The average monthly number of recipients in the Supplemental Nutrition Assistance Program (SNAP, formerly food stamps) reached 46.6 million, or about one in every six Americans, in FY 2012, according to the most recent USDA data. The weak economy was principally responsible, as the 2007-2009 downturn and slow recovery increased the pool of eligible recipients from an average of 38.9 million a month in 2007 to 50.7 million a month in 2010. At the same time, the program’s participation rate—the percentage of eligibles receiving benefits—went up from 65.8 percent to 75.2 percent. (The basic pattern here is more noteworthy than the specific estimates, given some recent changes in USDA’s method of calculating participation rates.)

Underlying these numbers is a little-noticed trend: a large and growing gap between the participation rates for urban and rural Americans. Individuals living in urban areas participate at lower rates: in 2010, they had a participation rate of 72.9 percent versus 85.6 percent for those in rural areas. This differential—now, more than 12 percentage points—has grown in recent years. It was less than 5 percentage points during 2003 and 2004. Going back to 1997 and earlier, eligibles in urban areas actually participated at higher rates than eligibles in rural areas. In 1995, for instance, the urban participation rate was 77.8 percent, compared with a rural rate of 65.5 percent.

What’s going on here? Three possible explanations deserve attention: benefit size, ease of program access, and preferences about receiving public benefits.

First, if urban cases tend now to receive smaller benefits, relative to rural cases, we might expect households in urban areas to be less likely to participate. The available data back to 1997 don’t support this, however. The average benefit for urban cases has consistently been slightly higher than rural cases—proportionally, by 2 to 7 percent—since 1997, without any discernible shift in this pattern.

Second, if access to SNAP benefits has become relatively easier for rural cases than for urban cases, we might expect the observed flip-flop in participation rates. Indeed, the availability of online, mail-in, or phone-in options for program application and renewal have likely benefitted rural cases more than urban cases. Prior to such administrative improvements, people in rural areas needed to travel greater distances to visit a local welfare office than people in cities did. One would not expect, however, these developments by themselves to raise rural participation rates above urban rates. We need to look further.

Third, if people living in urban areas have become less inclined to receive public benefits, it would support the urban-rural trend in SNAP participation rates. This indeed appears to be a contributing factor, reflecting the shifting racial-ethnic composition of the urban poor. Specifically, Hispanic households now represent a much larger percentage of urban households with incomes below 200 percent of the poverty level. Studies of both SNAP and Medicaid indicate lower rates of participation among eligible Hispanics.

This latter explanation has implications for future growth in SNAP caseloads. Against the backdrop of an improving economy, which will bring down the number of people eligible for the program, the long-term upward trend in program participation rates is likely to be moderated by a continued urban-rural imbalance in SNAP participation rates.