Congressional Budget Office
IMF Fiscal Affairs and Research Departments
Conference on Fiscal Policy
Implementation Lags of Fiscal Policy
Doug Elmendorf
Director
June 2, 2009
The American Recovery and Reinvestment Act
(ARRA)
During Congressional debate from mid-December to
mid-February, the U.S. economy was contracting
rapidly.
The Federal Reserve cut the funds rate essentially to
zero in December and thus was out of ammunition for
its principal policy weapon.
As a result, there was a perception of great urgency
in providing fiscal stimulus.
2
How Important Was Timeliness?
Many forecasters expected a large gap between actual and
potential output to persist for some time. (In CBO’s forecast
the output gap was 7 percent of potential in 2009 and 2010
and 5 percent in 2011.) Therefore, policies that provided
stimulus for an extended period of time seemed appropriate.
Moreover, fiscal stimulus that ends before the economy has
started to regain its footing runs the risk of exacerbating
economic weakness when the stimulus ends.
Still, with the economy contracting rapidly, many analysts
(and politicians) wanted stimulus to work quickly in order to
mitigate further deterioration in the economy.
3
Components of the ARRA
Total
Amount
Share Disbursed by End of
Fiscal Year
($ billions) 2009 2010 2011
Discretionary spending (Highways,
mass transit, energy efficiency,
broadband, education, state aid)
308 11% 47% 72%
Entitlements
(Food stamps, unemployment
compensation, health IT, Medicaid
matching rate, refundable tax
credits)
267 32% 73% 91%
Revenues
(Personal tax credits, business,
energy, infrastructure)
212 31% 116% 119%
Total 787 24% 74% 91%
4
Estimated Budgetary Effects of ARRA
Billions of Dollars
Mandatory Outlays
Discretionary Outlays
Revenues
5
Spend-Out of Discretionary Budget Authority Usually
Takes Time
Share spent by end of:
1
st
year 2
nd
year 3
rd
year
Highways 27% 68% 84%
Water Projects 4% 24% 54%
Defense 65% 88% 96%
Average 60% 85% 95%
7
Special Factors for ARRA
ARRA included a number of greatly expanded or
brand new programs, which slows spend-out.
Can the White House apply pressure so that
spending occurs more quickly?
Legislation could:
– Waive environmental reviews
– Allow contracts to be awarded without competitive bidding
– Award money to fastest movers rather than by formula
across jurisdictions
– Offer financial incentives to contractors for faster completion
– Give deadlines for states to obligate money
8
Infrastructure Outlays as a Result of ARRA
Billions of Dollars
9
Entitlements
Transfer payments to individuals (such as
unemployment benefits) go out from the federal
government quickly.
Aid to states goes out quickly if it builds directly on
existing formulas (such as increased federal support
for Medicaid) but more slowly otherwise (such as
education support).
– Key macro question: How quickly and how much do state
budgets respond? Implementation lags include not just time
between enactment and changes in federal outlays or
receipts, but also time between federal budget changes and
changes in GDP.
10
ARRA Outlays to Date (Through May 22
nd
)
Agency
Budget
Authority
Outlays
Through May 22
nd
Outlays
as a Percentage of
Budget Authority
Education
97,407 1,735 2
Health and Human Services
62,279 17,674 28
Transportation
48,120 69 <1
Energy
38,735 44 <1
Labor
36,423 3,695 10
All Other Agencies
96,077 13,476 14
Total
379,041 36,693 10
Millions of Dollars
11
Revenues
Changes in withholding occur quickly.
Rebates take several months, and more during tax-
filing season.
Deferring or eliminating future tax increases affects
the budget only with a lag. How quickly does it affect
household spending?
12
Estimated Effect of 2008 Rebates
20082007
11500
11000
10500
10000
9500
9000
Actual Income
Actual Consumption
Excluding
Rebates
Excluding
Rebate
Effect
Billions of Dollars
13
How Can Implementation Lags Be Reduced?
Have projects “shovel-ready.” But is this practical on
a large scale?
Pick tax and spending changes on the basis of
implementation speed. But this requires compromise
on other aspects of desirable fiscal policy: Who would
be helped? What additional goods and services
would society consume?
14
Estimated Macroeconomic Effects of ARRA
Fourth Quarters of Calendar Years
2009 2010 2011 … 2019
Real GDP (Percentage change from baseline)
Low estimate of effect 1.4 1.1 0.4 -0.2
High estimate of effect 3.8 3.4 1.2 0
Unemployment Rate (Percentage-point change from baseline)
Low estimate of effect 0.5 0.6 0.3 0
High estimate of effect 1.3 1.9 1.0 0
Employment (Millions of jobs change from baseline)
Low estimate of effect 0.9 1.2 0.6 0
High estimate of effect 2.3 3.6 1.8 0
15
Effect of ARRA on the Output Gap
Gap Between Actual and Potential GDP as Percentage of Potential GDP
20152010200520001995
8
6
4
2
0
-2
-4
-6
-8
-10
-12
Without
Stimulus
Legislation
With
Stimulus
Legislation
16
