The FAO indicator—This indicator is based on national level food balance sheets. It is published annually and, among other things, is used to monitor the Millennium Development Goal (MDG) for hunger. It estimates on a global scale the number of persons in a country whose daily food availability does not provide the minimum amount of energy (kilocalories). It can be criticized for (a) the possibility of errors in food balance sheets caused by uncertain data received from national level data inputs and (b) possible biases in the parameters used to reflect the inequality of consumption within countries (the coefficient of variation). de Haen and Klasen concluded that the FAO method, in principle, was sound theoretically, but that there were major uncertainties and gaps in the data base, so that at present its accuracy is very much open to question. Also, it does not generate actionable information to identify and monitor priorities at the national and sub-national levels.
Food consumption surveys—Nationally representative household surveys are an increasingly important source of data on food security. They are being conducted more frequently and with rising accuracy. Information on food consumption is derived by converting food expenditure information into consumption quantities and calories. Surveys provide a more direct assessment of food energy deficiency at the household level, compared to the FAO method, and provide direct measures of the intra-national inequality of food intake. As with the estimates derived from food balance sheets, household surveys may face problems with data accuracy. There are also concerns about their high cost, timeliness, coverage, and comparability between countries and over time.
Anthropometric measures—These measures are based on nationally representative surveys such as the Demographic and Health Surveys (DHS) or integrated household expenditure surveys (e.g., living standard measurement surveys). They may include anthropometric measurements for all household members, or only for children under a given age (often for children less than 5 years of age), or for women and young children in the household. The data for children are compared to an international reference standard (WHO, 2006) to derive prevalence rates of stunting (low height-for-age), wasting (low weight-for-height), and underweight (low weight-for-age). They are actionable indicators that can be used to target specific interventions and to monitor changes in nutritional status as well as responses to programs and policy changes. There is increasingly good coverage of these indicators internationally and comparability between countries (especially for indicators derived from the DHS). One of the limitations of this indicator is that it is often collected only on children—and, in many cases, their caregiver (such as in the DHS). The data are also usually not collected yearly, preventing the monitoring of short-term trends. The DHS surveys also fail to gather data on important covariates such as income, although integrated household expenditure surveys that collect anthropometric data do have information on total expenditure (a good proxy for income). One advantage of anthropometric data is that they can also be used to derive indicators of overweight and obesity and provide some information on the nutrition transition. But there are also questions about their comparability over time and between countries. In particular, the nutrition transition leading to heavier children may erroneously suggest improvements in underweight (the key MDG indicators), and the method is extremely sensitive to assumptions about even small genetic differences in height and weight potential among populations.