Do We Need a New Stimulus Package? Estimates of Spending and Tax Multipliers
About This Event

Harvard economist Robert J. Barro discussed whether the government's enactment of the first stimulus package has reduced or increased private spending, whether public-sector hiring has lowered or raised private hiring, and whether an additional stimulus package should be enacted. He explained his recent research, which shows that, over five years, Listen to Audio

Download Audio as MP3
an extra $600 billion in public spending deters $900 billion in private spending.

Event Summary

WASHINGTON, DC, JULY 23, 2010--Harvard economist and AEI visiting scholar Robert J. Barro presented research on the government's recent stimulus package and its impact on private spending. In his study, Barro used long-term macroeconomic data to estimate the size of government spending and tax multipliers and their isolated effects on real GDP. He adopted an innovative approach to calculating the multiplier effect of government spending on the economy, focusing on the defense outlays during periods of war and peace. An empirical analysis of defense spending patterns throughout World War II and the Korean War shows a positive expenditure multiplier of approximately 0.4 to 0.5, predicting a fifty-cent increase in GDP for every $1 increase in government spending. Although the overall effect is positive, a ratio less than 1:1 suggests that private spending is being crowded out, particularly in investment and the purchase of consumer durables. Thus, Barro predicted that over a five-year period a new stimulus package of $600 billion would deter nearly $900 billion in private spending, a loss that an already destabilized and largely indebted economy cannot afford.

  • "The estimated multipliers are all significantly positive but also significantly less than 1. So if I estimate a multiplier to be positive, then the estimate is that, in the short-run, the gross domestic product goes up by 50 cents on the dollar. But since the multiplier is less than 1, there's something not left over, so you have to get a cut-back in other parts of GDP because government purchases are only one part. So if government purchases go up $1 [and] GDP goes up less than $1, other parts of the GDP are estimated to be crowded out by that heightened government activity, totally up to a crowding out of 50 cents on the dollar. And it turns out that, empirically, most of the crowding out is in investments, including purchases of consumer durables. It's not saying that GDP goes down, but it's saying that it goes up less than 1:1."
    --Robert J. Barro, Harvard University and AEI

  • "Nondefense purchases underlie the sorts of expenditure multipliers that the administration is currently using . . . multipliers on the order of 2, 2.5. I'm going to try to point out where they get these spending multipliers that underlie the main estimates from the Congressional Budget Office or from the Council [of Economic Advisers] or from individual administration economists. Adding the nondefense government expenditures variable in the system means I am going to make the assumption that all of the association between this variable and the dependent variable, which is the growth rate of GDP, is causation from the government to the GDP. That is, there's no reverse causation. And if you make that assumption, you estimate a spending multiplier of about 2.6, which is completely in accord with the numbers that they're using. So I think mostly what this is picking up is that when the GDP is growing rapidly, that this spending is increasing and vice versa. I think it's mostly reverse causation. And that's one reason why when you look at a different sample starting in 1930 this effect disappears."
    --Robert J. Barro, Harvard University and AEI

  • "If you look at the magnitude of the fiscal stimulus in terms of how big it actually was, it can't possibly have had a major effect, one way or the other, on what's happened to GDP in 2009 and into 2010. It's not big enough actually to have a big enough effect even if you think the multiplier is some crazy thing, which I think the administration thinks. If you're trying to 'dis-stimulate' the economy in the short-run, 2010, I would say you're better off with a given fiscal deficit focusing on the tax side, and on the tax side I think I have some further evidence that I think it's better to focus on cutting tax rates rather than throwing money at people. That means it's particularly a mistake not to extend the tax reform of 2003."
    --Robert J. Barro, Harvard University and AEI


Event details

View complete summary.

Speaker biographies

Robert Barro is the Paul M. Warburg Professor of Economics at Harvard University, a research fellow at the National Bureau of Economic Research, and a visiting scholar at AEI. Also a coeditor of Harvard's Quarterly Journal of Economics, Mr. Barro has written extensively on macroeconomics and economic growth, and his research topics include empirical determinants of economic growth, the economic effects of public debt and budget deficits, and the formation of monetary policy. At AEI, Mr. Barro will be exploring the effects of the stimulus bill on the U.S. economy.

Also Visit
AEIdeas Blog The American Magazine

What's new on AEI

Rebuilding American defense: A speech by Governor Bobby Jindal
image Smelling liberal, thinking conservative
image Stopping Ebola before it turns into a pandemic
image All too many reasons for pessimism about Europe
AEI Participants


Robert J.
AEI on Facebook