12 Fiscal Policy Prelim Disc Draft 2018.09.11

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

RESPONDING to the GLOBAL FINANCIAL CRISIS

What We Did and Why We Did It

The Fiscal Response to the Great Recession:

Steps Taken, Paths Rejected, and

Lessons for Next Time

Jason Furman

"#$%& $%' "() **)'+ " ,-.// 0" $*'" 1) 2) %3 /*" 4*3 25 2 62) "7 /% "4 *2 )% *$25 /*$0 " 8- /*/95- 3: ";$3/*"<6%9$*"=$3">)?62+"

>5%)-2/% " /7" 2 .)" @.52 )" #/63) " A$25 /*$0 " 8- /*/95- "B /6*-5 0 "C A8BD "$*'">)?62+" E33532$*2 "2/"2.)" 1%)35')*2" 7/%"

8-/*/95-"1/05-+"7% /9"FGGHI!GJ"1%5*-5?$0" >)?62+">5% )-2/%" /7"2.)" A8B"$*'"E335 32$*2" 2/"2 .)"1%)35 ')*2"7 /%"

8-/*/95-"1/05-+"7% /9"FG!!I!K"$*'"B.$5%9$*"/7"2.)"B/6*-50 "/7" 8-/*/95- "E'&5 3)%3 "7% /9"FG!KI!L:"M.53" ?$?)%"

N)*)752)'"7%/9".)0?760"-/*&)%3$25/*3"$*'"-/99)*23"7%/9"O$N%5)0"B./'/%/=IP)5-.J"M59"O)52 .*)%J "A)0 05) "Q5 $*RJ"

895"A$S$96%$J"#$*S"1$603/*J" B.%5325 *$"P/9)%J" ;$+",.$9N$6R.J";T*",2)5*33/*J"$*'"Q$%%+",699)%3:"M/9"

P)'N6%*"9$')"*69)%/63")U- )00 )*2") '52/% 5$0 "36R R)325 /*3" $*'"P/*)$0 ">)3$5 "$*'" @503/*" 1/=)00 "4 44 "?% /&5')'"

/6232$*'5*R"% )3)$% -."$33 532 $*-)J "$*'" >)N/%$." V-B0)00 $*"2 %5 )'"&$ 05 $*20+ "2 /"S))?" 9)"/*" 2%$- S"=52 ."2 .53" ?%/W )-2: "

Note: The views expressed in this draft are strictly those of the author(s).

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Introduction

The fiscal response to the Great Recession started when President Bush signed the

Economic Stimulus Act of 2008 on February 13, 2008 and finished when the payroll tax

cut enacted under President Obama expired at the end of 2012. Congress enacted at least

18 different laws that explicitly included discretionary fiscal stimulus totaling over $1.5

trillion during those five years, with about half of that coming from the American

Recovery and Reinvestment Act signed into law by Obama on February 17

2009.

The

stimulus was 54 percent tax cuts, 19 percent individual transfers, 11 percent state and

local fiscal relief and 16 percent public investment—with nearly all of that public

investment coming in the Recovery Act. The discretionary fiscal stimulus averaged 2.0

percent of GDP over those five years, boosting the level of GDP from what it otherwise

would have been by a maximum of 3.4 percent in the third quarter of 2010 with a smaller

effect thereafter. In addition, automatic stabilizers brought the total magnitude of the

countercyclical fiscal response to an average of 3.4 percent of GDP, the largest fiscal

response to a recession in U.S. history.

The first section of this paper describes the origins of these stimulus measures and

the second section summarizes their magnitude and composition. The third section

addresses some of the evidence evaluating the efficacy of the stimulus. The fourth section

describes some of the approaches that were considered by the Obama transition team

and Administration but not implemented. Finally, the last section discusses some lessons

for the future.

This paper is focused on the macroeconomic analysis of the fiscal response to the

financial crisis, and the extent to which it prevented a deeper recession, and helped

speed the economic recovery. It only briefly discusses other goals, like protecting the

most vulnerable as well as improving infrastructure, healthcare, broadband and energy

efficiency. These are all very important topics but largely outside the scope of this work.

I. The Three Phases of Stimulus

The fiscal stimulus went through three phases: (i) an initial set of responses in 2008 that

aimed to be “timely, targeted, and temporary”, (ii) a large response in the Recovery Act

in early 2009 that aimed to be “substantial, speedy, and sustained”, and (iii) a number of

laws passed in subsequent years that in retrospect could be described as “opportunistic,

extended, and under the radar.”

The 2008 Stimulus: “Timely, Targeted, and Temporary”

In 2007 the U.S. economy was deteriorating rapidly. Most notable was a rise in the

unemployment rate from 4.4 percent in March 2007 to 5.0 percent in December 2007—a

"M.53"'/)3"*/2"5*-06')"2.)"E02)%*$25&)"V5*5969"M$U"CEVMDJ"M%/6N0)'"E33)2"P)05)7"1%/R%$9"CMEP1DJ"/%"

36??/%2"7/%"2.)"O/&)%*9)*2I,?/*3/%)'"8*2525)3:"

"</%"$*")U-)00)*2"$--/6*2"/7"2.)"P)-/&)%+"E-2"$*'"523"0/*R)%I2)%9"32%6-26%$0"0)R$-+"3))"O%6*= $ 0'"CFG !F D:"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

magnitude of increase that, based on historical precedent, signaled the coming of a

recession that would cost many more jobs. Between September and December of 2007

the Federal Reserve reduced its target federal funds rate from 5.25 percent to 4.25

percent, but because of lags in the effectiveness of monetary policy this was not expected

to substantially bolster the economy until late 2008. In addition, the Fed took a number

of other actions, like expanding the discount window and establishing the Term Auction

Facility (TAF), that had unknown efficacy and timing. As a result, there was significant

concern that the economy would get worse before monetary stimulus set in.

On December 5, 2007, Martin Feldstein (2007) became perhaps the first

prominent economist to raise the possibility of fiscal stimulus. “The American economy

is now very weak and could get substantially weaker,” he wrote. “Current economic

conditions call for lowering interest rates and for enacting a tax cut now that is

conditioned on economic developments in 2008.” Lawrence Summers (2007) went on

to explicitly call for fiscal stimulus on December 19, 2007, saying it should aim to be

timely, targeted, and temporary.

By this time President Bush had already asked Treasury Secretary Henry Paulson

and his White House economics team to evaluate the need for a fiscal stimulus program.

In January, Paulson and the White House team reported back to Bush that a stimulus

plan was needed urgently and could be devised in a manner that would receive broad

bipartisan support and be quickly enacted. Bush wanted a package with the guiding

motivation being to put money into consumers’ hands quickly to provide a fast boost to

the economy. On January 18, 2008, Paulson unveiled the broad parameters of a stimulus

package, saying, “Our economy is growing slower than expected, and that means we need

to act quickly to put together a package that is temporary, simple enough to get enacted

quickly, effective at boosting growth and job creation this year, and large enough to make

a difference.” (Paulson 2008). This was enacted with strong cooperation by

Congressional Democrats in a matter of weeks and largely took the form of one-time

refundable individual tax rebates. The first electronic payments were made in April 2008

and the first checks mailed out in May (Internal Revenue Service 2008).

While the economic threat at the time appeared to be nothing worse than a typical

recession, there were two prime rationales for the stimulus: (i) fiscal action could provide

a faster boost to the economy than monetary policy, filling in some of the gap before the

interest rate cuts started kicking in; and (ii) given that the effectiveness of monetary

stimulus was uncertain, it was best to diversify the response by using multiple

instruments (Elmendorf and Furman 2008).

The Recovery Act: “Substantial, Speedy, and Sustained”

Despite the fiscal stimulus and further interest rate cuts, the economic and financial

crisis intensified, particularly in the autumn after Lehman Brothers failed, with the

unemployment rate rising from 5.0 percent to 7.3 percent over the course of 2008. By

early January 2009 the numbers showed that the economy had lost an average of

510,000 jobs a month for the previous three months, a loss that was subsequently

revised up to 647,000 jobs a month.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

As the economy weakened, the target size of stimulus plans increased. Running for

president, Obama originally proposed a fiscal stimulus in January 2008, proposed another

one in April, and expanded the size of the proposed stimulus over the course of the year,

including in June, August and October. Fiscal stimulus was also motivated by the concern

that the Fed, which had effectively cut its target rate to zero in December, 2008, was

running out of conventional monetary policy options and that all tools that relied on the

financial sector for transmission to the real economy were of uncertain and potentially

limited impact due to the crisis and associated breakdown of the financial system.

Building on these plans the transition team, which included Jack Lew, Dan Tarullo,

Jason Furman, and Austan Goolsbee, met with President-elect Obama on November 12

and presented him with the following rationale for a recommended $300 billion stimulus:

Broad expectation that economy will be anemic in 2009 and

beyond: job losses are accelerating; consumption deteriorating; Blue

Chip/WSJ Survey projecting negative GDP growth in Q1 2009 while other

analysts project negative growth for all of 2009; financially led recessions

tend to last longer—Goldman Sachs projects unemployment to peak at

over 8% and remain high for most of 2009 and 2010.

Running out of other options: not clear where economic jumpstart

will come from with tight credit markets, tapped out consumers, trading

partners following behind economic cycle and heading into recession and

a stronger dollar.

Appropriate tool in current environment: fiscal stimulus only real

option since monetary tools are exhausted and there is room for monetary

policy response if fiscal policy overshoots—recession risk much greater

than fears of inflation and crowding out (though inflation concern will

increase once a recovery is underway).

The question of whether or not the stimulus should be paid for was also debated

during this meeting. Economically, there was no need to pay for a stimulus, and

politically, the desire to move quickly and without complication to passing relief was

compelling. On the other side of the argument, some were worried about the impact that

rising debt would have on financial markets. Ultimately Obama decided to advance a

stimulus plan that could pass on a standalone basis but to make it clear it was part of a

longer-run budget that, taken as a whole, would make the debt sustainable.

At the time, even a $300 billion unpaid-for stimulus would have been considered

large, pushing the bounds of what was possible in Congress. Any stimulus was expected

to need the votes of the Blue Dog coalition, about 50 House Democrats concerned about

the deficit and debt who had explicitly said they would not vote for another unpaid

stimulus. House Democratic leaders had been considering a $150 billion stimulus, with

some talking about packages of up to $300 billion. Leading progressive economists were

advocating similar top-line numbers: a November 19

letter organized by the Center for

Economic Policy Research and signed by hundreds of economists including George

Akerlof, Dean Baker, James Galbraith, Lawrence Mishel, and Joseph Stiglitz called for a

$300 billion to $400 billion stimulus that would be “spent quickly” (Akerlof et al. 2008).

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

In an effort to create political space to pass a larger number through Congress the

transition team reached out to several signatories to encourage them to raise their public

requests, telling one economist speaking to the House Democratic caucus in the late fall

to call for $1 trillion in stimulus instead of the roughly $500 billion he had been planning

to recommend.

The economic news and forecasts continued to deteriorate. Macroeconomic

Advisers, a leading economic research and forecasting firm, made the largest negative

revision to a forecast in its history on December 8

. These developments plus the

filling out of the Obama economic team—Christina Romer and Summers were both

advocates of a much larger stimulus—led to a further upward revision of the number,

with a memo to Obama stating:

We believe that $600 billion in stimulus over two years would create 2.5

million jobs relative to what would happen in the absence of stimulus.

However, this falls well short of filling the job shortfall and would leave

the unemployment rate at 8 percent two years from now. This has

convinced the economic team that a considerably larger package is

justified… The memo outlines four alternative plan [sic] ranging from

$550 billion to $890 billion with the difference between them being the

state fiscal relief and tax proposals.

The upper end of this range was estimated to be sufficient to close half of the

projected 7 percent of GDP output gap. The memo’s rationale for not attempting to close

the full output gap included: fiscal stimulus was not the only economic tool, concerns

about spooking markets and raising interest rates with too large of a fiscal package, and

the view that it would always be possible to add to stimulus but might not be possible to

subtract from it.

At a meeting with the transition team on December 16

Obama decided to pursue the

largest stimulus that his team thought was politically feasible, his view being that political

constraints would be binding well before any economic concerns about market confidence

would be relevant. This was agreed to be a stimulus in the $800 billion range and Obama

left it to his political and economic team to develop a strategy to hit this target.

The team decided that coming out with an explicit budget number this large would

risk a backlash in Congress, slowing passage. In addition, it decided that proposing a full,

explicit plan could slow passage given the limited resources of the transition team and the

desire of Congress to put its stamp on the measure. Instead, publicly Obama continued to

push for the general concept of fiscal stimulus, using a job target instead of a cost target,

which was originally 2.5 million jobs on November 22 but revised up to “at least” 3 million

by December 20. Based on the economic team’s analysis, this job target corresponded to at

least $850 billion in stimulus. In addition, on January 8, 2009 Obama (2009a) listed a

specific set of goals for these investments and framed the issue by saying:

It is not just another public works program. It’s a plan that recognizes

both the paradox and the promise of this moment—the fact that there are

millions of Americans trying to find work, even as, all around the country,

there is so much work to be done. That's why we’ll invest in priorities like

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

energy and education; health care and a new infrastructure that are

necessary to keep us strong and competitive in the 21st century.

Behind the scenes work was ongoing to have a bill well underway before the

inauguration—with specific, detailed meetings with members and staff of Congress

beginning immediately after the December 16

meeting. Originally the transition team

privately asked Congress for a smaller number because it was concerned about a negative

reaction to a larger number, especially from the Blue Dogs, and expected that the total

would increase in the legislative process. This expectation was initially justified as the

House passed an $820 billion ten-year stimulus on January 28, 2009. The cost of the

Senate bill passed on February 10, 2009 had a slightly higher headline number, $838

billion, although the effective stimulus was smaller than in the House bill because the

Senate added an extension of a patch to the alternative minimum tax (AMT), a measure

that didn’t increase the effective magnitude of stimulus because it simply continued a

long-standing practice that would have been included in other legislation anyway. But

instead of the larger total package the Administration hoped and expected to get, the

conference agreement was smaller than either bill, coming in at an originally estimated

headline cost of $787 billion after three Senate Republicans insisted the cost come down.

The effective stimulus was even smaller because of the inclusion of the AMT patch.

Obama signed the Recovery Act into law on February 17.

The guiding philosophy for the Recovery Act was not “timely, targeted, and temporary”

but instead “substantial, speedy, and sustained.” The transition team evaluated a wide

range of potential provisions against several criteria: (i) how quickly would they spend the

stimulus money; (ii) what was their expected multiplier for GDP; (iii) how likely were they

to be made permanent (which was generally viewed as a minus because of concern about

the long-run deficit); and (iv) how “transformative” would they be for various public

purposes beyond immediately increasing GDP. The goal was to come up with a portfolio of

provisions that would span short-run stimulus to long-run transformation.

Originally the architects of the stimulus planned to have a limited set of areas for

“transformative” provisions centered on healthcare, energy, education and infrastructure.

All of these areas were chosen to build on campaign proposals that were more long-run

and structural in nature but could be plausibly (or in some cases, less plausibly) be

separated out from their broader context with a major down payment included in the

Recovery Act. In some cases, these more transformative provisions were motivated less

by their macroeconomic impact, instead heeding incoming Chief of Staff Rahm

Emanuel’s advice that you “never want a serious crisis to go to waste.” For example, the

campaign’s health information technology spending program and energy investment

program were put into the Recovery Act—while the fuller plans for health insurance

coverage and cap-and-trade were left to be enacted later. Based on additional input from

the transition team and Congress, more items were added to the list including subsidies

for broadband in underserved areas and funding for high-speed rail.

When Obama (2009b) signed the Recovery Act into law he said:

Today does not mark the end of our economic troubles. Nor does it

constitute all of what we must do to turn our economy around. But it does

mark the beginning of the end—the beginning of what we need to do to

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

create jobs for Americans scrambling in the wake of layoffs; to provide

relief for families worried they won’t be able to pay next month’s bills; and

to set our economy on a firmer foundation, paving the way to long-term

growth and prosperity.

The New York Times (Stolberg 2009) headline noted that “Signing Stimulus, Obama

Doesn’t Rule Out More”—which turned out to be the case in the next phase.

Post-Recovery Act: Opportunistic, Extended and Under the Radar

The unemployment rate continued to rise, hitting 9.4 percent in May 2009 (before the

bulk of the Recovery Act had even kicked in), which was well above what forecasters had

predicted as recently as early 2009. The economy was even worse, at least in part,

because the many external and internal forecasting models the transition team had used

insufficiently accounted for the role of the financial sector in exacerbating and, most

importantly, perpetuating the downturn. The Administration proposed more stimulus

packages, including in December 2009, September 2010, and the $447 billion American

Jobs Act in September 2011. All of these proposed a combination of public investment,

relief for individuals and states and public investment along the same lines as the

Recovery Act. Congress, however, had lost its appetite for major additional legislation

that was described as economic stimulus and did not enact another bill along these lines

after February 2009—although the House did pass an infrastructure-oriented stimulus

bill in December 2009 but it was not taken up in the Senate.

Nevertheless, many of the different elements proposed by the Administration or

desired by Congress were still passed, attached to other legislation like defense

appropriations bills or FAA reauthorizations. The largest opportunity came when the

2001 and 2003 tax cuts were expiring at the end of 2010. Obama had long called for the

expiration of the provisions in the tax cuts that solely benefited high-income households.

Republicans were in a stronger position after large gains in the mid-term elections. It

was possible that Obama could still have pushed through the expiration of the tax cuts,

but in that case he would have had no leverage to get any additional priorities in the bill.

Instead, he chose to ask Vice President Joe Biden to negotiate a deal with as much fiscal

stimulus as possible—specifically, trading a two-year extension of the high-income tax

provisions for a continuation of tax credits for low-income families with children and

replacing the Making Work Pay ($58 billion annually) tax credit with a larger but less

well-targeted payroll tax cut ($112 billion annually), and establishing 100 percent

expensing for business investment.

This phase of stimulus was necessarily opportunistic. The Administration generally

wanted as much fiscal support as possible, including increased public investment and tax

cuts. Congress generally wanted less stimulus, but was more supportive of additional tax

cuts than funding for infrastructure or other investments and more supportive of

extending existing provisions like expiring unemployment insurance benefits than of

devising new measures.

The net effect was a substantial additional stimulus: $657 billion above and beyond

the Recovery Act by the end of 2012, which resulted in an increase in total fiscal stimulus

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

as a share of GDP in 2010, maintained almost that level in 2011, and prevented a sharp

falloff in 2012. But this was still well below the size Obama requested and did not have

the composition the Administration desired (public investment was almost entirely

absent). The piecemeal nature of these additional provisions spread over at least 13

different pieces of legislation, most of which were not explicitly marketed as stimulus,

also sacrificed the potential benefit from clearly setting expectations and increasing

confidence that a more visible stimulus package might have generated.

II. The Size and Composition of the Fiscal Stimulus

Overall the cumulative discretionary fiscal stimulus totaled $1.537 trillion through the

end of calendar year 2012.

Nearly half of the fiscal stimulus came in the Recovery Act,

which provided a net new $712 billion for the economy through the end of CY 2012.

The

remainder came in at least 17 other laws that included the original 2008 rebate checks,

the cash for clunkers program to encourage auto purchases, a tax credit for hiring the

long-term unemployed, a tax credit for homebuyers, the payroll tax cut, and several

additional provisions.

Table 1

As shown in Figure 1, the overall discretionary fiscal stimulus averaged 2.0 percent

of GDP over the five years from 2008 through 2012, peaking at 2.7 percent of GDP in

2010. This was enhanced by automatic stabilizers which expanded unemployment

"M.)"2/2$0"=$3"Y!:XZ["2%5005/*"-/6*25*R"2.)"N6'R)2$%+"59?$-2"2.%/6R."FG!H"N)-$63)"3/9)"?%/&535/*3J"05S)"

N/*63"')?%)-5$25/*"$*'")U?)*35*RJ"%$53)'"9/*)+"5*"2.)"/62"+)$%3:"

"M.53"'577)%3"7%/9"2.)".)$'05*)"YL\L"N5005/*"5*"2.$2"52"%)70)-23"36N3)]6)*2"B^_"%)I)3259$2)3"$*'"$03/"

36N2%$-23"52)93"05S)"2.)"EVM:"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

insurance, means-tested programs and reduced taxes—averaging another 1.4 percent of

GDP in stimulus and bringing the total stimulus to as much as 4.6 percent of GDP in

2010 and an average of 3.4 percent over the entire period.

Figure 1

The largest category of stimulus was tax cuts for individuals, with total tax cuts

making up 54 percent of the fiscal stimulus as shown in Figure 2—although some of the

items formally classified as tax reductions are analogous in impact to increased

expenditures. Another 30 percent of the stimulus went for relief, either directly to

individuals or as fiscal relief for states. The remaining 16 percent of the stimulus was

devoted to outlays on public investment. Some of the larger or more notable items in

each of these categories along with their cost from 2008-12 are listed in the Appendix.

Figure 2

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

III. Evidence on the Impact of the Fiscal Stimulus

The principal basis for evaluating the fiscal stimulus should match its primary purpose:

raising GDP above what it otherwise would have been, thus preventing a deeper

recession and speeding economic growth. In addition, it would be desirable to analyze

the relative effectiveness of elements of the stimulus, but evaluating individual

provisions proved nearly impossible and those that have are not necessarily comparable.

Nevertheless, some general observations are possible.

Ex Ante Projections of the Macroeconomic Impact

The fiscal stimulus was part of a broader response to the crisis that included monetary

policy and policies affecting the financial, housing, and auto sectors. Evaluating the

causal impact of the entire response, let alone any individual element, is not

straightforward and may even be impossible because we cannot observe the baseline that

reflects what would have happened absent the policy response. For example, observing

that high unemployment rates follow a stimulus may establish nothing more than that

the stimulus was warranted in the first place. Conversely, observing the economy

recovering may just be the natural self-equilibration of the economy and not the result of

macroeconomic policy.

Time series econometric methods have attempted to figure out the impact of fiscal

expansions by separating out “endogenous” fiscal policy that is a response to where the

economy is from “exogenous” fiscal policy that is done for reasons unrelated to the state

of the economy. This method is impossible for the Recovery Act or other fiscal responses

because these are basically like a single data point and the fiscal response was clearly an

endogenous response to the fact that the economy was widely expected to weaken

substantially in the future.

As a result, most of the estimates of the macroeconomic impact of the Recovery Act

and other fiscal measures have been estimated ex ante. Specifically, they draw on

research that was done before the Great Recession to predict the impact that a given

fiscal policy would be expected to have on the economy. Generally, this is done by

assigning a set of multipliers (spread out over time) to different types of fiscal measures.

The multipliers used by the Council of Economic Advisers (CEA) ranged from 1.5 for

public investment outlays and for income and support payments, which were expected to

generate more economic activity as the initial round of stimulus was spent and re-spent

in the economy, to 0.1 for business tax incentives, reflecting the expectation that they

would mostly be saved by businesses in the form of larger retained earnings. CEA’s

multipliers are shown in Table 2, they generally fall in about the middle of the range for

the Congressional Budget Office’s (CBO) low and high multiplier estimates.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Table 2

Estimates for the macroeconomic impact of the Recovery Act alone are provided in

Table 3 generally based on these types of multipliers, showing it adding between 0.7

percent and 4.1 percent to the level of GDP in 2010, when almost all forecasts predicted

it would have its peak impact. Blinder and Zandi (2015) used similar methods to

estimate the impact of the broader set of fiscal expansions.

Table 3

These same methods can be used to estimate the macroeconomic impact of the full

set of fiscal stimulus legislation on GDP starting in 2008, which is shown in Figure 3.

The total raised the level of output by a maximum of 3.4 percent of GDP in the third

quarter of 2010 with the effect remaining around 3 percent of GDP through the nearly

the end of 2011 and falling off to about 2.2 percent in the beginning of 2012. In terms of

growth rates, this means that Q4/Q4 growth rates were raised substantially in 2008 and

2009 but were roughly unchanged in 2010 and 2011 and reduced in 2012 as the addition

to the level wore off. This is equivalent to adding about 3 million jobs at the maximum

point of the stimulus or a cumulative total of about 10 million job-years over the five-

year period the stimulus was in effect.

[

"M.)"$*$0+353"-/9N5*)3"2.)"96025?05)%3"5*"M$N0)"F"=52."2.)"0$R"32%6-26 %) "5* "B 8E "C FG G H D:"M . )"0$ R 3"$ %)"=.+"

2.)")77)-2"/7"2.)"753-$0"3259606 3"/* "/6 2?6 2"53"9$ 5*2$ 5*)' "$2"$".5R. )%"0)& )0")&)* "$72)%"2. )"753-$0"3259 606 3"

523)07"32$%2)'"2/ "N ) "= 52. ' % $=*:"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Figure 3

Ex Post Cross Sectional Effects

In general, it is difficult to evaluate any individual provision of the stimulus on GDP

because numerous provisions were going into effect simultaneously. Nevertheless,

research has exploited different types of variation to estimate the impacts of some of the

individual provisions. These impacts, however, cannot easily be compared to each other

because of different estimation strategies, noise, and because many of them did not

estimate national multipliers.

A number of papers have effectively used random cross-state variation in different

components of the fiscal stimulus to estimate the state-level macroeconomic impact of

these particular measures. Chodorow-Reich et al. (2012), for example, study the impact of

higher Medicaid matches on state economies by looking at the portion of Medicaid

matching that was based on pre-recession Medicaid spending and thus was not simply the

result of the recession. They find a state-level multiplier of about 2. Conley and Dupor

(2013) used a similar method for highway spending and found a much smaller multiplier.

Dupor and McCrory (2017), Dupor and Mehkari (2016), and Wilson (2012) examined

broader Recovery Act spending based on formula grants and found multipliers in between

these two. The different estimated multipliers are probably an artifact of different

modelling strategies, however, and thus not comparable. Chodorow-Reich (forthcoming)

provides more harmonized estimates studying these three sets of stimulus and finds

multipliers of 1.53 to 2.29, with imprecise measurement meaning that none of them are

statistically significantly different from each other at a 5 percent level. These estimates

suggest clearly that increased federal transfers to states through a higher Federal Medical

Assistance Percentage for Medicaid and other investments funded through state budgets

were macroeconomically effective.

Feyrer and Sacerdote (2011) use state- and country-level spending on a variety of

types of stimulus grants (again, excluding tax cuts and some individual-level transfer

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

payments) and find a smaller multiplier of 0.5 to 1.0, although this reflects

heterogeneous estimates including a multiplier of -0.7 to -3.3 for education and police

spending but much higher multipliers like 2 to 2.3 for low-income support and 1.8 for

infrastructure and other grants. Dube, Kaplan, and Zipperer (2015) also use county-level

stimulus spending, finding slightly larger effects than Feyrer and Sacerdote (2011).

Translating state-level multipliers into national multipliers is not straightforward

because of many factors, most notably the spillovers from one state to another and other

general equilibrium effects (Nakamura and Steinsson 2014 and Farhi and Werning 2016).

Chodorow-Reich (forthcoming), however, argues that in many cases the policy-relevant

multiplier should assume no interest rate response and deficit financing, in which case

he shows that state-level cross-sectional multipliers are a rough lower bound on what the

relevant Keynesian national multiplier would be. He summarizes his analysis as finding

that his own results and other cross-sectional results “implies a no-monetary-policy-

response deficit-financed national multiplier of about 1.7 or above. This magnitude falls

at the very upper end of the range found in a recent review article based mostly on time

series evidence (Ramey 2011). Thus, cross-sectional multiplier studies suggest the

national multiplier can be larger than often assumed.”

A few other provisions of the stimulus have also been studied, although most of

these estimates do not meaningfully change the ex-ante parameters originally assumed

when the stimulus passed.

The one case where the evidence suggests a substantially

larger effect than originally assumed is for bonus depreciation. The CEA estimates

assumed a multiplier of 0.1 for business tax provisions, similar to CBO’s low estimates.

Zwick and Mahon (2017), however, studied the actual data and found large effects of

temporarily expanded equipment depreciation on investment, particularly for small

and financially constrained firms. This finding is consistent with some of the thinking

that went into designing these provisions in the first place, which were intended in part

as an interest free loan whose benefit would be related to the cash flow in the initial

years but whose cost was the much lower present value to the government. The original

bonus depreciation, for example, provided $50 billion of tax reductions in the first two

fiscal years as firms shifted depreciation allowances earlier but then most of this

money was recouped in future years, leaving the 10-year cost at only $7 billion. If only

10 percent of the $50 billion was spent as additional investment that would lead to a

multiplier of nearly 1 evaluated against the entire fiscal cost of the measure.

"M.)"FGG\"%)N$2)"$*'"V$S5*R"@/%S"1$+"=)%)"326'5)'"N+"1$%S)% ")2"$0:" CFG!KD"=./"0//S)'"$2"2.)"59?$-2"/*"

-/*369?25/*"$*'",$.9J",.$?5%/J"$*'",0)9%/'"CFG!FD"=./"%)05)'"/*"3)07I%)?/%2)'"36%&)+3J"=52."2.)"7/%9)%"

75*'5*R"0$%R)%"59?$-23"2.$*"2.)"0$2)%:"M.)"59?$-2"/7"2.)"-$3."7/%"-06*S)%3"?%/R%$9"53"')N$2)'"=52."V5$*"$*'"

,675"CFG!FD"$*'"#/)S32%$J"1600)%"$*'"@)32"CFG!LD"75*'5*R"39$00"2/"*)R$25&)")77)-23"=.50)"O%))*J"V)0`)%J"

1$%S)%"$*'"P/W$3"CFG!\D"75*'5*R"$"0$%R)%")77)-2a$02./6R."2.)"% )3 /0625/*"/7"2.53"')N$2)".$3"05220)")77)-2"/*"2.)"

)UI$*2)"%)36023"%)?/%2)'")$%05)%"N)-$63)"-$3.I7/%I-06*S)%3"=$3"/*0+"YK"N5005/*"/7"2.)"/&)%$00"Y!:Z"2%5005/*"

32596063:"_2.)%")&$06$25/*3"/7"3?)-575-"?%/R%$93"5*-06')"^)%R)%J"M6%*)%J"$*'"b=5-S"CFG!\D"75*'5*R"0$%R)"

)77)-23"/7"2.)"<5%32I259)"#/9)N6+)%3"M$U"B%)'52"$*'"B./'/%/=IP)5-.J"B/R05$*)3)J"$*'"($%$N$%N/6*53"

C7/%2.-/95*RD"75*'5*R"39$00)%"0$N/%"9$%S)2")77)-23"/7"2.)"6*)9?0/+9)*2"5*36%$*-)")U2)*35/*3:"

"M.)")U?$*35/*"/7"A)2"_?)%$25*R"Q/33)3"CA_Q3D"=$3"$03/"N$3)'"/*"$"35950$%"0/R5-J"N62">/N%5'R)"CFG![D"

75*'3"2.$2"9/3 2"/7"2. )3) "$' '525/ *$ 0"?$ +9 ) *23 "=) %)"3$ &) '"N +"75%93"%$ 2.) %"2.$ *"3? ) *2a$02./6R."2.53"'/)3"

*/2"%60)"/62"$''525/*$0"5*&)329)*2"5*"7626%)"+)$%3:"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Overall these results are not fully informative about how to design the composition

of stimulus. But they do suggest that the multipliers used by CEA, CBO and others may

have been too low, especially for some items like business tax incentives and possibly

state fiscal relief, and that the actual aggregate impact of the Recovery Act and other

stimulus programs may have been larger than previously estimated.

Evaluating Other Goals for Fiscal Stimulus: Protecting the Most

Vulnerable

In addition to its overall macroeconomic motivation, the Recovery Act and other fiscal

stimulus measures were designed, in the terminology of their creators, to “protect the

most vulnerable.” This category included both long-term poor who were viewed as being

more vulnerable to a downturn as well as people who temporarily fell on hard times by

losing their jobs. In total $289 billion was spent on direct aid to individuals like the

unemployed and those receiving the Supplemental Nutrition Assistance Program

(SNAP), but much of the $836 billion in tax cuts and $167 billion in state and local fiscal

relief also disproportionately benefited the same households.

The labor market did, indeed, disproportionately inflict harm on more vulnerable

populations. From the fourth quarter of 2007 through the fourth quarter of 2010 the

overall unemployment rate rose 4.7 percentage points but the unemployment rates for

those with less education and for African Americans and Hispanics rose much more

sharply, as shown in Figure 4.

Figure 4

The tax and transfer system, both the automatic measures already in the law and the

additions in the Recovery Act and other legislation, did a remarkable job in protecting

households from these job losses. Overall the poverty rate would have risen by 4.8 percentage

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

points from 2007 to 2010 absent these programs, but with them it only rose by at most 0.4

percentage points, and the increase may have been significantly less (Furman 2017).

A more comprehensive picture based of how the Great Recession affected incomes

from 2007 to 2010 is shown in Figure 5, which is based on CBO (2018) data for non-

elderly households.

The largest reductions to market incomes were for the bottom two

quintiles (due to the disproportionate job losses) and top quintile (due to the large declines

in capital income, although some of this is an artifact of the timing of capital gains). After

taking account of taxes and transfers, the bottom quintile for non-elderly households

experienced substantial income gains even in the face of a massive recession. The second

quintile also enjoyed small gains while taxes and transfers absorbed the bulk of the losses

for the third and fourth quintiles. For the highest quintile households, losses were similar

both before and after taxes and transfers. In other words, from the perspective of the fiscal

system low- and moderate-income households got a bailout, not high-income ones.

Figure 5

"M.)3)")3259$2)3"=)%)"'/*)"7/%"<6%9$*"CFG!LD"N+"@59)%")2"$0:"$2"2.)"B/069N5$"1/?60$25/*"P)3)$%-."

B)*2)%:"M.)3)"*69N)%3"63)"2.)"B6%%)*2"1/?60$25/*",6%&)+"'$2$"/*"?)/?0)c3"3)07I%)?/%25*R"/7"=.$2"?6 N 05-"

N)*)7523J"05S)",AE1J"2.$2"2.)+"%)-)5&)':"E3"53"=5')0+"6*')%32//'"C):R:J"V)+)%J"V/SJ"$*'",6005&$*"FGGHDJ"

2.)3)"*69N )%3"36 N32$ *25$00+"6 *' )%32$ 2)"$-26 $0"2$S) I6?"/7"2.)3)"?%/R%$93"$*'"2.63"/&)%32$2)"2.)"5*-%)$3)"

5*"?/&)%2+"= .) *"2. )3)"? %/R %$9 3"$%) ")U?$ *' )' :"

"M)-.*5-$00+"2.53"53"$"-/9N5*$25/*"/7"B^_c3"-$2)R/%5)3"d./63)./0'3"=52."-.50'%)*e"C=.5-."5*-06')3"$"39$00"

*69N)%"/7")0')%0+"./63)./0'3D"$*'"d*/*)0')%0+"./63)./0'3"=52./62"-.50'%)*e:"B^_c3"/&)%$00"*69N)%3J"

5*-06'5*R "2. )")0')%0+J"2)00"$"&)%+"35950$%"32/%+:"

"E00"/7"2.)")3259$2)3"W632"3./="-.$*R)3"5*"5*-/9)"$*'"'/"*/2"%)70)-2"-.$*R)3"5*"=)$02.J"7/%")U$9?0)"2.)"

0/33"/7"./9)3"/%"32/ -S "9$%S)2"=) $02 . :"M. ) "3? ) -57 5-") 3259$2)"7/%"2. ) "2/ ? "] 6 5* 250)"53 "' )? ) * ' )* 2"/ * "2 . )"

2%)$29)*2"/7"N/2."2.)"%)$05`$25/*"/7"-$?52$0"R$5*3"5*-/9)"$*'"2.)"?$+9)*2"/7"-$?52$0"R$5*3"2$U)3"N62"2.)"

75*'5*R"/7"%)'6- 25/* 3"5*"N / 2."9 $ %S) 2"$* '"$ 72)%I2$ U"5*-/ 9 )"./ 0' 3"5*"$* +"-$3):"M. )"/ 2.) %"0)$'5* R"3/ 6%-) "7/%"

-/9?%).)*35&)"5*-/9 )"' $2$a2.)"@/%0'"4*)] 6 $052+" > $2$N $3 )" N $3)'" /*" 9 )2./' 3" ')3-%5N)'" 5*" 15S)22+J",$ )`J"

$*'"b6-9$*"CFG![DJ".$3"3/9)=.$2"'577)%)*2"75*'5*R3"/&)%$00J"5*"?$%2"N)-$63)"/7"2.)5%")U-0635/*"/7"2.)"

5*-/9)"/ * "- $ ? 52$ 0"R$ 5* 3"N 6 2"5* - 06 35/ * "/ 7"2 . )"2$ U )3 "/ * "2. )3 )"R $ 5* 3:"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Evaluating Other Goals for Fiscal Stimulus: Public Investment

As shown in Table 4, the Recovery Act included about $300 billion in public investments

that were motivated not just by increasing short-run GDP but also by other goals like

enhancing long-run growth through more infrastructure or increasing energy efficiency.

The prior and subsequent stimulus was almost entirely tax cuts and relief for individuals

and states. This paper does not evaluate these goals.

Table 4

One goal of the Recovery Act was to spend investment funding as quickly as

possible through so-called shovel ready projects. It was well understood that some

measures, like high-speed rail and electronic medical records, would take some time to

deliver their benefits but the highway program was designed to spend out relatively

quickly. In fact, the actual spendout for the highway program was even faster than

initially predicted—with 80 percent of the funding actually spent by 2011, well above the

projection of 55 percent by that year (Figure 6).

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Figure 6

IV. Alternatives That Were Not Taken

In putting together the Recovery Act and subsequent responses a number of different

alternatives were contemplated. Some were rejected by policymakers in the

Administration or failed to pass Congress. The following describes three of those

alternatives, the thinking that went into them, and potential considerations for future

fiscal stimulus.

New Jobs Tax Credit

During the campaign, Obama proposed a $3,000 tax credit for employers for each net new

additional job added in 2009 and 2010 (Obama-Biden campaign 2008). The transition

team further refined the idea and Obama signed off on proposing it as part of the Recovery

Act. Congress rejected the idea because lawmakers were skeptical of the efficacy of a

measure designed to add jobs at a time when firms were rapidly shedding jobs.

Obama proposed a retooled version in January 2010 that would have provided a tax

credit of $5,000 per net new job and a bonus for firms that raised total wages, with both

capped at $500,000 in credits per firm (White House 2010). The credits would have

been paid out quarterly and applied against payroll taxes so that even firms with tax

losses would have been eligible. This proposal evolved into the Hiring Incentives to

Restore Employment (HIRE) Act passed by Congress in March 2010, but it was

ultimately very different because it was a tax credit for hiring the long-term unemployed

rather than an incremental tax credit for adding net new jobs.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

The theory behind the new jobs tax credit was that it would have the same

Keynesian boost as any tax cut while providing additional bang-for-the-buck in terms of

higher employment, effectively raising the number of jobs per unit of GDP. This

theoretical intuition was bolstered by academic evaluations of the 1977-78 New Jobs Tax

Credit, which found that it was effective at promoting hiring (Perloff and Wachter 1979,

Bishop 1981). Later in 2009 CEA and Treasury did extensive modelling and found that a

new jobs tax credit would cost about $20,000 per net new job added. About 90 percent

of the subsidy would go to jobs that would have been created anyway but the other 10

percent would have provided a net increase in new jobs. Even with this unbalanced ratio,

the cost per job created would have been considerably less than CEA’s (2014) implied

employment estimates of around $125,000 per job-year for the Recovery Act as a whole.

State Sales Tax Holiday

The transition team considered a state sales tax holiday as a way to shift consumption

forward by encouraging households to take advantage of temporarily lower prices.

Specifically, the transition team contemplated proposing something like a $250 billion

fund that would be available to states carrying out certain sales tax reductions. States

without sales taxes would have been allowed to use the money for income tax cuts. The

idea was ultimately not proposed to Congress.

The principal argument for the proposal was to take advantage of intertemporal

substitution, effectively setting a very negative real interest rate for consumer purchases

that would make it more attractive to make purchases in the short run before sales taxes

went up again. This would have acted as an additional stimulus on top of the normal

Keynesian response to the additional demand associated with tax cuts. The federal role

would have been simply to approve state plans and disburse money according to a

specified formula. States’ prior experience with sales tax holidays had demonstrated that

such a plan would have been relatively straightforward to administer. Concerns with the

proposal included worries that it could inhibit spending in the run-up to the tax or hurt

future demand by pulling spending forward. In retrospect, given how protracted the

downturn was, it was probably economically wise that this proposal was left on the

cutting room floor. Nonetheless, it has a number of benefits that merit consideration by

policymakers in future short, sharp downturns.

Expiring Debit Cards

A final idea that received substantial consideration during the transition but was

never proposed was to deliver rebates on a debit card. The debit card could have been set

with a use-it-or-lose-it feature so that, for example, anyone who did not spend all of the

money by the end of 2009 would have lost the remaining balance. The goal of the

proposal was to make the refund more salient than just getting a check or having fewer

taxes withheld, while also helping to ensure that people spent the money more quickly,

raising the short-term multiplier relative to traditional tax credits. Extensive work was

done with Treasury staff who believed that it would have been administratively possible

to provide such cards for all Americans at essentially no transaction cost to the federal

government because card issuers would bid for a contract that would earn them

interchange fees and float interest. Ultimately, however, it was considered too

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

administratively risky to design and launch a massive consumer-facing program in a

short period of time. Further effort and contingency planning on this idea may be

warranted because of its potential upside relative to conventional tax cuts.

V. Lessons for the Future

The experience with the 2008-12 fiscal stimulus and economic research that has been

conducted in the years since then provides six lessons for future fiscal stimulus.

LESSON 1: DISCRETIONARY FISCAL STIMULUS CAN BE PARTICULARLY

EFFECTIVE, ESPECIALLY AT THE ZERO LOWER BOUND. Prior to 2008

economists were generally skeptical of discretionary fiscal stimulus, concerned about

potentially large lags in recognizing the problem, passing a legislative response,

implementing the legislation, and having it take effect. In general, monetary policy was

considered superior on all of these counts and regarded as the first and potentially only

line of defense in downturns.

The 2008-12 experience suggests that these fears were overblown. Policymakers

proposed a fiscal response in January 2008, essentially just as the recession began

according to subsequent analysis by the National Bureau of Economic Research (NBER)

business cycle dating committee. Initial legislation was passed a month later and as the

situation became more severe at the end of 2008 the Recovery Act was passed within a

month of Obama’s inauguration. Many of the provisions were implemented quickly, with

electronic refunds first sent out in April 2008, two months after the first stimulus was

passed. Similarly, the reductions in withholding included in the Recovery Act were

implemented within months. Moreover, the evidence suggests that fiscal measures act on

the economy more quickly than with monetary policy.

Looking ahead, fiscal policy may become an even more important tool for aggregate

demand management. Research following the Great Recession has helped to build a

“new view” of fiscal policy—which in many ways is actually a rediscovery of the old

Keynesian liquidity trap views (see Furman 2016 for a summary). In particular, the new

view holds that if the equilibrium interest rate has fallen, then monetary policymakers

will more often hit the effective lower bound for interest rates—limiting conventional

monetary policy and increasing the importance of fiscal policy.

Moreover, fiscal policy may be particularly effective when interest rates are at the

lower bound. Traditional concerns about crowding out through higher interest rates may

be superfluous when interest rates are stuck at the effective lower bound: even with

additional demand the desired interest rate will not rise. Fiscal policy could also lead to

“crowding in,” either through an accelerator mechanism that increases growth rates and

thus investment growth or by raising expected inflation and thus lowering real interest

rates (Hall 2009; Christiano, Eichenbaum, and Rebelo 2011; Woodford 2011).

Concerns about fiscal space may be overstated in a highly depressed economy with

interest rates at the effective lower bound because fiscal expansions may raise GDP by

more than they raise debt—thus reducing the debt-to-GDP ratio. This result has been

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

found in a variety of settings, including the Fed’s main macroeconomic model FRB-US

(Reifschneider and Summers as reported in DeLong, Summers, and Ball 2014), the

Organisation of Economic Co-operation and Development’s (2016) NiGEM and FM

models, the International Monetary Fund’s modelling (Gaspar, Obstfeld and Sahay

2016), simulations by DeLong and Summers (2012) and regression-based estimates of

past fiscal stimulus by Auerbach and Gorodnichenko (2017). On the other hand, there is

still some reason for caution: Romer and Romer (2017) found that countries that go into

financial crises with higher debt have smaller fiscal responses and worse macroeconomic

outcomes. (It is not clear if their result reflects high-debt countries having little economic

space for stimulus or making political mistakes about too little stimulus.)

LESSON 2: THE POLITICAL SYSTEM MAY TIRE OF FISCAL STIMULUS

PREMATURELY, SO TRIGGERS AND OTHER AUTOMATIC MECHANISMS

MAY BE WARRANTED—INCLUDING STRONGER AUTOMATIC

STABILIZERS. The transition team thought that if the initial stimulus was too small or

the economy worsened further it would not be very difficult to persuade Congress to

provide more fiscal stimulus. This view did not seem unreasonable at the time: in the

face of a much milder recession in 2001 Congress passed bills it described as fiscal

stimulus in 2001, 2002, 2003, and again in 2004, nearly three full years after the end of

the recession. In a time of widely appreciated national economic emergency it should

have been even easier to get more stimulus, especially given that lawmakers would

normally have wanted to be seen doing something closer to the 2010 election.

This view proved less wrong than popularly understood when you look at the more

than 13 subsequent bills that included additional stimulus. But this stimulus still fell well

short of the Administration’s subsequent requests and, in general, stimulus was removed

well before it was economically sensible to do so—creating a fiscal drag starting in 2011

as the total magnitude of stimulus started to phase down.

The difficulty stems from three challenges. The first is that, paradoxically, the worse-

than-expected macroeconomic outcomes reduced the desire to take more

macroeconomic measures. Even though the bulk of the unexpected deterioration of the

economy happened by early-to-mid 2009, before the bulk of the Recovery Act went into

effect, this was viewed by some as evidence that the law had not worked, making future

stimulus counterproductive. The second is concerns over the deficit—which nearly

reached 10 percent of GDP in 2009, putting the debt on a course to eventually more than

double as a share of GDP. The final reason was partisan politics, and in particular

Congressional Republicans not wanting to cooperate with the Administration in passing

more fiscal measures.

One way to address the political failures in the future would be to make the initial

stimulus package contingent on economic outcomes. For example, the Recovery Act

could have included an annual tax credit that would be in effect every year the

unemployment rate was above 7 percent. Such a trigger would have delivered more

consistent, predictable, and potentially larger fiscal support in subsequent years.

"E*"59?/%2$*2"-$&)$2"')?)*'5*R"/*"2.)"?/0525-$0"-/*2) U2f"57"2 . )"P ) -/ & )% +"E - 2". $ ' "5* -06 ' ) ' "36 - . "

-/*25*R)*2"9)$36%)3"=.50)"S))?5*R"=52.5*"2.)"Y\GG"N5005/*"N6'R)2$%+"05952"3)2"N+"B/*R%)33"2.)*"52"=/60'"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Even better would be to expand the automatic stabilizers permanently. In general,

the United States has smaller automatic stabilizers than most other advanced economies,

largely because their magnitude is highly correlated with the size of government—and

the United States has a generally smaller government.

Such automatic stabilizers could be measures that automatically expand (like

unemployment insurance and SNAP). In fact, the health insurance tax credits and

Medicaid expansion in the Affordable Care Act (ACA) will increase automatic stabilizers

in future years. Alternatively, automatic stabilizers could be contingent on national or

state economic data, perhaps full federal funding for lengthening unemployment

insurance benefits or higher Medicaid matching rates in states with high or rising

unemployment rates.

LESSON 3: IMPORTANCE OF MAKING SURE THAT STATES AND

LOCALITIES DO NOT UNDERCUT THE FEDERAL FISCAL STIMULUS. As the

federal government was expanding fiscal support for the economy in the wake of the

Great Recession, states and localities were cutting spending and thus undoing a

meaningful portion of this stimulus. This was unusual. If State and local spending had

followed the average pro-cyclical pattern of six of the previous seven recoveries and

everything else was the same, the GDP growth rate would have been 0.6 percentage

point per year higher in the five years following the trough.

The reductions in state and local purchases may have been specific to this recession

because of the substantial and lagged effects of reduced housing prices on property tax

revenue. Nevertheless, the causes of the large departure from previous experience are

not fully understood and the 2001 experience was also worse than the historical

experience. That raises the concern that, in the future, state and local fiscal policy could

once again run against federal efforts or be insufficiently supportive of them. Moreover,

even optimal subnational policy will be insufficiently fiscally supportive in a recession

because states and localities do not take into account the substantial benefits of their

stimulus measures on other states.

This highlights the importance of steps to support state and local spending in future

recessions and their aftermath. One approach would be to legislate discretionary fiscal

relief, either general fiscal relief or labelled for a specific purpose—with the

understanding that it would be fungible. Another and potentially better approach would

be to make state fiscal relief an automatic stabilizer that is triggered when the national

and/or state unemployment rates rose.

LESSON 4: THE IMPORTANCE OF MAINTAINING THE CONFIDENCE OF

TAXPAYERS WHEN UNDERTAKING A LARGE, RELATIVELY RAPID

SPENDOUT OF FUNDS. The Recovery Act included unprecedented transparency and

accountability measures, including an independent Recovery Accountability and

Transparency Board, regular and timely reporting of all contracts and other information,

.$&)"$-26$00+"N))*"smaller"6?7%/*2:"M.)"%)$3/*"53"2.$2"B^_"$*'";BM"3-/%)"0)R530$25/*"?%/N$N505325-$00+"$*'"

)&)*"$"-.$*-)"2.$2"36-."$"2%5RR)%"=/60'"32500"N)"5*")77)-2"5*J"3$+J"FG!Z"=/60'"-/6*2"2/=$%'3"2.)"-/32"

%)]65%5*R"/2.)%"?%/ &535/ *3"2/ "N)"39$00)%"2/"752"5*"2.)"Y\GG"N5005/* "059 52:"

"8U-06')3"2.)"!H\G"%)-)335/*"N)-$63)"/7"/&)%0$?"=52."2.)"!H\!g!H\F"%)-)335/*:"

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

and an easily accessible website to share all of this public information. As a result, only a

miniscule percentage of the contracts were considered fraudulent and even Republican

Congressman Darrel Issa, a frequent critic of the Administration, acknowledged that the

new recipient reporting standards were “the key to the success of the Recovery

Accountability and Transparency Board in catching and preventing fraud, waste, and

abuse in stimulus spending.” (House Oversight Committee 2011).

LESSON 5: GROWING EVIDENCE THAT TAX CUTS MAY BE MORE

EFFECTIVE THAN PREVIOUSLY APPRECIATED, POTENTIALLY IF

TARGETED AT LOWER-INCOME HOUSEHOLD AND ALSO AT

ACCELERATED DEPRECIATION. Beyond the specific evidence that emerged from

studying the Recovery Act itself, economic research focused on a longer period of time is

finding growing evidence that tax cuts, in particular, may have a higher multiplier than

previously appreciated. Romer and Romer (2010) use a narrative method to identify

exogenous tax shocks and find that the impact of a tax cut worth 1 percent of GDP grows

over time to about 3 percent of GDP after 10 quarters, an effect that may reflect

expanded demand in a traditional Keynesian manner or increased supply through the

incentives to work. Barro and Redlick (2011), Mertens and Ravn (2013) and Mertens and

Montiel-Olea (2017) find impulse responses from tax cuts that are similarly large. Zidar

(2017) finds large responses as well but also finds substantial heterogeneity, with a much

larger multiplier for tax cuts for low-income households than high-income households.

Most of these papers are focused on tax multipliers but they generally find larger

effects than the papers focused just on spending multipliers (e.g., Ramey 2011 and

Ramey and Zubairy 2018). In some cases, the papers provide estimates for both tax and

spending multipliers on a consistent basis (e.g., Carlino and Inman 2016, Andrés, Boscá,

and Ferri 2016 and Ramey 2018) and also find larger tax multipliers than spending

multipliers—in some cases by substantial magnitudes.

Zwick and Mahon (2017), discussed above, also finds very large responses to

accelerated depreciation, especially compared to the cost to the government. These

responses may be specific to a financial crisis, however, because the cash flow

improvement was especially potent and the present value cost to the government was

especially low given modest federal borrowing rates.

All of this said, the general evidence on generic tax cuts—like any macroeconomic

evidence—is not definitive and in some cases is puzzlingly larger than one would expect

based on the economic theory of either Keynesian multipliers or possible supply-side

responses. Moreover, it is possible that both spending multipliers and tax multipliers

have been underestimated in the past.

Overall, however, the evidence suggests some upward revision in tax multipliers

relative to what was believed in 2009 and 2010, especially for tax cuts for lower-income

households and for businesses.

LESSON 6: SUGGESTIVE SUPPORT FOR THE NOTION THAT IT WOULD BE

BENEFICIAL TO INCREASE THE NUMBER OF JOBS PER UNIT OF GDP. The

Great Recession had a particularly severe impact on the labor market even conditional

on the large increase in the unemployment rate, including sustained record high long-

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

term unemployment rates, a spike in people working involuntarily part time, and a large

reduction in the labor force participation rate. Similar labor market dysfunctions

followed the much shallower 2001 recession as well.

In many ways the labor market experience was much worse in the United States

than in other advanced economies. Germany experienced a similar decline in GDP

without the large increase in unemployment that the United States had, as shown in

Figures 7a and 7b. Instead, hours worked were reduced in Germany, spreading the

employment decline across a much larger number of people.

Figure 7a Figure 7b

While Germany followed some different policies and has very different institutional

arrangements than the United States, their experience highlights the potential for

increasing the number of jobs conditional on a given amount of GDP—not just focusing

on maximizing GDP. One way to do this is through a new jobs credit, described above,

which CEA estimated would add one job year for each roughly $20,000 spent—giving it

a considerably higher employment impact than generic stimulus. Another possibility

would be to encourage more states to adopt the type of short-term compensation found

in Germany that provides unemployment insurance not just for people fired from jobs

but also for people who have their hours involuntarily reduced.

VI. Conclusion

The stimulus that started in 2008, was greatly expanded in 2009, and somewhat

expanded and extended further in the several years thereafter, was an integral part of

the overall macroeconomic response to the financial crisis. Absent these measures the

recession would have been much deeper and more prolonged—potentially even more

so than conventional models indicate because of the possibilities of self-fulfilling

vicious cycles and persistent losses in output. This stimulus acted in a synergistic

manner with monetary and financial policy, helping the U.S. economy fare better than

many historical precedents and outperforming other countries in the wake of the Great

Recession (CEA 2017).

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Nonetheless, there were significant shortcomings in the fiscal response, largely the

result of the political difficulty of convincing Congress to pass sufficiently large stimulus

bills starting with the Recovery Act in 2009 and the failure to win support for additional

stimulus that was both large enough and weighted more towards public investment.

Much of the stimulus was designed with reasonably good aggregate macroeconomic

models, models that have been further corroborated by the experience of the 2008-12

stimulus bills. Knowledge of the different impacts of different types of measures,

however, was much less certain than the point estimates published by different

macroeconomic forecasters might lead one to believe. We know only a little more today

and it is likely that policymakers will still be flying blinder than should be the case when

designing the specific composition of the next fiscal stimulus. But, the evidence for the

overall importance of fiscal stimulus is stronger than ever.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Appendix

Some of the larger or more notable items in each of these categories along with their cost

from 2008-2012 were:

o! Individual Tax Cuts

o! Making Work Pay ($112 billion). A tax credit that was generally $400 for

individuals and $800 for married couples, phased out for higher-income

households. It was originally passed as part of the Recovery Act and lasted

for 2009 and 2010.

o! Payroll tax cut ($207 billion). A 2 percentage point reduction in Social

Security payroll taxes, with the cost reimbursed to the Social Security

Trust Fund. Originally passed for 2011 and extended through the end of

2012.

o! Tax credits for low- and moderate-income households with children

($41 billion). The Recovery Act expanded the Earned Income Tax Credit

for married couples and households with three or more children and also

increased the refundable portion of the child tax credit. These provisions

were subsequently made permanent.

o! Business Tax Incentives

o! Bonus depreciation and expensing ($180 billion). Originally allowed

businesses to deduct 50 percent of the cost of their equipment

investments up front and was expanded to 100 percent expensing of

equipment investment from September 9, 2010 through the end of 2011.

o! Discharge of business indebtedness ($43 billion). Allows certain business

repurchasing specific types of debt to pay taxes on the cancellation of that

debt income over five years beginning in 2014.

o! Tax credits related to clean energy and energy efficiency ($12 billion). The

Recovery Act included a number of tax credits for clean energy including

extending and expanding the energy tax credit.

o! Aid to Directly Impacted Individuals

o! Unemployment insurance expansions ($240 billion). Full federal

financing for benefits for up to 99 weeks of unemployment in high

unemployment States, initially also included temporarily increased

benefit amounts. In addition, the Recovery Act and subsequent legislation

included unemployment insurance reforms.

o! Increased Supplemental Nutrition Assistance Program (SNAP)

($38 billion). A temporary bump up to SNAP benefits that started in the

Recovery Act, was partially rescinded in two subsequent pieces of

legislation, and was phased out by November 2013.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

o! Temporary Assistance to Needy Families (TANF) Emergency Fund ($5

billion). A temporary increase to TANF that could be used by States for

purposes including public employment and hiring subsidies.

o! State Fiscal Relief

o! Increased Federal Medical Assistance Percentage (FMAP) ($100 billion).

Increased Federal matching for State Medicaid programs originally

passed as part of the Recovery Act and extended in the summer of 2010.

o! State educational assistance ($63 billion). Federal subsidies for States and

localities to fund efforts to prevent teacher layoffs. Included the “Race to

the Top” program, a competitive grant for States undertaking major

reforms.

o! Public Investment Outlays

o! Increased investments in the highway and rail program ($28 billion).

Largely formula funding for the traditional surface transportation

programs.

o! Transportation Investment Generating Economic Recovery (TIGER)

Grants ($3 billion). Competitive grants available to States and localities

for a variety of transportation purposes.

o! High-speed Rail ($2 billion). Grants to support State high-speed and

inter-city rail projects passed as part of the Recovery Act.

o! Health information technology ($15 billion). Including, incentive

payments to medical providers to adopt and meaningfully use electronic

medical records, with penalties for those that took the subsidies without

meaningfully using them.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

References

Akerlof, George, et al. 2008. “Economists’ Letter to Congress.” November 19.

(cepr.net/documents/publications/Economists_letter_2008_11_19.pdf).

Andrés, Javier, José E. Boscá, and Javier Ferri. 2016. “Instruments, Rules, and Household

Debt: the Effects of Fiscal Policy. Oxford Economic Papers 68(2): 419-443.

Auerbach, Alan J. and Yuriy Gorodnichenko. 2017. “Fiscal Stimulus and Fiscal

Sustainability.” NBER Working Paper 23789. Cambridge, MA: National Bureau of

Economic Research.

Barro, Robert, and Charles J. Redlick. 2011. “Macroeconomic Effects From Government

Purchases and Taxes.” Quarterly Journal of Economics 126(1): 51-102.

Berger, David, Nicholas Turner, and Eric Zwick. 2018. “Stimulating Housing Markets.”

NBER Working Paper 22903. Cambridge, MA: National Bureau of Economic

Research.

Bishop, John H. 1981. “Employment in Construction and Distribution Industries: The

Impact of the New Jobs Tax Credit.” In Studies in Labor Markets, ed. S. Rosen.

Chicago: The University of Chicago Press.

Blinder, Alan S., and Mark Zandi. 2015. “The Financial Crisis: Lessons for the Next One.”

Policy Futures Paper. Washington: Center on Budget and Policy Priorities.

Carlino, Gerald, and Robert P. Inman. 2016. “Fiscal Stimulus in Economic Unions: What

Role for States?” Tax Policy and the Economy 30(1):1-50.

Chodorow-Reich, Gabriel. Forthcoming. "Geographic Cross-Sectional Fiscal Multipliers:

What Have We Learned?" American Economic Journal: Policy.

Chodorow-Reich, Gabriel, John Coglianese, and Loukas Karabarbounis. Forthcoming.

“The Macro Effects of Unemployment Benefit Extensions: A Measurement Error

Approach.” Quarterly Journal of Economics.

(https://scholar.harvard.edu/files/chodorow-reich/files/ui_macro.pdf).

Chodorow-Reich, Gabriel, Laura Feiveson, Zachary Liscow, and William Gui Woolston

(2012). “Does State Fiscal Relief During Recessions Increase Employment? Evidence

from the American Recovery and Reinvestment Act.” American Economic Journal:

Policy 4(3): 118-145.

Christiano, Lawrence, Martin Eichenbaum, and Sergio Rebelo. 2011. “When is the

Government Spending Multiplier Large?” Journal of Political Economy 119(1): 78-121.

Congressional Budget Office (CBO). 2018. “The Distribution of Household Income, 2014.”

Report 53597.

Conley, Timothy G., and Bill Dupor. 2013. “The American Recovery and Reinvestment

Act: Solely a Government Jobs Program?” Journal of Monetary Economics 60(5):

535-549.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Council of Economic Advisers (CEA). 2009. “Estimates of Job Creation from the

American Recovery and Reinvestment Act of 2009. Report. May.

_____. 2014. “Chapter 3: The Economic Impact of the American Recovery and

Reinvestment Act Five Years Later.” 2014 Economic Report of the President.

Washington: United States Government Printing Office.

_____. 2017. 2017 Economic Report of the President. Washington: United States

Government Printing Office.

DeLong, J. Bradford and Lawrence H. Summers. 2012. “Fiscal Policy in a Depressed

Economy.” Brookings Papers on Economic Activity 43(1): 233-297.

DeLong, Brad, Larry Summers, and Laurence Ball. 2014. “Fiscal Policy and Full

Employment.” Center on Budget and Policy Priorities.

Dobridge, Christine. 2016. “Fiscal Stimulus and Firms: A Tale of Two Recessions.”

Finance and Economics Discussion Series 2016-013. Washington: Board of

Governors of the Federal Reserve System.

Dube, Arindrajit, Ethan Kaplan, and Ben Zipperer. 2015. “Excess Capacity and

Heterogeneity in the Fiscal Multiplier: Evidence from the Obama Stimulus Package.

Working paper. (http://econweb.umd.edu/~kaplan/stimulus_effects.pdf).

Dupor, Bill, and Peter B. McCrory. 2017. “A Cup Runneth Over: Fiscal Policy Spillovers

from the 2009 Recovery Act.” The Economic Journal 128(611): 1476-1508.

Dupor, Bill, and M. Saif Mehkari. 2016. “The 2009 Recovery Act: Stimulus at the

Extensive and Intensive Labor Margins.” European Economic Review 85: 208-228.

Elmendorf, Douglas W., and Jason Furman. 2008. “If, When, How: A Primer on Fiscal

Stimulus.” Washington: Brookings Institution.

Farhi, Emmanuel, and Ivan Werning, 2016. “Chapter 31—Fiscal Multipliers: Liquidity

Traps and Currency Unions.” Handbook of Macroeconomics 2: 2417-2492.

Feldstein, Martin. 2007. “How to Avert Recession.” Wall Street Journal. December 5.

(https://www.wsj.com/articles/SB119682440917514075).

Feyrer, James, and Bruce Sacerdote. 2011. “Did the Stimulus Stimulate? Real Time

Estimates of the Effects of the American Recovery and Reinvestment Act.” NBER

Working Paper 16759. Cambridge, MA: National Bureau of Economic Research.

Furman, Jason. 2016. “The New View of Fiscal Policy and Its Application.” Remarks at

Global Implications of Europe’s Redesign Conference. New York. October 5.

_____. 2017. “Reducing Poverty: The Progress We Have Made and the Path Forward.”

Remarks at the Center on Budget and Policy Priorities. Washington. January 17.

(https://obamawhitehouse.archives.gov/sites/default/files/page/files/20170117_fur

man_center_on_budget_poverty_cea.pdf).

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Gaspar, Vitor, Maurice Obstfeld, and Ratna Sahay. 2016. “Macroeconomic Management

When Policy Space is Constrained: A Comprehensive, Consistent and Coordinated

Approach to Economic Policy.” IMF Staff Discussion Note.

Green, Daniel, Brian T. Melzer, Jonathan A. Parker, and Arcenis Rojas. 2018.

“Accelerator or Brake? Cash for Clunkers, Household Liquidity, and Aggregate

Demand.” NBER Working Paper 22878. Cambridge, MA: National Bureau of

Economic Research.

Grunwald, Michael. 2012. The New New Deal: The Hidden Story of Change in the

Obama Era. New York: Simon & Schuster.

Hall, Robert E. 2009. “By How Much Does GDP Rise if the Government Buys More

Output?” Brookings Papers on Economic Activity 40(2): 183-249.

Hoekstra, Mark, Steven L. Puller, and Jeremy West. 2017. “Cash for Corollas: When

Stimulus Reduces Spending.” American Economic Journal: Applied Economics

9(3): 1-35.

House Oversight Committee. 2011. “Issa Introduces Sweeping Open Government,

Spending Transparency Reforms.” Press release. June 13.

Internal Revenue Service. 2008. “Economic Stimulus Payments on the Way; Some

People Will See Direct Deposit Payments Today.” News Release.

(https://www.irs.gov/newsroom/economic-stimulus-payments-on-the-way-some-

people-will-see-direct-deposit-payments-today).

Mertens, Karel, and José L. Montiel Olea. 2017. “Marginal Tax Rates and Income: New

Time Series Evidence.” NBER Working Paper No. 19171. Cambridge, MA: National

Bureau of Economic Research.

Mertens, Karel, and Morten O. Ravn. 2013. “The Dynamic Effects of Personal and

Corporate Income Tax Changes in the United States.” American Economic Review

103(4): 1212-47.

Meyer, Bruce D. Wallace K. C. Mok, and James X. Sullivan. 2009. “The Under-Reporting

of Transfers in Household Surveys: Its Nature and Consequences.” NBER Working

Paper 15181. Cambridge, MA: National Bureau of Economic Research.

Mian, Atif, and Amir Sufi. 2012. “The Effects of Fiscal Stimulus: Evidence from the 2009

Cash for Clunkers Program." Quarterly Journal of Economics 127(3): 1107-1142.

Nakamura, Emi, and Jón Steinsson. 2014. “Fiscal Stimulus in a Monetary Union:

Evidence from US Regions.” American Economic Review 104(3): 753-792.

Obama, Barack. 2009a. “Address at George Mason University in Fairfax, Virginia.”

January 8. Online by Gerhard Peters and John T. Woolley, The American Presidency

Project. (http://www.presidency.ucsb.edu/ws/?pid=85361).

_____. 2009b. “Transcript: Obama Remarks At Stimulus Signing.” February 17. Online

by Brian Montopoli, CBS News. (https://www.cbsnews.com/news/transcript-

obama-remarks-at-stimulus-signing/).

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Obama-Biden campaign 2008. 2008. “Barack Obama and Joe Biden: A Rescue Plan for

the Middle Class.” https://www.scribd.com/document/6523359/Barack-Obama-

and-Joe-Biden-s-Rescue-Plan-for-the-Middle-Class.

Organisation for Economic Co-Operation and Development (OECD). 2016. “Chapter 2:

Using the Fiscal Levers to Escape the Low-Growth Trap.” In OECD Economic

Outlook, Volume 2016, Issue 2. Paris: OECD Publishing.

Parker, Jonathan A., Nicholas S. Souleles, David S. Johnson, and Robert McClelland.

2013. “Consumer Spending and the Economic Stimulus Payments of 2008.”

American Economic Review 103(6): 2530-2553.

Paulson, Henry. 2008. “Press Briefing by Treasury Secretary Henry Paulson and

Chairman of the Council of Economic Advisors Ed Lazear.” Washington. January 18.

Online by Gerhard Peters and John T. Woolley, The American Presidency Project.

(http://www.presidency.ucsb.edu/ws/?pid=76285).

Perloff, Jeffrey M. and Michael L. Wachter. 1979. “The New Jobs Tax Credit: An

Evaluation of the 1977-78 Wage Subsidy Program.” American Economic Review

69(2): 173-179.

Piketty, Thomas, Emmanuel Saez, and Gabriel Zucman. 2016. “Distributional National

Accounts: Methods and Estimates for the United States Data Appendix.” WID.world

Working Paper Series 2016/4. December. (https://wid.world/document/t-piketty-e-

saez-g-zucman-data-appendix-to-distributional-national-accounts-methods-and-

estimates-for-the-united-states-2016/).

Ramey, Valerie A. 2011. “Can Government Purchases Stimulate the Economy?” Journal

of Economic Literature 49(3): 673-685.

Ramey, Valerie A. 2018. “Ten Years After the Financial Crisis: What Have We Learned

from the Renaissance in Fiscal Research,” Manuscript prepared for Prepared for the

NBER Conference “Global Financial Crisis @10.

Ramey, Valerie A., and Sarah Zubairy. 2018. “Government Spending Multipliers in Good

Times and in Bad: Evidence from US Historical Data.” Journal of Political Economy

126(2): 850-901.

Romer, Christina D., and David H. Romer. 2010. “The Macroeconomic Effects of Tax

Changes: Estimates Based on a New Measure of Fiscal Shocks.” American Economic

Review 100(3): 763-801.

_____. 2017. “Why Some Times Are Different: Macroeconomic Policy and the

Aftermath of Financial Crises.” NBER Working Paper 23931. Cambridge, MA:

National Bureau of Economic Research.

Sahm, Claudia R., Matthew D. Shapiro, and Joel Slemrod. 2012. “Check in the Mail or

More in the Paycheck: Does the Effectiveness of Fiscal Stimulus Depend on How it is

Delivered?” American Economic Journal: Economic Policy 4(3): 216-250.

PRELIMINARY DISCUSSION DRAFT September 11–12, 2018

Stolberg, Sheryl Gay. 2009. “Signing Stimulus, Obama Doesn’t Rule Out More.” The New

York Times. February 17. (https://www.nytimes.com/2009/02/18/us/politics/18web-

stim.html).

Summers, Lawrence. 2007. “Risks of Recession, Prospects for Policy.” Remarks at the

Brookings Institution: “State of the U.S. Economy”. Washington. December 19.

(https://www.brookings.edu/wp-content/uploads/2012/04/20071219_summers.pdf).

The White House. 2010. “Small Business Jobs and Wages Tax Cut.” Fact Sheet.

https://obamawhitehouse.archives.gov/sites/default/files/rss_viewer/fact_sheet_s

mall_business_jobs_wages_tax_cut.pdf.

Wilson, Daniel J. 2012. “Fiscal Spending Jobs Multipliers: Evidence from the 2009

American Recovery and Reinvestment Act.” American Economic Journal: Economic

Policy 4(3): 251-282.

Woodford, Michael. 2011. “Simple Analytics of the Government Expenditure Multiplier.”

American Economic Journal: Macroeconomics 3(1): 1-35.

Zidar, Owen M. 2017. “Tax Cuts for Whom? Heterogeneous Effects of Income Tax

Changes on Growth and Employment.” NBER Working Paper 21035. Cambridge,

MA: National Bureau of Economic Research.

Zwick, Eric, and James Mahon. 2017. “Tax Policy and Heterogeneous Investment

Behavior.” American Economic Review 107(1): 217-248.