Did Stimulus Dollars Hire the Unemployed?

Garett Jones, Daniel M. Rothschild

In an effort to boost hiring and job creation and to invest in a variety of domestic infrastructure programs, Congress passed and the president signed the American Recovery and Reinvestment Act (ARRA), commonly known as the economic stimulus package, in 2009. ARRA represented one of the largest peacetime fiscal stimulus packages in American history. But little is known about the ways in which organizations and workers responded to the incentives created by the bill.

To address the lack of knowledge about ARRA funding, we surveyed hundreds of firms, non-profits, and local governments that received ARRA funding. We collected over 1,300 anonymous, voluntary responses from managers and employees that allow us to better understand what happened at the organizations that received contracts funded by ARRA spending. This bottom-up study of ARRA is the first of its kind. We hope that others, especially government agencies, will build upon this on-the-ground analysis.

The survey asked a number of questions critical to analyzing the effectiveness of ARRA: Were new workers mostly hired from the unemployment lines or did they get "poached" or "raided" from other organizations? Did workers "game" the unemployment insurance system by waiting until benefits ran out before taking a job? Did Davis-Bacon prevailing wage laws force organizations to pay above market wages to new hires?

The quantitative survey results here complement our interview-based results in a companion paper (Jones and Rothschild, 2011) and suggest a number of interesting results, some expected and others surprising:

• ARRA funds led to worker hiring and retention at stimulus-receiving organizations that responded to our voluntary survey (Figure 1). On average an organization that received stimulus funds equal to 10 percent of its annual revenue reported retaining or hiring workers equal to 5 percent to 6 percent of its workforce [Tables 1, 2]. ARRA funds did more than just sit in bank accounts or pad company profits. However, our data don’t tell how many of these hires were for were part-time or temporary jobs. Our in-person interviews indicated companies frequently included parttime and temporary jobs in reported job totals.
• Hiring isn’t the same as net job creation. In our survey, just 42.1 percent of the workers hired at ARRA-receiving organizations after January 31, 2009, were unemployed at the time they were hired (Appendix C). More were hired directly from other organizations (47.3 percent of post-ARRA workers), while a handful came from school (6.5%) or from outside the labor force (4.1%)(Figure 2). Thus, there was an almost even split between “job creating” and “job switching.” This suggests just how hard it is for Keynesian job creation to work in a modern, expertise-based economy: even in a weak economy, organizations hired the employed about as often as the unemployed.
• Only about half (47 percent) of responding ARRA-receiving organizations said it was easier to hire high-quality workers than before the financial crisis. The rest said hiring good people was either as hard as (41 percent) or harder than (12 percent) before the financial crisis of 2008 (Appendix B).
• There was no tendency for stimulus funds to go to organizations that found it easy to hire good people. Under Keynesian theory, government spending has its greatest effect when targeted toward sectors of the economy with slack; by this job-focused measure, stimulus funds were poorly targeted (Tables 3, 4).
• However, there was a weak tendency among non-government organizations for stimulus funds to go to those organizations that said “things had been slow” before ARRA funds arrived. There was no such tendency among government organizations (Table 7, 8, 9).
• Workers did not game the unemployment insurance (UI) system when deciding when to take a job. There was no unusual tendency to take jobs right around the time of UI benefit expiration (Appendix C).
• Among organizations required to pay prevailing wages, 38.2 percent thought that they could have hired workers at wages below the Davis-Bacon prevailing wage (Figure 3) while another 17 percent were unsure. This meant higher costs for the federal government and fewer jobs created.