You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Missing the ARRA Forest in the Reporting Trees

In early November we noted some atypically lousy reporting from the Washington Post concerning recovery programming. Unfortunately, it looks like the incredibly shrinking Washington Times also decided to get into the unbalanced reporting on ARRA game.

Rather than unfairly characterizing programs, this time the reporting focused on… well, reporting itself. Using data from the Franklin Center for Government and Public Integrity, the Times noted that $375 million in recovery dollars was delivered to nonexistent zip codes.

Now, I’m not going to get into the business of defending reporting failures—which these clearly are. But the Times already does a good enough job of explaining how these can be viewed as unintentional clerical errors (courtesy of Ed Pound, the ARRA Transparency Board Communications Director).

Instead I want to focus on the flip side: how these inaccurate zip codes actually show how much we’re getting right. Based on the most recent detailed, ZIP-level data available via Recovery.gov, recipients reported project data for over $158.7 billion in projects. That means the $375 million in faulty ZIP data accounts for just 0.2% of total reporting.

Or, to reflect just how solid ARRA’s new reporting standards are for fact-based governance, the public is getting accurate location data on 99.8% of its investments. I don’t know about you, but that is pretty impressive. 

And here’s why such unbalanced reportage is a problem in this particular arena. For too long the federal government sent out billions in programmatic spending with nary a way for the public to track geographic delivery. ARRA began the process of changing that, and that added accountability will create new avenues for policy research and, in the end, opportunities for better governance.

Let me be clear. Watchdogging our government is a valuable activity--but putting those findings in context is just as necessary as pointing a finger in the first place.