Poverty, food insecurity, and nutritional outcomes in children and adults

https://doi.org/10.1016/j.jhealeco.2003.12.008Get rights and content

Abstract

Using data from the National Health and Nutrition Examination Survey, we examine the relationship between nutritional status, poverty, and food insecurity for household members of various ages. Our most striking result is that, while poverty is predictive of poor nutrition among preschool children, food insecurity does not provide any additional predictive power for this age group. Among school age children, neither poverty nor food insecurity is associated with nutritional outcomes, while among adults and the elderly, both food insecurity and poverty are predictive. These results suggest that researchers should be cautious about assuming connections between food insecurity and nutritional outcomes, particularly among children.

Introduction

The US Government has been producing official poverty estimates for over 35 years. The intent of the poverty measure is to identify households with “inadequate resources to obtain basic living needs,” where basic needs are considered to be food, clothing, and shelter (Citro and Michael, 1995).

Poverty in the United States is determined by comparing a family’s actual income to a poverty line that was originally a multiple of the amount of income deemed necessary to purchase a “thrifty food basket” (Orshansky, 1965). However, in recent years, the idea that official poverty rates measure actual physical deprivation has been criticized. For example, Slesnick (1993) finds that consumption-based measures of poverty are substantially lower than those based on income, and Jencks and Mayer (1989) argue that the official poverty measure does a poor job of measuring material hardship. A recent National Academy of Sciences report concluded that the official poverty measure is flawed and made numerous recommendations for improving it (Citro and Michael, 1995).1

Partially in response to criticisms of the poverty measure, the government has also collected more direct measures of material hardship such as the food insecurity measures. These measures are widely used to measure “hunger” in America. This paper examines the extent to which poverty and food insecurity are predictive of several measures of nutritional status. We use a unique dataset, the National Health and Nutrition Examination Survey III (NHANES III). In addition to the dietary recall information collected in other data sets, NHANES III conducts a physical exam of the participants, collecting weight and height and analyzing their blood. This examination data provides measures of nutritional status for individuals of all ages that is not subject to recall or proxy bias. Thus, these data allow us to more precisely examine the way in which correlations between poverty, food insecurity, and nutritional outcomes vary by age. This last point is particularly valuable given that many anti-hunger programs (such as school nutrition programs) target people of particular ages.

We find that once poverty is controlled for, the food insecurity questions are predictive of the nutritional outcomes of older household members, but are not consistently related to the diets of children. In contrast, poverty predicts nutritional outcomes among both preschoolers and adults. Moreover, there are interesting differences between racial and ethnic groups, with the effects of poverty being generally greater for black and Hispanic children than for whites, while the effects of food insecurity are largest for the Hispanic elderly.

We note one important caveat to our conclusions from the outset. Most studies that examine food insecurity use a summary measure based on a series of questions. The NHANES III does not contain the entire series of questions, so a direct examination of the summary measure cannot be undertaken. However, the questions that are available in NHANES III are very similar to those available in the CPS, and these questions are very highly correlated with the summary measures. Thus, we interpret our results to be a strong indicator of what would be obtained if the entire set of food insecurity questions were available in NHANES III.

The rest of the paper is organized as follows. We describe the data set in Section 2 and the construction of variables in Section 3; Appendix A gives significantly more details about variable construction. Section 4 provides a regression analysis of the relationship between the variables and measures of nutritional status. We end with a discussion and conclusion in Section 5.

Section snippets

Hunger, nutritional deprivation, and food insecurity

There is some confusion in the literature about what is meant by “hunger”. Writing in the Journal of the American Dietetic Association (Lenhart and Read (1989), p. 1269) state that hunger is “a condition resulting from chronic under-consumption of food and/or nutritious food products”. However, frank starvation is very rare in the United States. The 1984 President’s Task Force on Food Assistance chaired by Clayton LaForce was established to look into other issues of concern, including reports

Data

We rely on data from the National Health and Nutrition Examination Survey III for our analysis. This national survey was conducted between October 1988 and October 1994 and included nearly 34,000 respondents, aged 2 months and older. The survey over-sampled blacks, Mexican Americans, younger children, and older persons to assure adequate representation and includes weights to make the sample nationally representative.

NHANES III collects information in the usual survey domains, including

Methods

We estimate models of the following form:Outcomei=α+Measureiβ+Xiγ+εiwhere Outcomei is a consumption or nutritional outcome measure, Measurei a vector of poverty or food insecurity measures, and Xi are other characteristics that describe the composition of the sample. The coefficients β measure the differences in outcomes associated with poverty and food insecurity.

We estimate the regressions separately for the 2−5, 6−11, 12−17, 18−64, and 65+ age groups. For each subgroup, we first estimate the

Results

Regression estimates are presented in Table 3, Table 4, Table 5, Table 6, Table 7, Table 8. Each panel/column combination represents a separate regression. For example, Table 3 presents the results for 16 different regressions. We withhold the other coefficient estimates for the sake of brevity.

Table 3 displays results for the 2−5 age group. The table indicates that children in poverty have less healthy diets (by 1–2 points on the HEI), are three percentage points more likely to be overweight,

Discussion and conclusions

Significant resources have been devoted to developing measures of food insecurity and hunger in the United States. This study adds to the growing literature that seeks to investigate the meaning of these measures by comparing the estimated effects of food insecurity indicators to those of the traditional measure of material hardship, poverty.

Our results lead to several conclusions. First, poverty is predictive of poorer nutritional outcomes among preschoolers and adults, though not among school

Acknowledgements

Bhattacharya and Haider gratefully acknowledge the financial support from the USDA/Institute for Research on Poverty Small Grants Program. The authors thank Alison Jacknowitz for superb research assistance and are grateful for the helpful comments from the participants at the USDA Small Grants Workshop in Madison, Wisconsin; the USDA Small Grants Conference in Washington, DC; the Dartmouth Labor Seminar; and the University of Maryland Demography of Inequality Seminar. All opinions expressed

References (26)

  • W.H. Dietz et al.

    Symposium on the identification and prevalence of undernutrition in the United States: introduction

    The Journal of Nutrition

    (1990)
  • E. Kennedy et al.

    The healthy eating index: design and applications

    Journal of the American Dietetic Association

    (1995)
  • D. Rose

    Economic determinants and dietary consequences of food insecurity in the United States

    The Journal of Nutrition

    (1999)
  • K. Alaimo et al.

    Food insufficiency and American school-aged children’s cognitive, academic, and psychosocial development

    Pediatrics

    (2001)
  • M. Andrews et al.

    Household food security in the United States in 1995: results from the food security measurement project

    Family Economics and Nutrition Review

    (1998)
  • Andrews, M., Nord, M., Bickel, G., Carlson, S., 2000. Household Food Security in the United States, 1999. Economic...
  • Bickel, G., Nord, M., Price, C., Hamilton, W., Cook, J., 2000. Guide to Measuring Household Food Security, Revised...
  • Borjas, G., 2002. Food Insecurity and Public Assistance, NBER Working Paper #9236, NBER, Cambridge...
  • Citro, C., Michael, R. (Eds.), 1995. Measuring Poverty. National Academy Press, Washington,...
  • Currie, J., Bhattacharya, J., 2001. Youths at nutritional risk: malnourished or misnourished? In: J. Gruber (Ed.),...
  • Cutler, D., Glaeser, E., Shapiro, J., 2003. Why have Americans become more obese? Journal of Economic Perspectives,...
  • Derrickson, J., Anderson, J., Fisher, A. 2000. Concurrent Validity of the Face Valid Food Security Measure. Discussion...
  • Dixon, L., Winkleby, M., Radimer, K., 2001. Dietary Intakes and Serum Nutrients Differ Between Adults from...
  • Cited by (403)

    • Developing New Scales of Personal Food Security

      2023, Journal of the Academy of Nutrition and Dietetics
    View all citing articles on Scopus
    View full text