FIGURE 2-5 Trends in prevalence rates of food insecurity and very low food security in U.S. households, 1995-2011.
NOTE: Prevalence rates for 1996-1997 were adjusted for the estimated effects of differences in data collection and screening protocols used in those years.
SOURCE: ERS, 2012. Calculation by ERS based on Current Population Survey Food Security Supplement data.
fact, SNAP recipients are twice as likely as SNAP-eligible nonrecipients to report being food insecure (Tiehen et al., 2012). (A similar contradiction is seen in Figure 2-6, presented later in this section.)
Because participation in SNAP is not likely to be unrelated to food security status, a selection problem arises in evaluating the effect of the program on food insecurity (Currie, 2003). Studies evaluating nonrandom selection by nonexperimental statistical methods (e.g., Gundersen and Oliveira, 2001; Kreider et al., 2012; Mykerezi and Mills, 2010; Yen et al., 2008) generally have found that SNAP reduces food insecurity.
A broader metric of the effect of SNAP on the well-being of individuals and households is the antipoverty effectiveness of the program. While the provision of food assistance has a modest effect on household work effort (Hoynes and Schanzenbach, 2011), it increases household resources for the purchase of food and thus should reduce the incidence and severity of poverty by freeing up income for the purchase of other goods and services (Tiehen et al., 2012; Ziliak, 2011). Ziliak (2011) used data from the Annual Social and Economic Supplement of the CPS to estimate the number of persons lifted out of poverty by SNAP in any given year from 1999 to 2009. Figure 2-6 shows that the antipoverty effectiveness of SNAP increased over the decade, with