Previous - Next
Introduction
Evidence of nutrition/productivity links
Women's BMI and productivity
Summary and conclusions
References
Discussion
E. Kennedy and M. Garcia
International Food Policy Research Institute, 120017th Street, N.W, Washington, DC 20036, USA
The paper reviews the evidence to date on the nutritional links with productivity and then goes on to provide results from a multi-country study on the effects of increasing body mass index (BMI) on productivity. The research relating nutritional status to work capacity is more consistent than the research linking nutritional status to productivity. None of the studies to date elucidate the pathways through which improved nutrition improves economic productivity. In addition, many of the studies that have been conducted on the nutrition/wage links have been based on samples that contained a disproportionate number of male subjects. The few studies that have disaggregated data by gender report different results for men and women.
Research conducted at IFPRI is presented to examine the trends in BMI for men and women across countries and for Kenya to examine the relationships between various measures of nutritional status - BMI and height - and energy expenditures in women. BMIs of men show a more consistent relationship with increasing household income than do the BMIs of women. In the case of the Gambia and Kenya, the mean BMI of women decreases with increasing household income.
One reason for the apparently low response of BMI to increasing household income in Kenya is the time allocation patterns of women. Women in the Kenya sample spend the largest proportion of their day in home production activities which are energy intensive.
In examining the relationship
between nutritional indicators and the time devoted to work, the results suggest a
significant, positive association between both BMI and height and the amount of time
devoted to work. In the models presented, both BMI and height appear to increase the
capacity to carry out work. It is difficult to value much of this work time since a
disproportionate share is devoted to home production activities. Some of the more classic
methods of measuring economic productivity, such as measuring wage rates, are not relevant
for women in this setting. The data from Kenya suggest that more appropriate measures for
specifying the value of women's work need to be developed in order to capture some of the
nutrition/productivity links which may exist.
Much attention has focused recently on the links
between nutrition and productivity. In particular, enhanced human capital is seen as a key
component of increasing agricultural productivity in developing countries where
labour is still a major input into crop production.
Improved health and nutritional status are two elements contributing to this improved human capital.
This paper reviews the evidence on
the nutrition/productivity links with a particular emphasis on body mass index
(BMI)/productivity relationships. The paper then discusses some newer research related to
women's nutritional status and links to productivity. The paper ends with a discussion of
the policy implications for future work in this area.
The classic work of Ancel Keys (Keys et al., 1950)
provided some of the earliest evidence on the nutrition/productivity links. Male
volunteers subjected to diets with progressively lower levels of energy dramatically
reduced their physical activity. It was thus assumed that the opposite would also be true;
that is, as nutritional status improved, the energy expenditure level of individuals would
increase.
It may be intuitively appealing to believe that better-nourished individuals are able to be more productive. However, the nutrition/productivity links are difficult to establish. Rather than assuming that improved nutritional status leads to increased productivity, it is equally plausible that increased productivity leads to increased income which in turn improves nutritional status.
Many of the studies in economics have dealt with this bias of simultaneous effects by developing models which predict the nutrition input variables based on exogenous factors such as prices and household demographic variables. Strauss (1986), using data from Sierra Leone, used the predicted household energy intake per capita to explain household farm production. The results suggest that household energy consumption was a positive, significant determinant of farm output.
A similar approach was used by Sahn & Alderman (1988) with 1980/81 data from Sri Lanka. Here again, predicted household energy consumption per capita was used as the measure of nutritional status and related to wage earnings. Since not all sample households earned a wage, a two-staged process was used in the analysis. First, a probit analysis was used to predict the probability of being a wage earner. This estimate along with the predicted household energy per capita was included in the wage equation. Curiously, household energy per capita was a significant, positive determinant of men's but not of women's wages. This differential result between men's and women's productivity is a finding which emerges in a number of the studies linking nutrition to productivity.
Both the Strauss (1986) and Sahn & Alderman (1988) studies relied on extant data and thus were limited to the use of household energy values as the only measure of individual nutritional status. Clearly, a measure of individual nutrient consumption and more importantly a measure of an individual's nutritional status would have strengthened the analysis. This was done by Deolalikar (1988) using data from India. The study used both individual energy intake and weight-for-height to explain wage rates and farm output. Energy intake was not a significant determinant of wage rates but weight-for-height was at the 5% level of significance. Similarly, weight-for-height, but not energy consumption, had a significant, positive effect on farm output. The author concluded that nutritional status, as measured by weight-for-height, is important in determining labour productivity.
Haddad & Bouis (1991) advanced our state of knowledge by employing individual energy intake, BMI and height measures in explaining agricultural wage rates. Equations to instrument individual energy consumption and BMI were developed and used in the wage equations. Height but not BMI was a significant, positive determinant of wage rates. The authors conclude that substantial lifetime losses may occur if adults who were stunted in childhood due to poor health and nutrition rely on agricultural wages as their primary source of income. The authors also go on to caution about the finding of a lack of an energy/wage rate association given 'a vast nutrition literature would argue that relatively strenuous agricultural labor can only be sustained (without weight loss) through relatively high caloric intakes.' This last statement may or may not be true depending, in part, on how leisure time is affected.
Immink & Viteri (1981) found that in the strenuous task of cane cutting in Guatemala, it was the leisure time which appeared to be most affected by inadequate energy consumption. Men with deficient energy consumption decreased the energy intensity of their leisure time activities but not the amount of energy or time expended at work. When increased dietary energy was made available, the men did not increase the supply of units of work but rather become more active in their leisure time.
In contrast to the research of Haddad & Bouis which modelled the determinants of agricultural wages, Thomas & Strauss (1992) examine the nutrition/productivity links using wage rates in urban Brazil as the measure of economic productivity. Four indicators of nutrition are used as explanatory variables: height, BMI, per capita energy consumption and per capita protein intakes.
The results indicate that height is a significant determinant of the wages of both men and women in urban Brazil. However, BMI is a positive, significant predictor of male but not female wages. The Brazil results also suggest that per capita energy and protein intake are significantly related to wages but the positive effect of energy disappears rapidly indicating that it may only be the very malnourished for whom energy is a limiting factor for wage rates. The authors conclude that health (through improved nutrition) does provide an important return to labour in Brazil.
A particularly rigorous piece of work linking nutrition data on the same individuals who participated in energy and protein supplementation trials in Guatemala in the 1970s to measures of work capacity and productivity 15-20 years later was recently completed (Chung, 1992). Results from this work indicate that long term nutritional status -as measured by height - was a strong determinant of an individual's work capacity but short-term nutritional status (BMI) was not. However, work capacity (as predicted from anthropometry) is not a significant determinant of agricultural wages in this setting. The author goes on to indicate that without strenuous tasks, it is highly unlikely that nutrition-related productivity effects will become evident. However, this statement appears to be in contradiction to the urban Brazil data (Thomas & Strauss, 1992) where the authors do report a nutrition-wage link.
There are inconsistencies in the
nutrition/productivity literature. First, the research relating nutritional status to work
capacity is more consistent than the research linking nutritional status to productivity.
None of the studies to date provide evidence that elucidates the pathways through which
improved nutrition improves economic productivity. In addition, many of the studies that
have been conducted on the nutrition/wage rate links have been based on samples that were
disproportionate in having so many male subjects. The few studies that separate data by
gender - Sahn & Alderman (1988) and Thomas & Strauss (1992) - report different
results for men and women. This further obscures the causal mechanism between nutritional
status and economic productivity. It is unclear why improved nutrition in men but not
women would improve productivity.