View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

Economic Brief
APRIL 2009, EB09-04

What We Do and Don’t
Know about Discretionary
Fiscal Policy

At one time, the idea that the government should respond to an economic
downturn by using fiscal policy to stimulate aggregate demand would
have been met with little dispute among economists. Such actions follow
the“Keynesian”school of thought on fiscal policy, named after John
Maynard Keynes who laid its groundwork in 1936 with his General Theory
of Employment, Interest, and Money.

By Renee Courtois

There is a lack of consensus
about the effects fiscal policy measures
may have during a recession. But we can
draw some general conclusions
from economic theory.

However, by the 1970s Keynesian thinking on discretionary fiscal policy
(that is, fiscal policy that is implemented to stabilize the economy around
business cycles) fell out of favor with economists and policymakers alike.
Fiscal policy entailed significant lags between the time it became evident
that countercyclical fiscal policy was needed and when policy could
actually impact the economy. When the 1970s saw stagflation, or
concurrent falling output and rising prices, policymakers were convinced
that a new strategy was needed, and focus soon turned to the importance
of monetary policy.1
By the early 1980s, monetary policy enjoyed greater credibility, and a
consensus emerged that it was a much more successful tool than fiscal
policy for dealing with worsening macroeconomic conditions. Some
economists argued that there may be a role for discretionary fiscal policy
only should monetary policy hit the“zero bound”– that is, when the
policy interest rate approaches zero, a very rare occurrence indeed.
This, of course, is exactly what has transpired following the recession
that began in December 2007. In December 2008, the Federal Open
Market Committee, the Fed’s policymaking body, established a range for
the target federal funds rate, its primary policy instrument, from zero to
twenty-five basis points. Once interest rates fall this low, the Fed is limited
in what else monetary policy can do to stimulate the economy. This does
not mean the Fed is powerless to address the recession; for example,
it has undertaken“credit easing,”in which the Fed supplies large amounts
of liquidity by purchasing a variety of securities, even though the fed
funds rate stays the same. Regardless, the severity of the downturn
has led many economists and policymakers to reconsider the role of
discretionary fiscal policy in counteracting the recession.


We are in a peculiar position, however, regarding fiscal policy. Because
monetary policy has been the tool of choice for managing business
cycles for several decades, the literature on the theoretical and practical

implementation of discretionary fiscal policy is not nearly as developed.
Furthermore, economists differ famously on models that purport to
explain how fiscal policy affects the economy. Keynesian models tend to
focus on stimulating aggregate consumption in order to boost employment and economic output in a downturn, whereas Neoclassical models
focus on the process of resource reallocation following a shock from less
productive to more productive uses, to move employment and prices back
to equilibrium. They focus on different determinants of economic growth
– the former on aggregate demand, the latter on aggregate supply. Many
of the disagreements economists have over fiscal policy fundamentally
address the extent to which government provided goods and services can
enhance private economic activity. Specifically, economists differ on the
nature of assumptions within economic models that determine how
fiscal policy would impact the economy, yielding different fiscal policy
prescriptions in the face of an economic downturn.
Despite these differences, economists are not doomed to disagree on discretionary fiscal policy. Models haven’t yet been created that incorporate
all the“frictions”in implementing fiscal policy that we know exist in
reality. The result is that no matter which assumptions in theoretical
models one thinks are correct, what theory prescribes and what works in
practice can often be different. Therefore, it is also practical to evaluate the
likely effects of discretionary fiscal policy through a cost-benefit analysis.
The purpose of this Economic Brief is to map out factors that would be
relevant in a cost-benefit analysis of discretionary fiscal policy so that
readers can draw their own informed opinion on the matter.
ADDREssIng An EConomIC DoWnTuRn
Suppose the economy is hit by a negative business cycle shock that,
among many possible outcomes, reduces employment and overall
economic output. In the current recession, this took the form of a
severe housing downturn that transcended into a financial crisis that
significantly hindered overall economic activity.
It would be possible for policymakers to simply leave the economy
alone, since economists believe that an economy that has experienced
a negative shock will gravitate back toward its potential level of activity
eventually. However, this process can be lengthy and quite costly in terms
of lost income, jobs, and a host of other potential outcomes. In practice
this process would be aided by“automatic stabilizers,”or taxation and
spending rules already existing in the government’s budget that cushion
losses to disposable income inversely with the business cycle. Still, the
more severe the downturn, the less likely policymakers are to leave the
economy alone to correct itself.
Policymakers may instead look toward more active policies to buffer the
economic downturn, such as monetary policy. In a traditional downturn,

PAgE 2 EB09-04

the Federal Reserve will lower interest rates to stimulate business investment and consumer spending, thereby restoring output. Despite being
constrained by the zero bound, the Fed has in the present recession
actively pursued alternative ways to provide liquidity to financial markets.
In addition, most economists agree that troubles in banking and financial
markets are the fundamental problem that needs to be addressed.
However, the severity of the downturn has led to a renewed debate
over using fiscal policy, too, to buffer the economic downturn.
In this vein, the goal of discretionary fiscal policy can reasonably be stated
as mitigating losses in jobs and output resulting from a negative business
cycle shock.2 In addition, the point of fiscal stimulus is to affect jobs and
output during the economic downturn, as opposed to impacting the
economy after a recovery is already underway.
Fiscal stimulus generally takes the form of tax reductions (reductions
in government revenue) or spending increases (greater government
outlays). Stimulus-related tax cuts are intended to boost individuals’
disposable (after tax) income and thus boost consumption and aggregate
demand.3 Consumption is 70 percent of overall gross domestic product,
so consumers are a natural target for economic stimulus. Through
government spending stimulus, the government funds the production
of goods and services (either by subsidizing private activity or by
increasing government activity itself), and, by doing so, has the potential
to offset the negative shock by providing employment to both human
and capital resources that have been adversely affected by the negative
shock to output.
Much discussion around fiscal stimulus centers on the size of the“multiplier,” a measure of whether each dollar of government spending or tax
cuts boosts output by less than a dollar (where the multiplier is less than
one), more than a dollar (more than one), or a exactly a dollar (equal to
one). Estimates of the multiplier vary widely in economic literature, but
for purposes of a cost-benefit analysis, what matters is whether factors
that enhance a stimulus package’s effectiveness are likely to dominate
factors that detract from its effectiveness.
sTImuLus ThRough TAx CuTs
Stimulus-related tax cuts for individuals can come in the form of a
lump-sum tax rebate or a cut in tax rates. The effectiveness of each is
determined by whether people spend the corresponding increase in
disposable income and, if so, when.
Timing for tax cuts is relatively straightforward: when tax cuts impact the
economy depends on how quickly they’re implemented (when they put
“cash in hand”), whether people react to an announcement of coming tax

cuts in advance of their implementation, and how quickly people spend
whatever portion of them they ultimately use for consumption.
Most analyses of tax cuts reference the“permanent income hypothesis,”
which states that consumers prefer to avoid disruptions in their level of
consumption from one period to the next, so they draw down savings in
bad times and save in good in order to“smooth”consumption over their
life. According to this theory, changes in income will only have a large
effect on consumption behavior if the changes are permanent. This
implies that tax rebates and temporary tax cuts should have no large
effect on consumption, while permanent tax decreases should cause
relatively large increases in consumption.
A common argument against the efficacy of tax cuts, even permanent
ones, as stimulus to the economy is“Ricardian equivalence,”which says
that consumers who receive a tax cut are able to foresee that the tax
cut must be paid for by a future tax increase, causing them to save the
entirety of the tax cut with no boost to consumption.4 In order for Ricardian equivalence to hold in reality, individuals must be forward-looking,
and they must care enough about the future consumption of themselves
or their heirs to change their behavior as a result. If so, they may save
the tax cut to pay for the future tax increase, or leave bequests to their
children to do the same.
The inherent contradiction in cutting taxes to stimulate a limping economy is that theory implies tax cuts must be permanent to have a large
impact on consumption, but fiscal stimulus is used to address a temporary
problem (the economic downturn). Thus, it may not be credible for the
government to announce tax cuts as being permanent, either because
people presume that other fiscal priorities will prevail after the downturn
has passed, or because individuals judge that the government will raise
taxes in the future to finance the present tax cut. This is an understandably opaque issue for consumers, even if they are forward-looking, since
the U.S. government has in recent history appeared able to run deficits
for quite some time with no obvious macroeconomic side effects.5
There is one condition under which any type of tax cut might be used to
increase consumption more than basic theory suggests. Consumers who
face borrowing constraints (such as difficulty getting a car loan or a credit
card) are more likely to use a tax cut to make a purchase for which
they have not been able to save an adequate amount. For this reason,
individuals who are poorer have a higher marginal propensity to consume
(MPC) out of an increase in income. Further, borrowing constraints mean
these consumers must wait until the given tax cut is“in hand,”implying
they are not likely to react to an announced tax cut until it has actually
been implemented.

A final way in which tax cuts may boost output is if they improve incentives to work. This effect is exacerbated the more progressive the tax
system is because the marginal tax rate (the increased tax rate you face
with each additional dollar of income you earn) is higher under such a
system. When individuals choose to work more (increase their labor
supply), they often consume out of the added income they receive,
which boosts economic activity.
Empirically there is great uncertainty on the effects of tax reductions
enacted as part of economic stimulus, since there have been very few
episodes in history of stabilization-focused tax cuts from which to test.6
Also, measuring the impact of tax cuts implores economists to separate
out changes in output that are a direct result of the tax cut from changes
in output that are occurring anyway, as from business cycles. One consensus of empirical studies is that borrowing constraints are a very important
determinant of whether tax cuts or rebates are spent.7
sTImuLus ThRough govERnmEnT sPEnDIng InCREAsEs
In the textbook version of a spending stimulus, following a negative shock
to the economy the government will quickly pass and implement a
spending program applied directly to the human and capital resources
in the economy that have become idle as a result of the shock. As the
government works to employ idle resources, economic activity falls less
than it would have in the absence of government intervention, achieving
the goal of quickly buffering losses to jobs and output.
Clearly this process is much more complex in reality. Timing alone can
impede successful stimulus spending. The current recession was not
identified as such by the National Bureau of Economic Research until December 2008, a full year after it began. generally speaking, implementing
a stimulus package in time for it to address the downturn entails a very
quick turnaround between the time we figure out we are in a recession
(and indeed one bad enough to warrant a stimulus package) and the
time the government can agree on and pass a specific stimulus package.
Further, for stimulus spending to affect the economy during the downturn, the projects must be able to be implemented quickly: projects must
be“shovel-ready”(not requiring significant planning, research, and
development) and must be possible to implement without“bottlenecks”
(overburdening an industry with projects beyond its capacity to produce
in a timely fashion).
hoW To sPEnD WIsELy
By far the largest hurdle of government spending stimulus is what to
spend on. To be successful, stimulus spending must be targeted to uses
where there are idle resources. Alternatively, if stimulus spending is
targeted to industries that are at full employment, they will be forced to
compete with private projects for resources. For example, if there is full

EB09-04 PAgE 3

employment in road repair, then the government cannot stimulate road
repair without draining construction workers from another project, reducing the amount produced in that other endeavor. In this case government
spending displaces private economic activity. Furthermore, workers shifted
to these new uses may not be as productive at them as in what they were
previously doing.These“human capital leakages”need not be significant;
for example, a worker who moves from constructing bridges to repairing
roads may find the skill sets match quite well, with minimal losses in productivity. For workers who move from very different industries, however,
the losses of human capital may be larger and may take time to correct as
workers get trained in their new uses.
Despite human capital leakages, it may still be socially beneficial for
government spending to displace private spending if it creates output
that is more highly valued than the projects it displaces. However, since
the government does not face the market test for efficiency that private
enterprises do (it faces no competition and thus no profit motive), projects
chosen by the government can be less highly valued than the marketdriven projects they displace.Thus, to the extent that private economic
activity is displaced and human capital adjustments are significant,
the impact of stimulus spending will likely be reduced.
The government can avoid displacing private economic activity completely
if government spending is targeted to resources that are idle, since employing them in nearly any way will be a productive improvement that will
add to economic output. In one extreme, government spending will have
to target stimulus spending exclusively to idle resources.
Three factors determine the government’s ability to target stimulus spending toward idle resources. First, doing so requires the government
to have perfect information about where resources are idle. In theory the
government would have to identify specific firms that have laid off workers
and capital; in practice, the government might start by looking at unemployment rates across industries. Second, stimulus spending will affect complementary inputs. For example, since construction workers do not work
alone, stimulus on road repair will also affect the markets for trucks, shovels,
and concrete.To the extent that trucks, shovels, and concrete were less adversely affected by the economic downturn, some of the government’s
spending will compete with private resources for use of those inputs, offsetting production elsewhere.Third is the political reality that politicians face
pressure to direct spending toward projects that are of local interest, as
opposed to being beneficial for immediate boosts to the economy.
The first two factors imply it is not reasonable to expect that the government can target spending exclusively to idle resources; there will always be
some degree of waste.The third factor detracts from the likelihood that
government will choose to target the bulk of spending to idle resources.

PAgE 4 EB09-04

A caveat is in order. If the government is able to successfully target stimulus spending mostly toward idle resources, it still may not be desirable to
do so, for it matters why those resources were idle in the first place. Were
they idle because of a spillover from the overall economic downturn, or is
the firm or industry experiencing an overall structural shift downward
that has simply been intensified from the economic downturn? Declining
industries may well be successful targets for economic stimulus in the
sense that it boosts their production in the short run, but it does so at the
expense of subsidizing moribund firms in the long run. Such actions keep
resources in inefficient uses at the expense of lost opportunities elsewhere. Economic stimulus, arguably, should not impede these marketdriven structural shifts in the economy.
A famous parable of stimulus spending is that the government could successfully stimulate the economy by paying people to repeatedly dig holes
and fill them up again. While this would certainly employ people and add
to economic activity in the accounting sense, it would, arguably, not produce anything or provide a valued service. Further, the job would disappear as soon as stimulus funds went away, with no lasting benefit on the
economy. Thus, a common argument is that government stimulus spending should, to the extent possible, add to the productive potential of the
economy. However, projects that have these long-run benefits often entail
longer planning and investment stages than projects that will provide
short-run stimulus. While such projects may be well-justified because of
their longer-run benefits, and can thus gain traction under the momentum of increasing government spending, they should not be confused
with economic stimulus if they are not likely to impact the economy during the downturn.
Long-Run ConsIDERATIons
We have focused this discussion entirely around what makes a stimulus
package successful as a timely buffer to the economic downturn. This is
a short-run goal that ignores many possible outcomes of a stimulus
initiative. Chief among these other concerns are:
Long run efficiency: As mentioned above, stimulus spending
targeted to idle resources risks subsidizing inefficient firms or
industries. To avoid this, it should be directed to industries that are idle
as a result of the economic downturn, not a longer-run decline.
Concerns over the national debt: Both tax cuts and spending increases add to the budget deficit (reduce the budget surplus) and
thus bode negatively for government debt. Whether this is perilous
is an unresolved issue in macroeconomics. In theory the economy
can run a deficit forever as long as its rate of economic growth plus
the rate of inflation is high enough to finance the interest it must
pay on its debt. However, if the government must borrow substantially to fund the stimulus, it may cause upwards pressure on interest
rates, reducing the length of time over which the government can


sustainably run a deficit without macroeconomic consequences.
Ratchet effect: Are spending increases permanently increasing the size
of government? Spending programs can develop a life of their own
that often makes them longer-term than originally intended.
Issues of equity: If tax cuts reduce distortions by reducing the progressive nature of the tax code, is that desirable?

Finally, if there is a financial crisis, all bets could be off. Such crises can interrupt the flow of resources through the economy, which can reduce the
impact of any type of government spending (or private spending for that
matter). Historically, recessions in which there is also a financial crisis, or
that have been precipitated by a financial crisis, are longer in duration.
While monetary policy is the preferred tool among most economists for
blunting the negative effects of economic downturns, it is limited when it
hits the zero bound, resulting in renewed interest in fiscal policy to stimulate the economy. Fiscal stimulus is a daunting undertaking. There is a lack
of consensus on how fiscal policy measures to stabilize the economy are
likely to play out, but we can draw some general conclusions from economic theory:
A tax cut is more likely to be effective as stimulus if it:
Is implemented quickly
Is permanent
Reduces distortions, and thus provides incentives to work more
Is targeted to consumers most likely to spend them:
Those who are borrowing constrained (if it is possible to
identify them)
Lower income households with a higher MPC

There are minimal human capital leakages as resources move
The government-sponsored projects are of higher value than the
private projects that are dispalced.

The fiscal stimulus process is rife with paradox.Tax cuts should be permanent to have a large impact on consumption, but tax cuts announced as
being permanent may not be credibly so, undermining their impact.
Spending programs must be implemented quickly to impact the economy
during the recession, but spending for short-run gain is often at odds with
long-run goals. It is unfortunate that we don’t have better answers to these
questions at a time when the economy is in recession. As a result, some
economists would argue against using discretionary fiscal policy in the current environment, concluding that its potential benefits are outweighed by
costs.We will undoubtedly learn a great deal from the 2009 American Recovery and Reinvestment Act that we can apply the next time fiscal stimulus is considered.

Renee Courtois is an economics writer in the Research Department of the Federal Reserve Bank of Richmond.

Blinder, Alan. 2004.“The Case Against the Case Against Discretionary Fiscal Policy.”Center for Economic Policy Studies Working Paper 100, Princeton University.


government spending increases or tax reductions when the economy is at full employment may
have very different effects than what are discussed here. For example, see: Romer, Christina, and
David Romer. 2007.“The Macroeconomic Effects of Tax Changes: Estimates Based on a New Measure
of Fiscal Shocks.”National Bureau of Economic Research Working Paper 13264.


Of course, tax cuts for stimulus purposes can take many forms in addition to tax cuts to individuals.
Common examples include investment credits and changes to sales taxes. Such tax changes are explicitly advertised as being temporary to encourage consumers and businesses to move consumption
in the future to today. For a review of the literature on these see Blinder (2004).


A stimulus spending initiative will be most effective if it:
Is implemented quickly on projects that are“shovel ready”and without“bottlenecks”
Targets idle resources that are idle due to spillover effects of the economic downturn, not a longer-trend decline
To the extent that it is targeted to resources that are not idle:

Barro, Robert. 1979.“On the Determination of the Public Debt.”Journal of Political Economy
87: 946-971.


Blinder (2004).


Romer and Romer (2007).


Blinder (2004).

EB09-04 PAgE 5