What Do Fiscal Multipliers Say About Obama’s Stimulus Plan?

What Do Fiscal Multipliers Say About Obama’s Stimulus Plan?

By Shawkat Hammoudeh

Shawkat Hammoudeh is Professor of Economics and International Business at Drexel University. He can be reached at hammousm@drexel.edu.

There is currently a debate raging among first-rate economists and among famous politicians on the extent of the effectiveness of the $820bn fiscal package suggested by President Barack Obama to stimulate the US economy. This package includes a $550bn increase in spending for state and federal government projects and $270bn in tax cuts over two years, with the aim of creating 3-4mn new jobs. While the debate involves both the spending hike and tax cut components, the package is viewed particularly by some conservative economists and Republican politicians to be unreasonably skewed towards more government spending less tax cuts.

Government Spending Multiplier

The essence of the debate concentrates basically on the values of government spending and tax multipliers because these are necessary for the effectiveness of the package. Some conservative economists use the GDP identity:

GDP = Consumption + Investment + Government Spending + Exports – Imports

to show that increases in government spending must come from transferring savings from private spending (consumption and investment) to government spending, thus deeming the package in their opinion to be unsuccessful. In this case, one unit increase in government spending results in one unit decrease in private spending, leaving GDP unchanged. In the language of the multipliers, the government multiplier according to those conservative economists, including John Cochrane for example, is zero. Suzan Woodward and Robert Hall argue that this multiplier is one. This means a unit hike in government spending produces a unit increase in GDP. In terms of Obama’s spending figures, this means the government spending hike will increase GDP by $550bn over two years, which is equal to the total spending component in the package.¹  On the other hand, the first-rate conservative economist Robert Barro asserted in an op-ed article in the Wall Street Journal that this multiplier is less than one. He calculated the military spending multiplier to be 0.8 for the period 1943-44, implying that the government spending partially crowds out private spending and results in less than one unit increase in GDP for every unit increase in military spending.²  In other words, some of the government spending hikes will be wasted. Prof Barro contends that the peacetime multiplier is less than 0.8. But he neglects to consider that private spending in the form of house construction, production of consumer stables, cars etc was rationed during the war time, which should naturally give rise to a standard multiplier that is higher than the 0.8 military multiplier. He used his estimated static multiplier to argue against the effectiveness of the Obama spending plan, including government spending on long-run infrastructure projects.

It is true that infrastructure projects take time to be implemented, and in that aspect they are late in creating jobs and stimulating the economy, but they bear sweet fruits in terms of increasing economic growth and efficiency in the long run, if they are not ‘sugar-high’ projects. In the multiplier language, this means that the infrastructure multiplier is low and slow in the short-run but it should be higher than one in the long run, particularly if the economy is in a recession with a liquidity trap and hardly producing crowding out. In this case, not all resources in the economy are in use, and thus the crowding out of private spending by government spending is minimal. Additionally, during the spending process and the consequent increases in GDP, the economy creates money which should not come from existing savings. Therefore, the government spending multiplier is greater than one in a severe recession with a liquidity trap and idle resources like the one we are in now, and not zero as John Cochrane contends. That is, a one unit increase in government spending leads to more than one unit increase in GDP. In the case of Obama’s fiscal package, a $550bn increase in government spending, which also includes funds that will immediately go to the pockets of consumers, should lead to more than that amount in GDP increase.

Spending Multiplier And Job Creation

Some economists (eg the 2008 Noble prize winner Paul Krugman and Obama’s economists) estimate the spending multiplier to be about 1.5, implying that the designated hikes in government spending should contribute $820bn (roughly $550bn x 1.5) to GDP over two years. If it takes a $275,000 increase in GDP to create one new job as suggested by the Republicans, which is really an inflated number and represents the worst case scenario for President Obama, and if the multiplier is 1.5 as suggested by Prof Krugman and others, then the increase in GDP as a result of spending alone would lead to the creation of 3mn jobs ($550bn x 1.5/$275,000) over two years.³  If the multiplier is zero as suggested by Prof Cochrane, then the spending stimulus would not create jobs. If the multiplier is one as suggested by Profs Woodward and Hall, then the government spending hike should increase jobs by 2mn ($550bn/$275,000), which will be acceptable if the tax cuts are effective. If we use a multiplier of 0.8 as calculated by Professor Barro, allowing 0.2 crowding out in consumption and investment of the suggested $550bn, we would then waste 400,000 jobs [0.2 x $550bn/$275,000] as a result of the crowding out by Obama’s spending plan compared with when the multiplier is one. This figure is disheartening and probably unreasonable!

Based on all this analysis, I am willing to bid that the government spending multiplier is greater than one. Overall, we need the spending multiplier to be greater than one to create the 3mn jobs promised by the president over two years without cutting taxes. No one has the exact answer now. 

Tax Multiplier

The debate on the tax multiplier is usually less sizzling, except these days because of Obama’s expenditures-loaded package. There is recognition among some economists that the tax cut multiplier is greater than the government spending multiplier if the cuts involve marginal income, payoff and corporate rates and are not refunds and rebates, are permanent in nature and/or their channeling mechanism goes through capital investment.

Christina Romer, President Obama’s Chair of the Council of Economic Advisors (CEA), estimates the tax cut multiplier to be around 3 because it raises capital investment, according to Greg Mankiw, something not written in textbooks which rely on the Keynesian disposable income-consumption channel.⁴  Economists, particularly those who are supply-siders, argue that tax cuts have more impact on employment than government spending. Using Romer’s estimate for the tax multiplier and repeating the same calculation for the impact of Obama’s designated tax cuts on job creation, we would have 2.9mn new jobs, which is almost equivalent to what we would have from the much larger government spending component in the package.

Currently,some economists and politicians are arguing for an increase in the $270bn tax cuts and the cut composition (favoring more of expensing and marginal rate cuts and less of refundables) and also a concurrent decrease in the $550bn government spending in Obama’s fiscal package to create the desired 3mn jobs or more over the two years. Needless to say, if the number for the Romer tax multiplier is correct, which is very doubtful, we would create more than 3mn jobs by just cutting taxes by a total of $300bn ( an increase of $30bn over what is suggested in the package), with no need for the corresponding hikes in government spending. In reality, while tax cuts on the margin have investment-stimulating merits, there is no guarantee that the tax relief will not be used for savings or paying down debts on the part of the consumer, particularly when the cut is not permanent. People try to smooth consumption over time as explained by the permanent income hypothesis. In fact, the true temporary lump-sum tax multiplier is close to zero as was evident from President George W Bush’s tax cuts.

The logic of a fiscal stimulus also contradicts the assumption that the tax cut is permanent. When they are temporary, the temporary tax multiplier is smaller than the temporary government spending multiplier, and thus tax cuts are not as effective in stimulating the economy as government spending. It is also likely that a temporary increase in government spending is more effective than a permanent hike over the common time period because of having less impact on the stream of future after-tax incomes than the permanent one. Therefore, there is no assurance that tax cuts will do the necessary economic stimulus. Moreover, tax cuts increase government indebtedness, worsening the government debt/GDP ratio and increasing the debt servicing burden. They do favor the private sector over the public sector, which is not undesirable. But when it comes to the production of public goods, which is very important in Obama’s package and socially desirable now, the private sector under-supplies especially in the current circumstances. Then the question is: who will build enough public highways, bridges, electric grids, schools etc to increase economic growth and raise the standard of living? This crisis offers us a rare opportunity to both cut taxes to stimulate business investment, to produce the public goods and infrastructures we are pensively waiting to have, and to help the American people and the world at large. Hence, we need spending hikes and sizable tax cuts to do the job, but also more government spending to invest in infrastructures. The multipliers play a major role in the composition. At the end, we should not forget the “confidence multiplier” of the stimulus package.⁵  

Notes

1.   See http://woodwardhall.wordpress.com/2008/12/11/measuring-the-effect-of-infrastructure-spending-on-gdp/

2.   See http://online.wsj.com/article/SB123258618204604599.html

3.   Most estimates center on $1mn as the true cost per job and $60,000 if the number is adjusted for increase in tax revenues as a result of economic stimulus. See, for example, http://www.sltrib.com/opinion/ci_11556149

4.   See http://gregmankiw.blogspot.com/2008/12/spending-and-tax-multipliers.html

5.   See http://www.economist.com/blogs/freeexchange/2009/01/blanchard_roundtable_1.cfm

Copyright MEES 2009.