How We Know the Stimulus Is Working

• The federal fiscal stimulus has helped lead the U.S. economy out of recession.
• This conclusion rests on statistical analysis, compared with a counterfactual of no stimulus.
• The Congressional Budget Office estimates results similar to those from Moody's Economy.com.
• Until the labor market turns around, criticism over the stimulus will persist.

President Obama's White House job summit this week was an attempt to tame a potentially dangerous political issue for the administration. Although the recession is over, the unemployment rate has risen sharply since Obama's inauguration—remaining in double digits even after November's surprising dip—and the American public is restless. Consumer confidence is off its lows seen earlier this year but remains well below its long-run average. The president's approval rating has been falling.

Obama's political opponents, meanwhile, have stepped up criticism of the unprecedented fiscal stimulus enacted early this year. For example, former Massachusetts governor and potential 2012 GOP presidential candidate Mitt Romney labeled the stimulus "a miscalculated failure" in a recent op-ed.

Romney is wrong. The stimulus is working and has been a major contributor to the economy's turnaround. This has been our view for some time, and recently the Congressional Budget Office said essentially the same thing. Job losses would have been much more severe without the government's support. Yet a political problem remains; although the economy is expanding, it is still losing jobs on net. And until the labor market sees clear improvement, President Obama's political stature will suffer.

Direct and indirect effects

We have several ways of knowing that the stimulus is supporting growth and job creation. The first is through the direct reports of stimulus recipients, required as part of the American Recovery and Reinvestment Act. Based on these reports, the administration said the stimulus has created or saved 640,000 jobs through the end of October. Some questions surround these data, in particular how to count a job that has been "saved" because of stimulus funding. State governments, for example, have used stimulus funding to avoid cutting programs and jobs in order to balance their budgets. Determining how many jobs were not cut, however, involves at least some conjecture. Still, the federal Government Accountability Office found recipients of federal funding "appear to have made good-faith efforts to ensure complete and accurate reporting."

Yet this understates the total impact of the stimulus on jobs. The state and local government numbers, for instance, include only direct effects of the federal aid. If 640,000 more workers had lost jobs, they would have spent less, reducing economic activity and leading to job losses in other industries such as retail, transportation/distribution and the like. Other parts of the stimulus such as the payroll tax credit and expanded unemployment insurance benefits have also boosted the economy through indirect impacts. And since the reports account for only about a third of ARRA spending to date, these other components comprise the bulk of the stimulus.

Hard data says it's working

Economic data over the past year also provide evidence that the stimulus is working. President Obama signed the stimulus into law on February 17. Real GDP, which had declined at an annualized rate of 5.9% during the fourth quarter of 2008 and first quarter of 2009, fell only 0.7% in the second quarter and expanded 2.8% in the third quarter. Consumer spending in particular had plunged starting in the spring of 2008 but has recorded substantial gains since March. And job losses, which averaged almost 700,000 per month in the first quarter of this year, fell to 500,000 in the second quarter, to 300,000 in the third quarter, and are below 100,000 per month so far in the fourth quarter.

None of this alone proves conclusively that the stimulus is working. It could be that Americans grew more confident after President Obama's inauguration and would have started to spend again even without the stimulus. Or that the recession had simply run its course and the economy would have turned around, stimulus or no.

Counterfactual analysis

To test that proposition, a counterfactual is needed: How would the economy have performed without the stimulus? Moody's Economy.com and the Congressional Budget Office have undertaken that task using similar approaches. The first step divides the stimulus bill—both its spending and tax cuts—into various components. The second step is to derive multipliers for each of these components, using historical data. For example, we estimate that each dollar spent on additional unemployment insurance benefits increases GDP over the following year by $1.61. The multipliers are combined with the actual spending to date under the ARRA to derive the change in GDP from the stimulus bill, relative to no stimulus. Finally, the historical relationship between GDP and employment is used to derive an estimate of the net impact of the stimulus on jobs.

It's important to note that most of the stimulus does not show up in the government component of national gross domestic product. One argument voiced by stimulus opponents, including some economists who should know better, is that federal government consumption and investment added less than 1 percentage point to annualized GDP growth in both the second and third quarters of 2009. This overlooks the fact that most stimulus spending ends up in other categories in the GDP numbers. Aid to state governments, for example, is why state and local government spending has been a plus for GDP growth, not a negative. And income support programs in the stimulus such as tax cuts and transfer payments are boosting consumer spending, which accounts for more than 70% of GDP.

An extra million jobs

Our results and those found by the CBO are in the same ball park. Moody's Economy.com estimates real GDP in the third quarter of 2009 was about 1.5% greater than it would have been without the stimulus. Indeed, without the stimulus, the recession would have continued at least into the third quarter. Overall, we estimate that through September, job losses from the recession would have totaled 8.2 million, rather than the 7.1 million actually seen. In other words, the stimulus has boosted employment by 1.1 million, relative to a no-stimulus baseline. The CBO estimates the stimulus increased real GDP between 1.2% and 3.2% through the third quarter and that employment is 600,000 to 1.6 million jobs higher than it otherwise would have been.

The Obama administration is also trying to make the point that things would be much worse without the stimulus. With job losses continuing and the unemployment rate still in double digits, that's a difficult sell. However, this is not unexpected. The peak in the unemployment rate follows the end of a recession with a substantial lag.

All this formed a backdrop for the White House jobs summit. President Obama needed to show that he understands the public's concerns and that he is doing what he can to boost job creation. The summit considered various steps to boost hiring, and Congress and the Obama administration are likely to propose additional support early in 2010 to sustain the recovery, although probably not with the "stimulus" label attached. Congress was recently urged to take several policy steps that would effectively boost job growth, including additional federal support for small-business lending, further aid to state governments, more aggressive efforts on mortgage modifications, and a job tax credit.

Given political and economic realities, however, there's only so much the administration can do. Job growth will pick up as the economy improves, but federal action can only induce employers to hire at the margin. The stimulus is working, but we won't see the results in the labor market until early next year.