Aside from health-care reform, probably the most divisive issue in Washington today is the $787 billion economic stimulus program.

One camp argues that the Recovery Act has done a good job in preventing the country from plunging into a more serious crisis than the one we’ve got, while another camp says the whole effort has had little effect and was an expensive mistake. We suspect the former conclusion is correct, but would like better evidence.

Most arguments are based on differing assessments of the first round of jobs data released in October. The Recovery Act’s exemplary transparency provisions cover direct federal contracts and some of the grants to states—a limited part of the overall stimulus program. They counted about 640,000 jobs created or saved so far.

There’s been much media attention given to instances in which job numbers seem to have been seriously exaggerated. Yet, there also are cases in which they were apparently understated. More than 2,000 contract and grant recipients report that their stimulus projects are more than 50 percent complete—yet they claim not to have created or saved a single job.

Clearly, there is confusion about the right way to report stimulus-related jobs. This is not surprising, given that tens of thousands of companies, government agencies and nonprofits are being asked to do this for the first time.

But the blame is not entirely with recipients. Instructions given by the Office of Management and Budget (OMB) did not adequately consider the variety of situations reporting entities would face. For example, exactly how does an employer determine which existing workers put on stimulus projects should be counted in computing the number of saved jobs?

Should they include only those who were about to be laid off before Recovery Act money was received—or should they also include those who might have been let go at some later point in the absence of those funds?

The furor over the quantity of stimulus-related jobs has obscured the question of their quality. Despite efforts of the Coalition for an Accountable Recovery (CAR) and others, the OMB decided not to collect data on wages, benefits or the number of full-time vs. part-time positions. Without this information, we cannot tell to what extent the Recovery Act is generating jobs that allow workers to support their families in a decent fashion.

Nor can we tell who is getting Recovery Act jobs. CAR also argued strongly for collection of data on the demographic characteristics and residential area of workers on stimulus projects. With information on factors such as race, gender and neighborhood of residence, taxpayers would be able to determine whether all communities are getting a fair share of Recovery Act employment.

The Recovery Act is a many-sided law meant to lessen what has turned out to be a more serious economic downturn than anyone expected. In the same way it may need to be adjusted because of the magnitude of the crisis, so must its reporting and transparency provisions be refined to reveal how well it is working and who it is serving.

Greg LeRoy is executive director of Good Jobs First, a nonprofit resource center on economic development accountability that is co-chairing the Coalition for an Accountable Recovery.