Using Data to Test and Tell - Transcript

Plenary Session Highlights with:
David E. Altig, Federal Reserve Bank of Atlanta
Eugenie L. Birch, University of Pennsylvania
Ann Carpenter, Federal Reserve Bank of Atlanta
Ira Goldstein, The Reinvestment Fund
Shannon Van Zandt, Texas A&M University–Galveston
Elizabeth Weigensberg, Chapin Hall at the University of Chicago

David E. Altig: The problem is the same in many ways, which is we have a multifaceted, multidimensional problem and we're trying to get our heads around it and need ways to get our heads around it, which really aren't amenable to sort of standard, simple, statistical analysis, and certainly not when you first address the problem where you're taking bigger and bigger and bigger sets of data, and bigger and bigger and bigger information sets, and larger numbers of indicators, and trying to figure out where your attention ought to be.

Just one other quick note on this "using data to make data." This had to do with the CWICstats initiative, which really kind of hit home for me of linking individual-level administrative data with other sets of data to get a big longitudinal, very complete picture. In the data-mining world this is where the gold is, is to actually begin to put together data sets that are disparate so that we can actually get real pictures of how people are living and how they are progressing, something that we really have far too little of at this point.

Eugenie L. Birch: Other people in the world are very much engaged in this sort of thing. What you can see here is the recent report of the OECD [Organisation of Economic Co-operation and Development], which has been coming up with well-being indicators as a result of the Stiglitz Commission. That has generated a lot of thinking about that the GDP [gross domestic product] is not enough to measure well-being—we need other sorts of things. Of course, the main idea is that we needed to frame these indicators in something, and we took the "livability principles," which had been created by the partnership. And the livability principles are quite aspirational, we had to operationalize them and we did that by defining what qualities they embodied. And you can see on the left-hand side what those were.

Elizabeth Weigensberg: Part of this report, we were able to come up with recommendations for the local workforce development system. Those recommendations included improved measures, including establishing longitudinal measures, establishing common measures with some flexibility, expanding what's measured, including assessment data, skill development, intern measures, engagement with the employer community and the broader community in which these organizations served, and really, ultimately, to try to improve that fragmented administrative data that's out there now to try to do an integrated data system.

Ira Goldstein: So what do we know? We do know that the early results are sort of intriguing, I think. That NSP2 [Neighborhood Stabilization Program] clusters seem to be doing a little better than the NSP1 clusters. We know that the higher risk score areas did not necessarily perform worse, in fact, it appears that they might actually be performing just a little bit better than areas where NSP activity went into less risky places at the time it starts. What we find is that investments mixed performed better in general where there were more investments in those places. We found that they actually did better in lower-income places, and in places where there were higher percentages of minority group members as measured by us, meaning African-American or Hispanic. Those are the generalized results. So the lesson learned there is "do it all," whatever you are doing, do it all, as much as you can do it.

Shannon Van Zandt: But Hurricane Ike was an opportunity, and I always am reluctant to call hurricanes "an opportunity," but as a researcher, they are. It gave us an opportunity to see if what we had predicted actually would be observed. We find that it's not enough to make knowledge about a program available, but we actually have to go to those areas that are in need of that aid because here we have a population that needs the help but they're not applying for it.

Ann Carpenter: So looking at who is coming back, whether or not they have appropriate housing (temporary and long-term housing), and, also I think from what I've seen in the literature and also in certain communities in Mississippi, planning in advance is really important. So having either a disaster specific recovery plan or just a really good quality, comprehensive plan that guides once something like this happens there are opportunities, but you also don't want the vulnerable populations, for example, to fall through the cracks. And so having the appropriate plan that addresses that and makes sure that everyone who wants to come back can come back. I think that is my biggest point there.