The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

This article presents data from four major surveys used to collect information on the extent of hunger in the United States. Three are periodic surveys conducted by U.S. government agencies: the Nationwide Food Consumption Survey (NFCS) and the Continuing Survey of Food Intakes by Individuals (CSFII) of the U.S. Department of Agriculture (USDA), and the National Health and Nutrition Examination Survey (NHANES III) of the U.S. Department of Health and Human Services (DHHS). The fourth survey, the Community Childhood Hunger Identification Project (CCHIP), was a project of a private advocacy group, the Food Research and Action Center (FRAC). All four of the surveys rely on adult responses about the children living in their households for information on food insufficiency among children. Adult responses on behalf of children may be subject to misreporting and bias, but there is no evidence reported of such bias in these surveys. Summary information about all four surveys is presented in Table 1.

Of the surveys, only the CCHIP was specifically designed to measure hunger as part of a wider range of experiences, similar to the concept of food insecurity discussed above. As the first effort to make scientifically defensible measurements of hunger, the CCHIP survey provided a foundation for the current research and measurement efforts discussed in this article. In contrast to those in the CCHIP, the hunger-related questions in the three government surveys were intended for exploratory data gathering. Because the USDA and the DHHS recognized the scientific limitations of their hunger-related survey questions, they adopted the term food insufficiency to reflect the lack of a precise definition of what the survey questions were intended to capture. The term food insufficiency is used throughout the rest of this article where responses to the government hunger-related survey questions are discussed. Where CCHIP results are presented, the term hunger is used, consistent with the CCHIP reports' terminology. While their terminology differed, both the government and the CCHIP surveys were focused on the same aspect of food insecurity, namely resource-constrained hunger.

There is a similarity in the approach to collecting information on food insufficiency in the NFCS, the CSFII, and the NHANES III that is due in part to the Food and Agriculture Act of 1977, which required the DHHS and the USDA to submit a plan to integrate their respective surveys.¹⁰ Since 1977, the USDA in the NFCS and the CSFII has asked the following question to identify the respondent's perception of the sufficiency of the household food supply:

"Which of the following statements best describes the food eaten in your household: (1) Enough of the kinds of food we want to eat, (2) Enough but not always the kinds of food we want to eat, (3) Sometimes not enough to eat, or (4) Often not enough to eat?"¹¹

Using this question, food sufficiency is defined as a response of 1 or 2 and insufficiency as a response of 3 or 4 and is reported for households and not for individuals.

In the NHANES III (1988-91), information on food insufficiency was collected by asking questions both about entire households and about individuals, but the data presented in this article reflect only the household responses. NHANES III used a question similar to that of the NFCS to determine household food sufficiency but without the option "Enough but not always what we want to eat" because the intent of the survey's designers was to capture quantitative, not qualitative, information, and field testing indicated that including both concepts in one question was too complex for some respondents.^12,13

In the CCHIP, children are classified as "hungry," "at-risk of hunger," or "not hungry" based on the responses of an adult in the household to a set of eight questions. Each of the questions asks about hunger (CCHIP uses the term "hunger" rather than the government's term "food insufficiency") in light of available resources. The eight questions contain two on household food sufficiency (for example, "Does your household ever run out of money to buy food to make a meal?"), two on hunger among adults in the family (for example, "Do you or adult members of your household ever eat less than you feel you should because there is not enough money for food?"), and four on child hunger (for example, "Do any of your children ever go to bed hungry because there is not enough money to buy food?"). A child is termed "hungry" if affirmative answers to at least five of the eight questions are obtained, which requires that at least one of the questions centered on children be answered in the affirmative. Children in families that provide affirmative responses for one to four questions are termed "at risk," and children in families with no affirmative responses are termed "not hungry."¹⁴

In addition to differences in defining hunger, there are other important differences between the government surveys and the CCHIP survey. First, while the government surveys were designed to yield estimates representative of the entire noninstitutionalized population, the CCHIP survey population is not as broadly representative. In the CCHIP, only low-income families with at least one child under 12 years of age were interviewed (such families are defined as those with incomes at or below 185% of the federal poverty line at the time the interview was conducted).¹⁵ The CCHIP survey was based on a nonrandom sample representing a mixture of locally administered statewide, districtwide, multiple-county, and single-county surveys. Responses to the local CCHIP surveys were combined, and national population weights were used to make national projections about the extent of hunger, but such projections may not actually be representative of the national population. Another difference is that, while the government surveys were conducted by professional surveyors, not necessarily from the same community or socioeconomic class as the respondents, the CCHIP was conducted by specially trained local residents in their own communities. The CCHIP approach was deliberately intended to ease surveyors' access to low-income neighborhoods and households and improve response rates.

The government surveys and the CCHIP are household surveys, and thus none of them includes the institutionalized population, homeless people, migrant workers, or Native Americans living on reservations—all groups that are likely to experience food insufficiency. Therefore, all projections based on these surveys will tend to underestimate the number of people experiencing food insufficiency, or hunger, in the United States.