A higher debt-to-GDP ratio makes investments in the debt issued by a
government more risky for two reasons.
First, revolutions—or other, more peaceful changes of government—happen.
One of the first things that any new government must decide is whether it is
going to honor the debt issued by the previous governments. Are these debts the
commitments of the nation, which as an honorable entity honors its
commitments? Or are these debts the reckless mistakes made by and obligations
of a gang of thugs, unrepresentative of the nation, to whom investors should
have known better than to lend money for them to steal?
The higher the debt-to-GDP ratio, the greater the temptation for a new
government to repudiate debt issued by its predecessor. Hence the riskier is
buying and holding a portion of a country’s national debt.
Second, even if there is no change in government, it is still the case that a
government can control the real size of the debt it owes through controlling the
rate of inflation. The (nominal) interest rate to be paid on government debt is
fixed by the terms of the bond issued. The real interest rate paid on the debt is
equal to the nominal interest rate minus the rate of inflation—and the
government controls the rate of inflation.
Thus a government that seeks to redistribute wealth away from its bondholders
to its taxpayers can do so by increasing the rate of inflation: the more inflation,
the less is the government’s debt worth and the lower are the real taxes that have
to be imposed to pay off the interest and principal on the debt. Whether a
government is likely to increase the rate of inflation depends on the costs and
benefits—and raising the rate of inflation does have significant political costs. But
the higher the debt-to-GDP ratio, the greater the benefits to taxpayers of a
sudden burst of inflation. When the debt-to-GDP ratio is equal to 2, a sudden
10% rise in the price level reduces the real wealth of the government’s
creditors—and increases the real wealth of taxpayers—by an amount equal to
20% of a year’s GDP. By contrast, when the debt-to-GDP ratio is equal to 0.2 the
same rise in the price level redistributes wealth equal to only 2% of a year’s GDP.
Thus the government’s potential creditors must calculate that the greater the
debt-to-GDP ratio, the greater are the benefits to the government of inflation as a
way of writing down the value of its debt, the more likely is the government to
resort to such inflation, and so the more risky are investments in the
government’s debt.
Thus a deficit is sustainable only if the associated steady-state debt-to-GDP ratio is
low enough that investors judge the debt safe enough to be willing to hold it.
Think of each government as having a debt capacity—a maximum debt-to-GDP
ratio at which investors are willing to hold the debt issued at reasonable interest
rates. If this debt capacity is exceeded then the interest rates that the government
must pay on its debt spike upwards—and the government is faced with either a
much larger deficit than planned (as a result of higher interest costs), or with
resorting to inflation or hyperinflation to write the real value of the debt down.
Deficits that lead to steady-state debt-to-GDP ratios higher than the
government’s debt capacity are not sustainable—and as economist Herbert Stein
said, if something cannot go on forever then at some time it will stop.
[Box: Another long-run effect: unpleasant monetarist arithmetic and
the incentives for inflation]
To be written…
Effects of Deficits: The Conventional View
Even if a given deficit as a share of GDP is sustainable, it still may have significant
effects on the economy, both in the short run and in the long run.
In the last chapter we saw that discretionary fiscal policy cannot help moderate
the business cycle: laws passed to increase the government deficit and stimulate
the economy are likely to take effect long after the phase of the business cycle,
and thus the need for stimulus, is passed.
What, then, can discretionary fiscal policy do?
It can change the level of the government’s debt.
Whenever the government spends more than it collects in taxes (and in
seigniorage from printing money), it must borrow. It sells the public promises
that the government will repay the principal it borrows with interest: threemonth Treasury bills, ten-year Treasury notes, thirty-year Treasury bonds, and
other securities. And these accumulated promises to pay make up the
government debt.
[Figure: U.S. government debt since the Civil War]
Usually governments run up large debts during wartime. Their survival, and
perhaps the survival of their nation and their civilization, is at stake. So during
major wars governments use all the tools they have to gain control of resources
for their fleets, armies, and airforces. And one of those tools is a very healthy
dose of government borrowing. The great peaks in U.S. government debt as a
share of total domestic product all come after the three major wars in which the
U.S. has been engaged: the Civil War, World War I, and World War II.
Usually during peacetime the size of the government debt as a share of GDP
falls. Economic growth raises real GDP and inflation provides an additional
boost to nominal GDP. As long as the government’s tax and spending programs
are not grossly out of whack—as long as the overall budget is not in enormous
deficit—in peacetime government debt tends to fall as a share of GDP.
Thus the violation of this empirical regularity in the U.S. in the 1980s—the
sudden emergence of peacetime deficits large enough to cause a substantial
upward jump in the debt-to-GDP ratio—was a great surprise.
[Figure: U.S. government debt to GDP ratio, 1970-present]
In part because of higher spending on defense and other programs in the 1980s,
in part because of substantial tax cuts, the Reagan presidency set in motion a
series of deficits that ended by nearly doubling the burden of the federal
government debt as a share of GDP. The rise in the debt was brought to an end
by three factors:

President Bush’s economic advisors, who persuaded him to go back on his
campaign pledge of “read my lips, no new taxes” and to negotiate a serious
deficit-reduction program including major reforms in congressional budget
procedures in 1990.

President Clinton and his economic advisors, who made deficit reduction the
highest priority of his administration in 1993.

A healthy dose of good macroeconomic luck.
A fair but rough assignment of credit would give 40% to those who planned the
1990 deficit-reduction program, 30% to those who planned the 1993 deficitreduction program, and 30% to sheer dumb good luck. The end of the era of
deficits was, by the late 1990s, trumpeted as an amazing and important political
success, a policy accomplishment that would significnatly improve the lives of
Americans.
[Figure: Federal spending and tax collections, 1960-present]
But how big a burden is a large national debt? How important was the doubling
of the debt as a share of GDP that took place in the 1980s and early 1990s?
One view—the view held by the majority of economists—is that the emergence
of such large government deficits has expansionary effects on the economy in the
short run, and contractionary effects on the economy in the long run.
Short Run
In the short run, the income-expenditure diagram tells us that a deficit produced
by a tax cut stimulates consumer spending. A deficit produced by an increase in
government spending increases government purchases. Either way, it shifts the
IS curve out and to the right: any given interest rate is associated with a higher
equilibrium value of production and employment. If monetary policy is
unchanged—if the LM curve does not shift—then output and employment rise in
response to the tax cut. A deficit is expansionary in the short run.
Caveats
Of course, the belief that deficits are expansionary—increase production and
employment—in the short run hinges on the Federal Reserve’s maintaining
monetary policy unchanged in response to the rise in the deficit. If the Federal
Reserve does not want inflation to rise, it will respond to the rightward
expansionary shift in the IS curve by tightening monetary policy: shifting the LM
curve back and to the left, neutralizing the expansionary effect of the deficit.
As the discussion in the previous chapter suggested, the decision and policy
implementation cycle for monetary policy is significantly shorter than the
decision and policy implementation cycle for discretionary fiscal policy. There is
little doubt that the central bank can keep legislative actions to change the deficit
from affecting the level of production and unemployment. The question is
whether it will.
The answer is “very probably.” The central bank is presumably trying its best to
guide the economy along a narrow path without excess unemployment and
without accelerating inflation. It has made its best guess as to what level of
aggregate demand leads along that path. In all likelihood its senior officials are
uninterested in seeing the economy pushed away from that path by the fiscal
policy decisions of legislators.
Hence it is very likely that the central bank will act to neutralize any effect of
changes in fiscal policy on the level of output and employment.
[Figure: central bank neutralizing effects of fiscal policy on output and
employment]
[Box: Interest rate forecasts and the Clinton deficit reduction
program]
To be written…
Open-Economy Effects
Such an increase in the government’s budget deficit also leads to an increase in
the trade deficit. The outward shift in the IS curve pushes up interest rates.
Higher interest rates mean an appreciated dollar, and so imports rise and exports
fall.
Up until now we have implicitly assumed that the composition of aggregate
demand has no effect on the productivity of industry: businesses have been
implicitly assumed to be equally happy and equally productive whether they are
producing consumption goods, investment goods for domestic use, goods and
services that the government will purchase, or for the export market. Yet this is
unlikely to be true. As you will recall from your microeconomics courses, the
point of international trade is to export those goods which your economy is
especially productive at making for goods which your economy is relatively
unproductive at making.
As large deficits that increase interest rates raise the value of the exchange rate,
export industries—likely to be highly productive—shrink as exports shrink,
reducing total productivity. Nobody, however, has a very sound estimate of how
large these effects might be.
[Figure: Budget Deficits and the International Sector]
Long Run Effects
In the long run, the increase in the deficit means that more of total product goes
for consumption or government purchases, and less remains for investment and
capital accumulation. The increase in the deficit reduces the investment share of
GDP. Thus the long-run growth path of the economy has a lower capital-output
ratio and a lower level of output per worker associated with the current level of
labor productivity.
But there’s more. A higher deficit means a higher debt, which means that the
government owes more in the way of interest payments to bondholders. Over
time—even if the level of the deficit is kept constant—the increase in interest
payment will require tax increases. And these tax increases will discourage
entrepreneurship and economic activity. In addition to the reduction in output
per worker resulting from the lower capital-output ratio, there will be an
additional reduction in output per worker: the increased taxes needed to finance
the interest owed on the national debt will have negative supply-side effects on
production.
That is the dominant view that economists hold of the effects of large
government deficits. But there are serious critiques of this dominant view, the
most powerful of which is mounted by Professor Robert Barro of Harvard
University. These alternative views are covered in chapter 17.
[Box: How large were the effects of the Reagan deficits?]
To be written…
Chapter summary
Main points
1. The standard measure of the budget balance—the one you see in the
newspapers and on television—is the “unified cash” balance. It is “unified” in
that it pulls together all of the federal government’s revenues and expenditures.
It is “cash” in that it counts cash paid out and cash paid in—and does not take
account of changes in the value either of assets or of liabilities.
2. The standard measures of the budget balance is not a very good measure of
fiscal policy.
3. Most people analyzing the government’s budget adjust the budget balance
first to remove the effect of automatic stabilizers and the business cycle on the
government’s balance, and second to remove that component of debt interest
which merely compensates bondholders for the inflation-driven real reduction in
the value of their principal.
4. A few economists go further, and adopt either “capital budgeting” or
“generational accounting” approaches to the federal budget.
5. The dominant view in economics is that a spending increase unaccompanied
by a tax increase—and thus financed by issuing debt—stimulates consumer
spending and raises GDP in the short run, but depresses savings and investment
and slows economic growth in the long run. But a challenge to this dominant
view comes from Robert Barro, who argues that a spending increase
unaccompanied by a contemporaneous tax increase should have no effect on
consumer spending or savings and investment. Why should our behavior
depend on when our agent happens to submit the bill we owe it?
Analytical exercises
[To be written]
Policy exercises
[To be written, and revised for each year within editions]