Disaster Caused Increased Food Stamp Participation in Washington?

In this New York Times article about the near-record rise in food stamp recipients over the past year, a graphic is included with a map showing that Washington State participants in the USDA's Food Stamp Program (FSP) has risen by about 25% -- unfortunately, the highest enrollment increase in the country.

Clark Williams-Derry over at Sightline, who I recently linked to on the related issue of public health and the wealth gap, read the NYT's article and took note of one cited reason for increased FSP participation: natural disasters. The NYT article was specifically referring to the effect of Hurricanes Katrina and Rita on FSP enrollment, not Washington State. However, Clark suggested that the December 2007 flood disaster in Western Washington (Lewis, Thurston, and Grays Harbor County) might be associated with Washington State's FSP statistics for last year.

Typically, rising food stamp enrollment is a clear sign of a slowing economy. But Washington's worst-in-nation food stamp surge likely had another cause: the flooding in west-central Washington late last fall that put many folks in shelters, and still more in dire financial straits.

He uses this supposition to support two claims: 1) That disaster's aren't natural and 2) that this provides even more reason for "stopping global climate change."

I had three reactions to Clark's interesting post:

1) Right on! The idea (fact!?) that disasters are not "natural" is a major theme that we teach here at WWU. It's great to see people who are not in disaster studies making this link and putting it out there.

2) Sightline really likes to use flooding in Washington as evidence for global climate change. I worry about over-selling this point. Flood disasters as the result of unsustainable residential, commercial, and economic development practices are a major issue we need to contend with regardless of climate change. And I would argue that development practices are ultimately more important (and easier) to focus on than the vague notion of "stopping climate change."

3) So what's the deal with the FSP and the Western Washington flood disaster anyway? Did the flooding increase FSP enrollment? And, if so, does this explain the 25% increase in FSP recipients in Washington State?

To answer these questions I did a bit of data compilation and crunching. For those who just want the answers without the explanation: Yes, the flooding likely increased participation in the FSP, but not for the reason you'd probably think. And no; it's unlikely that the flood disaster in Lewis, Thurston, and Grays Harbor Counties accounted for the 25% FSP increase in Washington State from 2006 to 2007.

The upshot is that Washington State lawmakers (and other concerned folks) should not brush of this 25% increase in FSP participants as the result of the December 2007 flood disaster. It's a signal, but not a big one, not even big enough to bump Washington State out of the top spot in terms of FSP participant increase compared to other states. It's important that we do figure out what the bigger signal is. (Anyone?) We need policies and plans to increase our resilience to flooding, but more importantly we need to first increase the resilience of folks to economic disaster who don't experience a natural hazard.

If you're interested in the lengthy details of what I found and how I found it, keep reading.



Before I get into things, let me say that this should be considered a "back of the envelope" analysis. I've done several hours of synthesis and analysis on this, but much more could be done and I wasn't able to find all the data I want.

So like what kind of data couldn't I find? Data on how many FSP participants enrolled as a result of the December 2007 disaster declaration in the three counties.

A minor issue.

So I did what any geographer geek would do: construct a census data spreadsheet to in order evaluate my two questions.

Question 1: Did enrollment in the FSP increase as a result of the December 2007 flooding?

To lighten the analytical load, I only looked at Lewis County for this question. After compiling the household and family income data from 2006 (the most current available), I realized that I didn't even need to compile the data to determine that more Lewis County residents were eligible (if not enrolled) for the FSP than normal.

The USDA has different eligibility requirements [pdf file] in the case of disaster declarations. In many instances, the income level (minus deductions or, in the case of disasters, losses) required to be eligible is higher. That is, it's easier for people to qualify for the disaster FSP.

Well, some people.

The maximum one-month income for a single person to qualify for the FSP goes from $1107 to $1416 -- a 28% increase -- in the event of a disaster declaration.

This appears to make sense -- in the case of disaster more money is available to assist more people in need. However...

The maximum one-month income for a family of four goes from $2238 to $2295 -- a measly 2.5% increase.

Does this make sense? It's significantly easier for an individual to be eligible for food stamps in the event of a disaster, but for a family of four it's negligibly easier.

Perhaps someone with more knowledge about this policy could tell us why the difference.

Regardless of the apparent disparity in disaster FSP eligibility, the fact remains that more people are eligible as the result of the higher income thresholds and thus more people are likely to be participants. In addition to the easier income requirements, of course one would assume that the disaster losses would increase the number of eligible people even without a change in the income requirement. That is, those people who would not qualify for the basic FSP based on income would qualify after the disaster because they can deduct their losses from their gross income.

I was curious which would have the bigger effect on eligibility: the easier income requirements or the increased deductions from flood loss.

Based on the little "what if" analysis I did with the census data, it appears the change in income level has a much greater effect on eligibility than direct disaster losses, in the case of Lewis County specifically. I estimate there was about a 40% more people eligible for the FSP just based on the income requirement difference. Conversely, the eligibility numbers were not very sensitive to losses, which I modeled as proportional to one month of home ownership cost -- data available in the census. In fact, I had to assume that every household incurred over 80% damage to their house to include the next income bracket, resulting in 92% more people being eligible for the FSP. (If anyone wants to look at the spreadsheet, just email me.)

Now obviously, this is as an estimate based on several assumptions, but I'm confident of the relative upshot: there was increased participation due to expanded eligibility, but that eligibility is more related to a policy decision (income threshold increase) than from direct losses from the flooding.

I don't have actual numbers on flood-related participants from the December disaster, but the USDA certainly keeps track of these numbers for all disasters [word document]. The USDA's data for past disasters, including Hurricanes Katrina and Rita, clearly shows an increase in FSP participation.

The first question that came to mind when I looked through their data from Katrina and Rita is why didn't the NYT (or whoever they interviewed) actually crunch the numbers to see IF those disasters contributed to the increase of participants in the Gulf Coast states? Because, honestly, it's rather unintuitive that a disaster that started in 2005 resulted in a bigger FSP enrollment spike between 2006 and 2007, then it did between 2005 and 2006. The disaster FSP benefit period after Katrina and Rita was, as far as I can tell, less than a year in all cases. So the NYT (or their source) is arguing that Katrina and Rita had knock on effects that lead to folks enrolling in the basic (non-diaster) FSP in 2007, but not in 2006. Disasters unfold. It certainly is possible that there are folks who live(d) in the Gulf Coast states who have fallen on worse times in the past year than in 2005 and 2006. But I'm not sure the numbers would be that large.

And this brings us to my second question....

Question 2: Did the December 2007 flood disaster in Western Washington result in the 25% increase in FSP participants in Washington State?

First of all, I'm not even sure if the USDA would consider the disaster enrollees from the December 2007 flood disaster part of 2007 reporting. The application period was December 10-14th, 2007. I'm guessing that all disaster FSP enrollees were authorized before 2008, but it's not unimaginable that they weren't.

Second, and most importantly, it doesn't appear to me that there were even enough eligible people in Lewis, Thurston, and Grays Harbor County to represent a dominant portion of the 25% increase of Washington State FSP participants. I estimate that in 2006 there were about 540,000 FSP participants in Washington State. A Washington State press release from 2007 say there are 500,000 participants in the state. The 2002 (the most recently available) Washington State Department of Social and Health Services "Blue Book" says there were 527,000.

The entire population in the three counties represents about 70% of the number of FSP participants in Washington State (depending on which of the above numbers you use!). In other words, about 35% of the population of the three counties would need to be new FSP participants as a result of the disaster to comprise the full 25% FSP participant increase for the entire state. Based on DSHS's "Blue Book," this represents over 100 times more FSP participants in Region 6 (which includes the three counties and eight others) than normal. My estimate for Lewis County was that an additional 18% of the population of Lewis County were eligible for the disaster FSP. Lewis County was the hardest hit, while Thurston County has the highest population, and certainly not all 18% of eligible people enrolled in the FSP.

The take home message is that, even though these numbers are very rough and the analysis quick, it's unlikely that the December 2007 flood disaster contributed to a majority of the 25% FSP participant increase in the State of Washington. My estimate is that the disaster represents 20 to 40% of the increase. So even without out the December 2007 floods, the State of Washington still had the greatest increase in FSP enrollment in the country between 2006 and 2007.

Now let's start talking about what we should do about this....

UPDATE: In response to Rebekah's comment, I looked a little harder at what counties were eligible for the disaster FSP. The info I read referred to just Lewis and Grays Harbor (and I included Thurston as an assumption for larger population numbers). The confusion was from the fact that more counties were deemed eligible later [Word doc]:

Those counties approved for the food stamp assistance are Clallam, Grays Harbor, Kitsap, Lewis, Mason, Pacific, and Thurston counties.

Deadlines for processing and receiving the benefits are January 2, 2008, for Grays Harbor, Lewis, Mason, Pacific, and Thurston counties, and January 7 for Clallam and Kitsap counties.

This snippet from DSHS answers two questions for me: 1) I haven't rerun the population count, but Thurston I think is still the biggest county, so the general conclusion above would remain the same. 2) The deadline for application was in 2008. The 25% increase in FSP recipients refer to the change from 2006 to 2007. So there were participants associated with the flood who could not be in that statistic.