The Impact of the Food Stamp Program on Transient and Chronic Poverty

Year: 2006

Research Center: Southern Rural Development Center, Mississippi State University

Investigator: Mills, Bradford, and Elton Mykerezi

Institution: Virginia Tech University

Project Contact:
Bradford F. Mills, Jr.
Department of Agricultural and Applied Economics
Virginia Tech University
314 Hutcheson Hall
Blacksburg, VA 24060
Phone: 540-232-6461


The United States has two relatively distinct types of poor; those consistently below the poverty line (chronically poor) and those temporarily exposed to poverty due to negative shortrun economic shocks (transiently poor). The social-safety-net needs of transiently and chronically poor families are likely to differ significantly, particularly with respect to food assistance. If food assistance is primarily used as an expenditure-smoothing mechanism, it will reduce transient poverty. If, on the other hand, food assistance is mainly used to support long-term expenditure levels, it will reduce chronic poverty. Yet little is known about the role of food assistance in ameliorating these distinct types of hardship. This study examines the differential role of USDA’s Food Stamp Program (FSP) on reducing household exposure to transient and chronic poverty.

The study uses Consumer Expenditure Survey (CEX) data from 1996 to 2004 to generate transient and chronic poverty measures from a short-term panel of quarterly family expenditures. In the application, transient poverty measures the component of poverty that stems from intra-annual (quarterly) variability in family expenditures over 1 year, while chronic poverty refers to the component of poverty associated with average expenditures below the poverty line over the same year. Specifically, a household is identified as chronically poor if expenditures averaged across the four survey quarters are below the U.S. poverty line. A family is defined as transiently poor if, on the other hand, expenditures in at least one quarter are below the poverty line but the family is not chronically poor. A severity-of-poverty measure is also employed in the study and is additively decomposed into its chronic and transient poverty components. Expenditures offer several advantages over income for measuring the dynamics of economic well-being in the short to medium run. More importantly, expenditures take into account the impacts of consumption smoothing as families use assets and credit markets to buffer shocks. Expenditures also appear to be less susceptible to systematic underreporting than income, particularly among poorer families.

The multivariate modeling effort focuses on identifying the determinants of the transient and chronic components of the severity of poverty. Most families have transient and chronic severity-of-poverty measures of zero, as well as zero FSP benefits over the previous year. FSP benefits and the severity-of-poverty measures are also likely to be jointly determined. Therefore, the severity of poverty and FSP participation is modeled as systems of censored equations. Due to changes in the way that FSP benefits are reported in the CEX, two different systems are estimated for each of the poverty measures. Under one specification, all data for years 1996-2004 are used, while alternative specifications use only data for years 2001-04.

Descriptive statistics for transient and chronic poverty measures indicate that transient poverty accounts for a large share of the economic hardship that U.S. families face.

The study shows that the incidence of exposure to a spell of transient poverty within a year is at least as great as the incidence of chronic poverty over the whole year. Further, the transient component accounts for approximately three-fourths of the total severity of poverty. Annual income-based poverty measures do not capture this important indicator of within-year economic hardship.

Poor families use food stamps to stabilize short-term expenditures.

Estimates of transient and chronic poverty systems of equations indicate that the common determinants of transient and chronic poverty are low human capital, minority status, and geographic location. However, the FSP primarily impacts transient poverty. Thus, poor and near-poor families appear to use the FSP as a short-term expenditure-stabilization tool rather than for long-term expenditure support. The parameter estimates from the transient poverty system imply that a 1-percent increase in FSP benefits for the average family of FSP participants reduces the transient component of the severity of poverty by 3.9 percent.

An alternative expenditure measure is then developed that smooths lumpy expenditures on homeownership and durable goods. Homeownership costs are replaced by the rental value of the property, and expenditures on durables are excluded from total expenditures. Adjusted expenditures are then normalized by the nondurable share of expenditures for families around the U.S. poverty line. As expected, the transient poverty component is much smaller, about a third of the total severity of poverty, under the alternative measure.

In terms of policy, the results suggest that a two-track FSP may be warranted. Given the primary impact of FSP benefits on transient poverty, fast-track eligibility and certification guidelines could enhance short-term program use and improve program effectiveness as an expenditure smoothing mechanism for transiently poor families. Alternative guidelines and certification procedures to enhance the ability of the FSP to meet the needs of chronically poor families also need to be explored.