Tuesday, July 11, 2006

Visualizing data: a food stamp example

Have you seen this inspiring example of data visualization with a public purpose? It is a conference presentation by Hans Rosling. I first saw the link on BoingBoing a couple weeks ago. It includes links to Gapminder, the authors of a clever Flash animation that allows the viewer to choose variables and time periods, and then watch animated bubble plots that describe economic and demographic information about the world's countries.

Hey, I wondered, why can't we do something fun like that for U.S. food policy?

The following animation illustrates the effect of economic conditions and welfare reforms on the Food Stamp Program caseload in the 1990s. Because the Food Stamp Program is an entitlement, the federal government is somewhat at the mercy of factors beyond its control, which change the number of eligible people from one year to the next.

Caseloads increased sharply in the early 1990s and declined even more sharply in the second half of the decade. In the heated debates that followed, some argued that the caseload fluctuations respond to economic conditions, while others placed the blame for caseload declines on the welfare reforms of 1996.

The animation shows each state's Food Stamp Program participation level (as a fraction of the state population) on the vertical axis, and the state unemployment rate on the horizontal axis. The size of the bubbles in the plot reflects the state population, so you can distinguish major states from minor ones.

Welfare reform occurred at different times in the various states -- the timing of the reform is indicated by a change in color from blue to red.

You can see for yourself that the caseloads appear to respond in the expected way to rising unemployment in the early 1990s and falling unemployment in the late 1990s. Welfare reform does not have a visible effect on caseloads. Following a suggestion from the Center on Budget and Policy Priorities, Nader Kabbani and I did an analysis that suggested that comparatively obscure policies about food stamp recertification periods might be more influential than the welfare reforms illustrated in the animation.

What is really cool is to see the diversity of state experiences with a clarity that a static graph could never achieve.

The "start" button starts the animation, and the "pause" button stops it. To restart, hit the "start" button again. I plan to start using such animations in my teaching, and will post more of this sort of thing in the coming year. Comments and suggestions welcome.

1 comment:

Anonymous said...

"In all of this banter about 'picking your poison', has anyone ever stopped to consider that there may be a healthy alternative to both HFCS and cane sugar. What happened to Honey?"

Well, honey is on the same list as HFCS with the nutrition nuts, because it's "predigested bee vomit." Maple syrup, maybe? Or is there a reason not to eat that also? Jeez!