Recovery Act Job Reporting Problems Persist

August 1, 2010

The fourth round of Recovery Act recipient data (covering the second quarter of 2010) has just been


on, and the numbers are again perplexing. The issue

I've written about before

— the large number of recipients putting a zero in the jobs column — endures, but first I want to address problems relating to those who do cite an employment impact of their ARRA contracts and grants.

The somewhat good news is that the total number of full-time-equivalent jobs associated with the forms of ARRA spending covered by the recipient reporting system rose to 755,000 from 682,000 during the previous quarter. Yet that gain of 73,000 jobs didn't put much of a dent in the 14 million-person army of the unemployed.

An examination of the spreadsheets underlying the totals reveals many more frustrating trends. The first is that just one of ARRA's scores of programs accounts for a disproportionately large share of the jobs. The State Fiscal Stabilization Fund (SFSF) accounted for 306,000 of the jobs, or 40 percent of the total. The problem is not with SFSF itself, which has done an important job in helping state governments weather the economic crisis, thereby allowing many public employees to keep their jobs and not swell the ranks of the unemployed.

Rather, it is the fact that just about everything else has been disappointing. Of the 73,000 grant and contract recipients providing employment numbers, 67,000 report fewer than 100 jobs. Scrolling through the spreadsheets, one is confronted with the dismaying sight of thousands of recipients reporting trivial numbers of jobs. Nearly 15,000 recipients report only a fraction of one job, down to the absurd listings for .01 jobs.

The average number of jobs per recipient is only about 10 — while the average amount of ARRA funding recipients have already gotten is $1.1 million. Excluding SFSF, the jobs average falls to about 6. Overall, the cost per job (in terms of ARRA funds already received) for non-SFSF programs is about $117,000, while for the dozen states with the highest SFSF numbers, the cost is only $72,000.

There is also a significant difference between grant recipients and contract recipients (the latter refers only to those working directly for the federal government). For grant recipients overall, the cost per job is $103,000, while among contract recipients it is $159,000. While some of the grant money ends up being used by states to award their own contracts, the discrepancy suggests that states do a better in creating job with their ARRA money than do federal contractors.

This brings us to the issue of the zero-job reporters. As in

previous quarters

, a large number of recipients (some 24,000) claim that their employees performed no work at all in connection with their ARRA grant or contract. Many of these projects have not yet started work or are not very far along. If we exclude those whose status is listed as "not yet started" or "less than 50% completed," we are left with 9,298 recipients. If we then exclude those reporting they have not yet received any ARRA funds (which is hard to believe when a project is at least half done), then we are still left with 8,566 zero-job reporters. This is up from 6,806 in the previous quarter.

What are we to make of the large number of ARRA recipients claiming either no jobs at all in connection with their contracts and grants or a microscopic employment impact? There are at least three possibilities.

The most benign is the matter of timing. Some of the projects reporting low job numbers are just getting started, even though we are well into the second year of the Recovery Act. Others may have already been completed before the last quarter or were winding down during that period, causing them to have few jobs to report. (The employment reporting is not cumulative.)

A more serious issue is the possibility that many recipients still do not understand the rules for job reporting. There was considerable confusion over the original regulations, which required recipients to calculate jobs created and jobs retained, so the Office of Management and Budget adopted a new system. Recipients are now supposed to tally all hours of work associated with ARRA projects, even if the position is not a new or retained one. It appears that many of them are still mixed up and are reporting unrealistically low numbers.

Finally, there is the possibility that many ARRA recipients are demonstrating the same inclination seen among U.S. employers in general these days: In the face of economic uncertainty, they are minimizing their hiring. In doing so, they are thwarting the intended job creation aspect of the Recovery Act.

When ARRA was first being debated, many critics charged that all that government spending would be rife with waste and fraud. That does not seem to have materialized on a significant scale, but perhaps we need an investigation of whether many employer recipients are abusing the Act by not using the funds to put more people to work.