Policy Research WORKING PAP-ERS Agricultural Policles Agriculture and Rural Development Department The World Bank October 1992 WPS 1 009 Labor and Women's Nutrition A Study of Energy Expenditure, Fertility, and Nutritional Status in Ghana Paul A. Higgins and Harold Alderman Women's nutritional status is reduced greatly by certain kinds of energy-expending work (especially agricultural tasks) and by "maternal depletion syndrome" in women with high fertility. Policy ResearchWozkingPapes disseminate the findings ofwork in progress ant encourage the exchange of ideas among Bank staffand allothers ltC d in developmentissues.Thesepapers. distributedby theResearchAdvisory Staff,cany thenarnes of the authors. reflect onlytheirviews.andshouldbe used and cited accordingly.The findings, interpretations. and eonclusions are the authors'own.Theyshould not be attributed to the World Bank. its Board of Direetors. its management, or any of its member countries. Policy Research| 1 3 3 _~1N Agricultural Pollels WPS 1009 This paper - a product of the Agricultural Policies Division, Agricultural and Rural Development Department-is part of a largereffort in the department to monitor the impact of agricultural policies on rural poverty. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Cicely Spooner, room N8-039, extension 32116 (October 1992, 41 paZes). Economic approaches to health and nutrition tive nutritional decline. But the "maternal have focused largely on measures of child depletion syndrome" remains controversial. nutrition and related variables (such as Much of the evidence to date has been impres- birthweight) as indicators of household produc- sionistic - or the results of studies based on tion of nutritional outcomes. But when dealing small, nonrandom cohorts. with adult nutrition, economists have to address an issue that has generated tremendous contro- Higgins and Alderman used a two-step versy in the clinical nutrition literature. instrumental variables technique to get consistent estimates of the structural parameters. Energy That issue is heterogeneity in an individual's expenditure, as embodied in individual time energy expenditures. Preschoolers' energy allocations over the previous seven days, was expenditure also differs, but the differences are found to be an important determinant of small enough to be ignored. Not so for adults, women's nutritional status. Time devoted to whose waking hours are devoted mostly to labor agricultural tasks, in particular, had a strong activities the energy costs of which vary enor- negative effect. mously. Variables measuring time allocation to various types of labor tasks were used to proxy The results also appear to confinn the differences in energy expenditure. existence of a maternal depletion syndrome. Perhaps more important, evidence was found of Parity has also been hypothesized to be an a substarntial downward bias of the calorie- important detenninant of female nutritional elasticity estimate when the energy expenditure health in high fertility countries - with rapid proxies were excluded. reproductive cycling contributing to a cumula- ThePolicy ResearchWorking PaperSeriesdisseminates thefindings of work under way in theBank. Anobjectiveoftheseries is to get these findings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official Bank policy. Produced by the Policy Research Dissemination Center LABOR AND WOMEN'S NUThITION: A STUDY OF ENERGY EXPENDITURE, FER LIY, AND NUTRIONAL STATUS IN GEANA- Paul A. Higgins Tulane UniversitY and Harold Alderman Aoricultural Policies Division Agricultural and Rural Development Department World Bank * Project initiated while the authors were, respectively, research support specialist and senior research associate at the Comnell UniversitV Food and Nution Policy Program. Partial support for the project from the Social Dimensions of Adlustment project of the, World Bank, and from U.S. Agency for Intemational Development Cooporotive AgreementAFROOO-A-O045-00, is gratefully scknowledged. The uthors would like to thank Insan Tunall for conmments and suggestions. TABLE OF CONTENTS Page Introducion ................ I Methodology ................ 4 Data ................ 13 Results ................ 16 Conclusio.s and Discussion ................ 20 References ................ 21 LIST OF TABLES Tablel .... 26 Table2 .... 28 Table3 .... 30 Table4 .... 31 Table5 .... 33 Table6 .... 35 TableAl .... 36 TableA2 .... 38 TableA3 .... 40 1. Introducion |he study of nutrition has been among the more fruitful applications of the economic theory of household production. In addition to using nutrition as an indicator of welfare, economists have incorporated nutritional variables into studies of labor productivity and wages, poverty, health, and ferdlity (Behrman and Deolalikar). Yet there is a danger in such intellectual border crossings; while nutritional status reflects the state of health of an individual as influenced by the intake and use of food or nutrients (Gibson), nutrient intake is only one in a complex of determinants of nutritional status. Diseases and parasites, and the individual's genetic endowment bave come to be recognized as important covariates conditioning the body's utlization of ingested food. But energy expenditure has been virtually ignored in economic approaches to the subject. Nutritional status is largely the result of individual net energy bance (Beaton 1983b). It is determined, in other words, by the person's energy epndkure as well as her calorie intake. Treating malntrition solely as a problem of inadequate food or nutrient availability can lead to perverse results - - for example, in the extreme, food-for-work progrms may fail to improve the nutritional status of participants if the increased labor effort required offsets the effects of the additional food. At the least, it can paint too narrow a picture of the nutrition problem for planners, causing them to disregard the interactions of non-food policies with nutrition, or to overlook possible aveues for improvement, such as the development of labor-saving devices. From a statstical standpoint, of course, neglecting energy expenditure differences in a population is likely to intrduce statistical bia. Many studies of nuritional status have reported sang positve income effects in reduced form or hybrid nutrtion equatons, even when input such as food or nutrien, morbidity, and health variables are induded. Since income per se is not an input into nutrition production, it is reasonable to ask whether the magnitude, perhaps even the sgnificance, of thes results may be the result of ms- 1 2 specification. If aveag onergy expenditure per unit of time detcreases with income in a population (or with assets that are correlated with expenditures), the effect of reduced energy expenditure could be incorrectly attributed to rising income;' if leisure is a normal good, then the income effect on leisure demand would simply reinforce the impact of this misspecification. Weak or insignificant impacts of calories on nutritional status have also been reported in the literature (e.g., Alderman and Garcia for nutritional status of children). While other plausible explanations, notably errors in the measurement of calorie intakes, have been put forward to explain similar puzzling results, it could equally well be the result of bias due to specification error. Calorie intakes and requirements ,ro dilectly correlated with levels.of individual energy expenditure (James and Schofield), whUe the energy costs of activities are presumably negatively related to indicators of nutrtional stants such as body weight or adiposity, other things being equal. Thus, failing to account for differences in energy expenditure would tend to bias the coefficient on calories downward, quite apat ftom any memen error effects. 'he issue of energy expenditure is especially important in studies of adult nutrition, since adult enegy use can be expeced to vary systematicaily within a population depending on activity level.2 An individual's energy expendiure is determined by her basal metabolic rate (BMR), and by the energy cm of her daily activities. The former is stochastic, and generally unobserved, but correlated with A, gender, and body ma (James, Ferro-Luzzi, and Waterlow). On the other hand, the non toc component of energy expediu is a function of the individual's time uses, and the intensity wih wh she puru them nam nd Schofield). 'Bonis and Haddad discuss this possibility with respect to the estimation of calorLe demand. 2Thls in not to may that energy expenditure can always be safely ignored in child studies (see Beaton 1983b). 3 Note that this problem is relat, though not identical, to the question of heterogeneous nutriet requirements across individuals (and over time) which currently bedevils the clinical nutrition literature (see B&tn 1983a, b; also Dasgupta and Ray). Whether individual requirements are fixed or adaptive, however, they are well-predicted by the energy costs of current activity levels, along with the aforementioned covariates of BMR (James, Ferro-Luzzi, and Waterlow). Hence, including current engy expendiure, or some indicator of time use correlated with energy expenditure, along with age, height and gender seems a way of minimizing this source of bias in a nutrition regression, providing, of course. that the simultaneity of these choices can be appropriately modeled. This study examines the determinants of the nutritional stas of adult women, using household survey data from Ghana. Its main contribution lies in exploiting time use data to estimate the contribution of individual energy expenditure differentials in determining nutritional status. The role of energy expenditures in contributing to female malnutrition is potendally mom important in sub-Saharan Africa than anywhere else in the world. African women tend to spend a ratively higher proportion of their time performing physically demnanding tasks, with relatively less leisure time, due mainly to their central role in agricultural production and distribution, and a lack of labor saving devices (Mueller; Lawrence et al. 1985; Singh et al.; Lamba and Tucker; Mebrahtu). Marked seasonal swings in energy expenditure, as well as in body weight and composition, and food availability, have also been documented among African women, especially in nral areas (Lawrence t al. 1989; Reardon and Maton). A secondary focus of the analysis is the role of fertility. In general, weight increases wih pari. Among undernourished, bigh fertility populations, however, indicators of nutritional status based on weight may decline with increased parity (Adair). Sub-Saharan Africa's average fertility rate (6.5 per woman, compared to 2.7 for East Asia and the Pacific, 2.0 for South Asia, 3.3 for Latin America and 4 the Caribbean3) Is the highest in the world. While acceptance of the notion of a maternal depletion syn. -rme is not universal (Winikoff and Castle express skepticism, for example), recent empirical evidence from a variety of settings suggests that rapid reproductive cycling indeed contributes to maternal nutritional depletion in high fertility coui tries (Merchant and Martorell; Huffman, Wolff, and Lowell; Adair et al.; Merchant, Martorell, and Haas 1990a, b). Since none of these studies treated parity as an endogenous choice variable, however, the possibility of statistical biased results cannot be ruled out. 2. Methodology 2.1. Theoretical Model. The theoretical model used here is based on those employed by Rosenzweig and Schultz (1983) and Schultz. Household members are assumed to behave as joint welfare maximizers with respect to individual health status, consumption, and time allocation.' Hovsehold utility is derived from both purchased and home produced goods, including nutrition and health. A joint household preference function governing household decisions over this choice set takes the frrm: i(1) U = U(Hf, Ci, C', LI', I, A), i =l,..J, where H' is the health status of household member i, C,: is member i's nonfood consumption vector, Cf is i's food consumption vector, L' is i's leisure time, gs is an unobserved variable capturing tastes and norms, assumed to be exogenous to current consumption decisions, and I Is household size. 3According to the World Development Report 1992. 4This assumption has been criticized for ignoring bargaining between household members (e.g., Folbre (19861; Manser and Brown; McElroy and Horney). Different intra-household models can be tested using individual specific non- labor income. The distinction between the two model. is often small in empirical applications, however (cf. exchange between Folbre (119841 and Rosenzweig and Schultz (19841), s0 Occam's Razor would seem to favor the former. Tho relationship between the nutritional status of each household member and nutrient and health inputs (as conditioned by the individual's health endowment and the household and community health environment and infrastructure) is governed by a production function of the form: (2) H' H(N', A', B', T', FP; Di, S, M, 0i), i= ,...,I where Ni, Ai, B', and E are vectors of member i's recent nutrient intakes, morbidity episodes,5 use of health care services, and energy expenditure, respectively; P' is i's total parity; D' is a vector of other fixed, observable individual characeistes of member i affecting her nutritional status; S and M are vectors of household and community fixed factors, respectively, that affect th6 nutritional status of household members; and VI is i's (unobserved) health endowment. The household maximizes (1) subject to (2) and its full incame budget constraint, generating input demands that enter the right hand side of (2) and which take the general form: (3) Z F r(Y, P; D, s, M, O,i= 1,..., where Z is a placeholder for (Ni, A', Bi, Ei, F), Y is exogenous income, and P is the complete vector of prices, broadly defined to include time as well as money costs. Note that in this specification neither prices nor income etr direcy int the production of health. Instead, they affect (2) idirectly, via the demand for inputs. 2.2. Indviduol Heterogenei. 5Because diseases can reduce the absorption of nutrients consumed, as well as depressing the anpptite, while fevers raise metabolic rates, the morbidity indicator may be thought of as conditioning the nutrient intake variable. 6 Empirical applications of the above modea face several potential pitfalls. Perhaps best known is the problem of unknown individual heterogeneity - in terms of the current specification, the inabiity of researchers to observe ,qi. To illustrate, consider the example of a household in which some members are inherently more robust than others, and thus better able to weather short term shocks such as food shortages. Family members are likely to be aware of this, and in lean periods may choose to allocae relatively more food to those who most need it in order to survive. Typically, researchers cannot observe such differences in individual endowments; yet the observed levels of some health and nutrient inputs will undoubtedly vary according to this individual attribute V. The resulting correlation between the inputs and the error term in equation (2) will bias the coefficients if they a:e estimated using ordinary least squares. There are essentially three possible responses to this difficulty. The first is simply to estimate equation (2) by ordinary least squares, and live with the possibility of bias. A case can be made for doing so, since the bias may be small, and the remaining options, while consistent, are often relatively inefficient (see Buse).' Another option is to use fill information methods to estimate the production function and input demands as a simultaneous system. This was the approach chosen by Rosenzweig and Schultz (1983), for example, in their analysis of birth weight production with endogenous inputs in the United States. Guilkey et al. also used this method in their study of birth outcomes in the Philippines. Explicitly modeling the full structural system is appealing, but this requires the researcher to specify the complete structure of the model, possibly increasing the likelihood of specification error. Perhaps more to the point in many applications, they can make insupportable demands upon the data set. The remaining option - which is related to the second - is to use an instrumental variables (IV) esimator. This method also may impose heavy demands on the data set. It is often difficult in practce 'For this reason it is interesting to compare the estimates of the preferred model, presented below, with the OLS estimates which appear in an appendix table. 7 . to find idendfyg restrictions for more than one or two Wndognous haldth or nutriet inputs, when many more are usually required. Even when sufficient plausible restrictions are avalable to identify the model, the instruments may perform poorly, leading to esdmates of the strucwal coefficients which are imprecise. Idoally, the vector M In equations (3) should contain a compete set of prices and wages, as well as other community variables affecting the demand for nutrien and halth inputs; examples include roads, distance to nearest clinic, quality of available health care services, clima, prevalnce of disease vectors In the local water supply and environs, local dietary and other customary practices, and the like. Ofto these are not observed; with the excepdon of local market prices and some locational indicators this is unfortunately true of the data set we use here. As is true of many integrated household data sets, however, ours was generated using cluster sampling techniques. Each cluster represents a single market and a relatvly homogeneous group of households, and interviews within each cluster were conducted over a short period of time. As such, there is likely to be virtually no intra-cluster price variation, while al of the other variables in M are by ddetion constant within a cluster. This fact suggests the possibility of takng the efcts of these missing variables in the instrumenting equations into account using clustr fixed effects or similar techniques. lbls is discussed in the next section. 2.3. Jdeadfyng Me Effects of Mfsing Comnwday F=tors Consider the following esdmatig equons, which may be viewed a linear approximations to equations (2) and (3): (4) Hi = ZpH + DHP + S,NtH + MH8H + e + (5) DZ= Dzz+SZP +J4+ 4 where the v subscript indexes villages (clusters) and i indexes individuals. As before, Z represents the vector of inputs into the production of nutrition; D, S and M denote observable individual, household, and village fuctors, respecively, which affect nutrition production or input demands; the v and X tem are unobserved personal and village characteristics. If most elements of the vector ?# In equation (5) are unavailable, then the unobserved component ,Z may contain the bulk of the information explaining the use of some inputs, meaning that any predicted values derived from the estimated coefficients of equation (5) are likely to he inefficient instruments. If any of the included predetermined variables are cor:elated with coz, then the parameter estimates from (5) would also be biased. A possible solution to this problem would be to use a community fixed effecs model to estimate (5), whereby the data are deviated from their cluster means7: (6) 2Z . (DZ - D)Z + (Sz -SZ)IZ + i Equivalently, this equation can be expressed in terms of dummy variables: (7) 7 = Dz + SY + X + i where X, is a vector of cluster dummy vaiables, such that the jth element is defined to be unity if individual i lives in cluster j, and zero otherwise. As long as all obsertions within a given cluster are collected in the same time period, then equaion (7) Is nearly equivalent to the more richly parameterized equation (5). ?Strauss (1990) uses a similar method to deal with the problem of missLng household variables in a study of child nutrition in Cots d'Ivoire. 9 Thus, the total variance in equation (5) may be partitioned into the within-cluster sum of squares about the cluster means, and the sum of squares of the cluster means about the grand mean, and their res- pective effects treated separately. Because Mz and dZ are constant across households in a given cluster, they are differenced out when the data are deviated from their means and thus need not be considered in the estimation of (6). For this reason, there is no possibility of bias due to the correlation between M! and c.z in these regressions. Thus, ,Sz and -yz may be consistently estimated. If consistent estimates of these pararneters were the primary goal of this exercise, then cluster fixed effects would be fully appropriate. As a means of obtaining the IV estimator of the nutrition production function, on the other hand, using cluster fixed effects to construct instruments may not be the best choice. Note that the vector MZ is dropped when estimating (6); yet MZ is apt to contain many exogenous variables useful for identifying instruments, even though it may be relatively sparse for the reasons given above. Furthermore, the fixed- effects estimator of equation (6) can use only the information contained in the intra-cluster variation; inter-cluster variation is ignored. For these reasons, the fixed-effects estimator of (6) is likely to be imprecise. The cluster means of the observed Z4,., which implicitly contain much of the information on the effects of the unobserved community variables in explaining inter-cluster variations in input use, may be included as instruments in the estimation of equation (4) along with the predicted deviations. The approach taken ifn the present studv is slightly different. Pither than estimating equation (6) in fixed effects, the cluster mean of the LHS variables in equations (5) is included as an instrument in equations (5); in effect, SZzZ y + MZbz + w0 is replaced by Z,.' As discussed above, the community means implicitly contain information on the missing cluster characteristics useful for identifying the instruments. While the pathways by which the missing community variables affect input levels cannot 8A similar technique is used by Alderman and Garcia. 10 be identified using this method, the impact of these missing variables on the production of nutridon, as they operate through the demand for inputs, can be consistently estimated.9 2.4. Model Specification In this study, women's nutritional status is measured by Quetiet's body mass index (hereafter BMI), defined as (weight/height), the most commonly used indicator of nutritional status for non- pregnant, non-lactating adults.'° BMI has a very low correlation with height, but is highly correlated with adiposity (Gibson, Fogel). It is also highly correlated with many health-related indicators, including mortality risk (Waaler). Nonetheless, there is as yet little agreement on BMI's distribution in healthy populations, and hence of appropriate cutoff values for evaluating health and mortality risk. One classification, provided by James, Ferro-Luzzi, and Waterlow, suggests that BMI levels between 18.5 and 23 be considered normal, with values above 23 being overweight." Individuals falling below 18.5 are assigned one of three c?uxories, ranging from grado I, or mild to moderate energy deficiency for BMI between 18.S and 17, to grade m, or severe wasdng, for BMI below 16. 'It might be argued that while this technique accounts for the problem of Indlvidual heterogeneity, the related problem of unobserved cluster heteroge- neity is not addressed. That is, w, and wx may not be independent, so that Lnalumion of the cluster means in the first stage estimation of the ZA may not result in instruments which are fully independent of the error tor in equation (4). Most of the obser-ed Lnter-cluster variation in nutritional status, however, is likely the result of fluctuations in levels of input use. These, of course, would be captured in the first-stage prediction. of the instruments. Moreover, Lndlcators are included in the structural nutritional status equation for agro-ecological zones, urban and semi-urban areas, and the capital and other relatively privileged cities. These should account for most of the remaining unexplained cluster heterogeneity. '*On the relative merit of dlfferent indices for adult nutritional status, cf. Gibson, and the references therein; also Smalley et al. ' Other cutoff values have been suggested by Royal College of Physicians; Dugdale; Paynct and Health and Welfare Canada. 11 An obvious input to nutrition is the individual's intake of calories in the previous period. In the present study this is not directly observed, and is proxied by per capita household calorie availability, as well as by covariates such as age and height."2 Calorie availability derives from the household's food demand, and thus flows from the interaction of household tastes and preferences with the budget constaint. The latter is de -nmined by the household's assets and human capital, relative prices, and the opportunities and limitations imposed by locale and season. Morbidity is the result of exposure to pathogens and parasites, as modified by individual choices and characteristics, and by the availability and quality of health care services. Nutrition affects susceptibility to illness, through its influence on the immune system, as well as being affected by it. Exposure, on the other hand, is largely a function of the level of community sanitation and the type of water supply, as well as by local prevalence of disease organisms, and local customs and practices. Demand for health care services is also a function of exogenous income and prices, as well as individual and household characteristics. However, availability of, and distance to, health care facilities are often more important determinants in developing countries, particularly in rural areas (see Gerler and van der Gaag). Quality of services, and the extent and diffusion of knowledge about health care practices within the community, may also play important roles but are generally not observed. Energy expendiure, as previously mentioned, is determined by the individual's basal metabolic rate, and by the energy costs of her daily activities. The former may be proxied by age and height (James, Ferro-Luzzi, and Waterlow); the latter is determined by the individual's time uses and intensities. 'hese are fuictions, in turn, of individual tastes, relative prices (including the value of time), othr 12 Adult height is generally considered predetermined since it is largely a product of early childhood nutrition, in interaction with the genetic potential of the individual (Martorell). Height may also be influenced by nutrition during adolescence, especially if pregnancy occurs before growth is completed (see Kennedy and Bentley and the references therein). 12 individual and household characteristics, and all of the local factors affecting the individual's choice set, including community norms, infrastructure, cropping patterns, climate, and season. Parity, too, is influenced by individual preferences, as well as prices. It is also a function of community standards and expectations, and of the woman's access to, and knowledge of, birth control methods. These latter factors are covariates of village infrastructure and locational variables, as well as numerous unobserved community characteristics affecting the state, and diffusion, of knowledge within the community (Bollinger). Thus a number of variables that enter the nutrition production function are significantly influenced by the community's environment and infrastructure, knowledge base, and norms and practices, all of which are unobserved. These include recent days of illness, consumption of heal care services, hours devoted to tasks demanding various levels of energy expenditure, and parity. Ean of these variables was ; Instrumented using cluster average values, in addition to other exogenous variables."3 Calorie availability, by contrast, was identified in the usual fashion. Additional individual characteristics included in the nutrition production function include age, intended to pick up developmental processes, height," which may proxy unobservable genetic endowments that contribute to overall physical robustness (Martorell), and education dunmmies which, for any given level of inputs, influence the knowledge and efficiency with which they are used in producing nutrition. For the most part household attributes, including the age-gender composition, information about the household head, household assets, and the educational stock of its members, enter the nutrition '3To avoid reintroducing indLvidual heterogeneity in these regressions, non-self cluster mean were used, calculated over all households in the cluster other than that of the individual in egstion. "Both age and height may also capture the residual effects of previous deprivation# e.g., the calamitous drought and famine which hit Ghana in 1983- 84. 13 equation via the instruments rather than directly in the structural production equation. This is in keeping with household production theory, which views nutritiopal status as the outcome of the levels of the proximate inputs chosen by the individual household members. The same holds for many of the observable commnunity factors. Nevertheless, household size was included directly in order to capture potential scale economies; and dummy variables for the Savannah and Forest agro-ecological zones, and for Accra and Kumasi, the major urban centers, were also included in order to account for any additional regional disparities not measured elsewhere. Finally, quarterly dummies were included to account for possible seasonal effects. Seasonal swings in nutritional status are certainly expected in Africa, particularly in rural areas (Lawrence et al. 1985). However, while the prices of most of the major staples exhibit marked seasonality, the impact of this on nutrition should be fully accounted for by the calorie availability pathway. Seasonal differences in disease prevalence, too, ought to affect nutrition indirectly, via the illness instrument. However, quarterly dummies were included in order to allow for the possibility of additional, unobserved seasonal factors. 3. Data The model is estimated with data from the 1987-88 Ghana Living Standards Survey (hereafter GLSS), a nationally representative, self-weighing random sample of nearly 3200 households, or roughly 15,000 people.'5 Community (cluster-level) prices were gathered for most of the calorie-dense foods and some nonfood items in a complementary market survey that was undertaken at the same time as the household interviews. Field staff made three purchases of each commodity and recorded the prices paid. Subsequently the purchases were weighed, giving a unit price. Missing prices were handled where I5For a discussion of the GLSS survey design and sampling methodology, see Scott and Amenuvegbe (1989). 14 possible by using the price from the closest available cluster. Rice prices, which were not recorded on the questionnaire, were obtained from regional agricultural market data. The definitions of all variables used in the subsequent analysis are giveu in Table 1. Descriptive statistics on these variables are presented in Table 2. A major aim of this paper is to construct a useful proxy for individual energy expenditure. The time use module of the GLSS, administered to all household members 7 years of age and older, repors hours devoted to "main" and "secondary" jobs during .he seven days prior to interview, as well as the time spent in the home performing non-market oriented tasks (e.g., preparing meals, fetching water and wood etc.) during the same period. Seventy-two percent of the women worked outside the home, and of these, 21 percent reported working at least two jobs in the previous seven days. ThIs, of course, is in addition to a weekly average of more than 20 hours of labor in the home, a category which more than 95 percent of the sample reported performing. Agriculture dominates the activities of Ghanaian women. Forty-three percent reported that agricultural tasks constituted their main job in the previous week; of those reporting a secondary job, 40 percent said that this, too, was in agriculture. The predominance of agricultural employment is, however, not apparent in the mean hours in agriculture reported in Table 2; those who did perform non-agricultural work reported more time in that activity and thus the average time spent in non-agricultural activities exceeds that in agriculture. It should be noted that the distinction between market time ("labor supply") and nonmarket time, often made in time allocation studies, is not germane to this discussion. Instead, what is required is that the average energy intensity of an hour's labor in any activity be more homogeneous within the given job ciass than across classes. Obviously, the more similar with respect to energy intensity the activities are within classes, and the more dissimilar they are between classes, the better job the time allocation 15 variables will do at representing variations in energy expenditure within the sample, and thus the population. It was decided to aggregate hours into just three categories for use in the regression analysis: agricultu2l, nonagricultural, and home labor. The low level of technology which prevails in Ghanaian agrculture, and the virtual lack of animal traction, assures that most if not all agricultural activities are labor-intensive and relatively demanding. On the other hand, the available GLSS time use data are insufficient to distinguish among gradations of the physical exertion required in the other occupations - e.g., between sales and service jobs. Thus the agriculture/nonagriculture distinction was a nawal one in this case. Home labor was included because it is such a significant component of women's time allocation in Ghana. Table 3 illustrates the sample distribution of nutritional stats for female respondents 18 years of age and above, using the categories suggested by James, Feffo-Luzzi, and Waterlow. By this stan- dard, 17 percent of women in Ghana during the sample period could be classified as undernourished (BMI less than 18.5); roughly two percent suffered from severe (Grade M) caloric deficiency, while the remainder fell into less acutely dangerous brackets. Table 3 also makes clear that rural inhabitants are generally thinner than their urban counterparts: rural women are observed in the lowest categories at rates twice that of urban dwellers. Wbile a significant share of Ghanaian women are undernourished, a surprisingly high proportion appear to be overweight, particularly in urban areas.16 This may simply reflect genetic differeces between the Ghanaian population and those used to derive the cutoff values. It may also reflect dietary habits which have lagged behind behavioral changes as the society becomes more sedentary. While less of a conoern ' Using GLSS data Alderman (1990) reported that males in Ghana are leaner overall, but that proportionately more women appear to be suffering from the severer grades of undernutrition, as well as from obesity. 16 than undernutritdon, this apparent high prevalence of obesity nevertheless may pose a potentially serious health problem. Table 4 shows the joint distribution of rural and urban BMIs, heights and weights by predicted household per capita expenditure decile (a proxy for permanent income)."7 Height, weight, and BMI all fail to display a consistent relatonship with per capita expenditures at the means. These observations, however, must be tempered by the knowledge that there is a substantial spatial component to the social disparities observed in Ghana: income, wealth, infrastructure and services, educational opportunities, and agronomic potential all tend to follow the same south-to-nortb gradient exhibited by the prevailing rainfall pattern. There are also ethnic differences which, while far more complex, correspond in a rough way to the simple coast-forest-savannah division. 4. Resuls Results from several alternative specifications of the nutritional status regression are presented in Table 5. Overall, the paten displayed by the coefficients on the time use variables tends to support the hypothesis that individual time allocation plays an important role in determining female nutritional status. They also are consistent with the interpretaion that time use variables, appropriately disaggregated, are useful for proxying the individual energy costs of routine daily activities. The coefficients indicate, for example, that an additional ten-hour day per week devoted to agricultural labor would result, ceteris paribus, in an expected reduction in BMI of 0.64 (nearly 15 percent of the range between obesity and the cutoff for mild chronic energy deficiency). By conast, the non-agricultural time "1The log of per capita household expenditures was predicted am a function of fixed household assets and human capital variables, as well as the cluster mean of the dependent variable (Alderman and Hiqgins). In the text below, this variable will be termed income to avoid confusion with onergy expenditures. 17 allocation variable has a positive coefficient, which is consistent with the more sedentary nature of non- agricultural labor in Ghana. The positive net effect of home labor, wbile smaller in absolute value than the other time use coefficients, was not expected. Some of the tasks elicited in this category (e.g. hauling water, gathering fuelwood) would seem to be as demanding as most farm tasks; others, such as childcare, cooking and clothes mending, are far less energy intensive. The comparatively small positive coefficient may reflect the net impact of disparate activities with widely varying energy demands which were inappropriately aggregated into a single category. Unfortunately the data allow no way of teasing these categories apart. The results also imply that the elasticities of other explanatory variables may be substantially biased if the variation in individual energy expenditure is not taken into account. The coefficient for calorie availability, in particular, is small, unstable, and uniformly insignificant in all specifications which exclude the time use variables. When the latter are included, by contrast, the magnitude of the parameter estimate for calories jumps dramatically, and becomes statistically significant (p-0.039)." The direction of bias when time allocation is excluded Is consistent with the discussion of the joint interaction of individual calorie consumption, requirements, and energy expenditures in the introduction. Similar patterns are observed for several other variables. For instance, coefficients for the forest and savannah zones, and urban and semi-urban areas, which "explain" a very large proportion of the variance of the dependent variable in the parsimonious models (columns [11 through [3]), drop by an average of eighty percent in absolute value in model 4 which includes the time use variables. Moreover, none of coefficients of the regional and sector dummy variables remain significant by the usually accepted crieria when time allocation is included. This indicates tha most, if not all, of the regional pattern observed in female nutritional status that is not explained by calories, parity, morbidity, and the "Moreover, this result in also robust across several other permutations (not reported here because of space limLtations), so long as the tim use variables ara Lncluded. 18 availability of health care services, workcs through regional differences in the pattern of women's daily activities, and the energy costs that they imply. Parity is associated with a negative net effect on nutritional status. The coefficient suggests that, on average, each additional pregnanc carried to term by a woman in Ghana implies an expected drop in her BMI of between 0.25 and 0.40, or roughly half of the impact of an additional day of work in agricultural labor each week, holding other factors constant. The coefficients on the dummy variables for secondary and post-secondary educational attainment shrink approximately 60 percent in absolute magnitude upon the introduction of parity into the regression, in the process becoming smaller than their estimated standard errors. This suggests that much of the apparent ameliorative effect of education on women's nutrition opera through its impact on ferdlity, rather than by directly influencing the health technology practiced in the home. A set of analogous regressions, which include an exogenous income instrument in place of calories, are presented in Table 6. As previously noted, income is not a proximate input in the produc- tion of nutritional outcomes. In effec, then, these equations are hybrids of structral and reduced forms; while formally inconsistent with household production theory, they may be usefully interpreted as conditional production fumctions (Strauss 1990). They are of interest primarily to shed further light upon the nature of the specification error discussed above. These regression results are generally quite similar to those reported in the previous table. lhe notable exception is that, in distinction to the calorie variable in Table S, the coefficient on incomes falls from 1.45 to 0.62 - i.e., by nearly sixty percent of its value - - as variables measuring morbidiy, health care, parity, and time uses enter the model sequentially. When predicted calories is added", the income variable shrinks to 0.2, and is no longer distinguishable from zero by the usual criteria. 9ThLi result is not presented here due to apace limitations. 19 ibis result illustrates an important difference in the impact of excluding time allocation variables in the two sets of equations. As discussed above, the relationship between caloLies and BEh is unambiguously positive; likewise, the partial correlation between energy expenditure and BM! is unambiguously negative. Since calorie intake and energy expenditure are positively correlated, excluding indicators of the latter will bias regression coefficients of the former in a negative direction. In the case of regressions involving an income instrument, on the other hand, the interactions are more complex. While energy expenditure is still negatively related to BM!, the sign of the relationship between energy expenditure and income is murkier, and may well be negative when occupation is not held constant. As suggested in the preceding discussion, BMI probably bears a nonlinear relationship to nutritional health, with both the lower and upper tails implying serious, if rather different, health problems. To focus on the likelihood of serious consequences due to energy deficiency, alteative specifications of both sets of regressions were run to model the probability of being malnourished. The dependent variable in these equations is a dummy variable which takes on a value of unity if BM! is below 18.5, and zero otherwise. The maximum likelihood probit estimates of these models are presented in appendix table 2. Similarly, since the data presented in Table 3 suggest that obesity may be a significant source of risk, even in a poor nation such as Ghana, probit regressions reported in appendix table 3 report the probability of a women being obese.' In rerds to undetition, lack of calorie availability seems to counts for less than overwork, lack of utilization of health care and parity. calorie availability (and income), however, do contribute to the probability of obesity. hisis consistent with the existence of non-linearities in the central results modeled and reported in Tables 5 and 6. Alternative, or additionally, the population of overweight X The categories, although mutually exclusive, are not exhaustive. The signs of most variables in Table A2 are expected to be the opposite of those in Table A3. Unlike the main results, the T statistics are not corrected for first-stage prediction error and should be taken as suggestive only. 20 women in Ghana are more homogeneous than those who are in chronic energy deficit and, hence, our prediction equations are more precise. 5. Conclusions and Discussion The poor health and nutrition of African women are among the most serious problems facing governments and development planners there. While statistics on morbidity and mortality in Africa are rare, there is ample evidence that women in Sub-Saharan Africa suffer extremely high rates of morbidity and mortality. Major causal factors include insufficient and seasonally flucating nutrient availability, high levels of energy expenditure, and high fertility, as well as endemic diseases and poor provision of health care services (cf. Kennedy and Bentley, and the references therein). The results of this study suggest (i) The demanding physical labor performed oy Ghanain women, especially in agriculture, but possibly also in food preparation (such as processing grain and pounding roots to make staples such as fifu) and other household tasks, has a significant negative effect on their nutritional status. Moreover, such nutritional impacts may affect the prospects for their children as well. This imnplie that the introduction of labor saving dev ^7 r , Ay have a direct impact of nutrition similar to the increase of food consumption. It also suggests that the energy consequences of public works projects involving women in physical labor be considered, especially when the programs are design to in response to chronic or acute food shortages. (ii) The extremely high fertility rate of Ghanaian women, in concert with disease and inadequate health care and nutrient availability, also takes a measurable toll. Thus, increased education for women and family planning programs may be considered facets of nuuritional planning as well as programs justified in their own right. It also underscores the importance of supplementation programs for pregnan and lactatng women for nrual women and women residing in the Savannah regions of the country. 21 REFERENCES Adair, L. 1991. "Weight-for Height and Body Mass Index in Nonpregnant Women." In: K. Krasovec and M.A. Anderson (eds.), Maternal Nutrition and Pregnancy Outcomes. Results of the meeting 'Maternal Anthropometry for Prediction of Pregnancy Outcomes" held 23-25 April 1990. Washington DC: Pan American Health Organization. Adair, L., B. M. Popkin, E. Bisgrove, and C. Barba. 1990. A Longitudinal Analysis of the Patterns and Determinants of Women's Nutrition In the Philfppines. Report No. 1, Maternal Nutrition and Health Care Program. Washington DC: International Center for Research on Women. Alderman, H. 1990. Nutritfonal status in Ghana and Its Determinants. Washington, DC: Cornell Food and Nutrition Policy Program Working Paper No. 1. Alderman, H. and M. Garcia. 1992. Food Security and Health Security. Explaining the Levels of Nutrition in Pakistan" World Bank Policy Research Working Paper #865 February 1992. Alderman, H. and P.A. Higgins. 1992. Food and Nutrftional Adequacy In Ghana. Washington DC: Cornell Food and Nutrition Policy Program Working Paper No. 27. Beaton, C. 1983a. "Adaptation to and Accomodation of Long Term Low Energy Intake." In: E. Pollitt and P. Amante (eds.), Current Topics In Nutrition and Disease. Vol. 11, Energy Intake and Acdvly. New York: Alan R. Liss for UNU. Beaton, G. 1983b. Energy in human nutrition: Perspectives and problems. Nution Revlewx 41(11): 325-340. Behrman, J. and A. Deolalikar. 1988. "Health and Nutrition." In: H.B. Chenery and T.N. Srinivasan (eds.) Handbook of Development Economics. Vol. 1. Amsterdam: North-Holland Publishing Co. Behrman, J., A. Deolalikar and B.L. Wolfe. 1988. Nutrients: Impacts and deerminants. World Bank Economic Review 2 (3): 299-320. Bollinger, L. 1992. Diffusion and adoption of contraceptive technology In developing countries. Presented at the Annual Meeting of the American Economic Association (mimeo). 22 Bouis, H. E. and L. J. Haddad. Forthcoming. Comparing calorie-income elasticities using calories derived from reported food purchases and a twenty-four hour recall of food intakes: An application using Philippine data. Juarnal of Developrnent Economics. Buse, A. 1988. Book Review. Journal of Economic and Business Statdsdcs 6: 141-142. Dasgupta, P. and D. Ray. 1990. "Adapting to undernutrition: The clinical evidence and its implications." In: J. Dreze and A. Sen (eds.), 7he Political Economy of Hunger. Dugdale, A.E. 1985. Family anthropometry: A new strategy for determining community nutrition. 7he Lancet: 672-673. Fogel, R.W. 1991. New Sources and New Techniques for the Study of Secular Trends In Nutridonal Status, Health, Mortality, and the Process of Aging. Cambridge, MA: NBER Working Paper No. 26. Folbre, N. 1984. Market opportunities, genetic endowments, and intrafamily resource distribution: Comment. American Economic Review 74 (3): 518-520. Folbre, N. 1986. Cleaning house: New perspectives on households and economic development. Journal of Development Economics 22 (1): 5-40. Gerdler, P., and J. van der Gaag. 1990. lhe Willingness to Pay For Medical Care: Evidence from TVvo Developing Countries. Baltimore: Johns Hopkins University Press. Gibson, R.S. 1990. Principles of Nutritional Assessment. New York: Oxford University Press. Guilkey, D., B. Popkin, J. Akin, and E. Wong. 1989. Prenatal care and pregnancy outcomes in Cebu, Philippines. Journal of Development Economics 30: 241-272. Haddad, L., and H. Bouis. 1991. The impact of nutritional status on productivity: Wage evidence from the Philippines. Oxford Bulletia of Economics and StatIstics 53 (1): 45-69. Health and Welfare Canada. 1988. Canadian Guidelines for Healthy Weights. Report of an expert committee convened by the Health Promotion Directorate, Health Services and Promotion Branch, Health and Welfare. Ottowa. Huffman, S., M. Wolff, and S. Lowell. 1985. Nutrition and fertility in Bangladesh: Nutritional status of nonpregnant women. 7he American Journal of Cainical Nutrion 42: 725-738. James, W.P.T., A. Ferro-Luzzi, and J.C. Waterlow. 1988. Definition of chronic energy deficiency in adults. European Journal of Clinical Nutrition 43: 969-981. 23 James, W.P.T., and E.C. Schofield. 1990. Human Energy Requirements: A Manal for Planuers and Nutritonists. Oxford: Oxford University Press for FAO. Journal of the Royal College of Physicians 17 (Nov. 1, 1983, pp. 1-53). Special issue on obesity. Kennedy, E., and M. Bentley. Forthcoming. "Women's Health and Nutrition in Sub- Saharan Africa." In: A. Sorkin and I. Siralgedin (eds.), Food Policy, Health, Nutriton, and the COanging Environment. Lamba, C., and K. Tucker. 1990. Work Patterns, Prenatal Care, and Nutritional Status of Pregnant Subsistence Farmers in Central Malawi. Washington DC: International Center for Research on Women. Lawrence, M., J. Singh, F. Lawrence, and R.G. Whitehead. 1985. The energy cost of common daily activities in African women: Increased expenditure in pregnancy? American Journal of Clinical Nutrition 42: 753-763. Lawrence, M., F. Lawrence, T.J. Cole, W.A. Coward, J. Singh, and R.G. '7Jhitehead. 1989. 'Seasonal Pattern of Activity and Its Nutritional Consequence in the Gambia. In: D.E. Sahn (ed.), Seasonal Variability in Third World Agriculture: The Consequences for Food Security. Baltimore: Johns Hopkins University Press for IFPRI. Manser, M., and M. Brown. 1980. Marriage and household decision-making: Bargaining -analysis. International Economic le.'itw41 (1): 3144'.. Martorell, R. 1991. "Maternal Height as an Indicator of Risk." In: Maternal Nutrition and Pregnancy Outcomes. Results of the meeting "Maternal Anthropometry for Prediction of Pregnancy Outcomes" held 23-25 April 1990. Washington DC: Pan American Health Organization. Martorell, R., and T. Gonzalez-Cossio. 1987. "Maternal Nutrition and Birth Weight.' In: Yearbook of Physical Anthropology 30: 195-220. McElroy, M.B., and M.J. Horney. 1981. Nash-bargained household decisions: Toward a generalization of the theory of demand. International Economic ReWew 22 (2): 333- 350. Mebrahtu, S. 1991. "Women, Work, and Nutrition in Nigeria." In: M. Turshen (ed.), Wom&.. and Health in Africa. Africa World Press, Inc. Merchant, K., and R. Martorell. 1983. Frequent reproductive cycling: Does it lead to nutritional depletion of mothers? Progress In Food and N4trition Science 12: 339-369. 24 Merchant, K., R. Martorell, and J. Haas. 1990. Consequences for maternal nutrition of reproductive stress across consecutive pregnancies. American Journal of Clinical Nutrition 52: 616-620. Payne, P. 1986. "Appropriate Indicators for Project Design and Evaluation." In: Food Aid and the Well-Being of Children In the Developing World. New York: UNICEF. Pitt, M. 1983. Food preferences and nutrition in rural Bangladesh. Review of Economics and Statistics 65: 105-114. Reardon, T., and P. Matlon. 1989. "Seasonal Food Insecurity and Vulnerability in Drought-Affected Regions of Burkina Faso." In: D.E. Sahn (ed.), Seasonal Variability in Third World Agriculture: The Consequences for Food Security. Baltimore: Johns Hopkins University Press for IFPRI. Rosenzweig, M., and T.P. Schultz. 1983. Estimating a household production function: Heterogeneity, the denand for health inputs, and their effects on birthweigbt. Journal of Political Sconony 92: 723-746. Rosenzweig, M., and T.P. Schultz. 1984. Market opportunities, genetic endowments, and intrafamily resource distribution: Reply. American Economic Review 74 (3): 521- 522. Rutishauser, I.H.E., and R.G. Whitehead. 1972. Energy Intake and Expenditure in 1-3 Year Old Ugandan Children Living in a Rural Environment. Brish Journal of Natridon 28 (1): 145-i52. Schultz, T.P. 1984. Studying the impact of household economic and community variables on child mortality. Population and Developmnent Review 10 (supp.): 215-236. Scott, C., and B. Amenuvegbe. 1989. Sample Designsfor theUlving Standards Surveys in Ghana and Mauritania. Living Standards Measurement Study Working Paper No. 49. Washington DC: World Bank. Singh, A M., A.M. Prentice, E. Diaz, W.A. Coward, J. Ashford, M. Sawyer and R.G. Whitehead. 1989. Energy expenditure of Gambian women during peak agricultural activity measured by the doubly-labelled water method. Britsh Journal of Nurition 62. Smalley, K. J., A. N. Knerr, Z. V. Kendrick, J. A. Colliver, and 0. E. Owen. 1990. Reassessment of body mass indices. American Journal of ainical Nurition 52, 405-408. Strauss, J. 1990. Households, communities, and preschool children's nutritional outcomes: Evidence from rural Cote d'Ivoire. Economic Development and Cldtural Change 38 (2): 231-262. 25 Strauss, J., and D. Thomas. 1989. The Shape of the Expenditure-Calorie Curve. Santa Monica, CA: Rand Corporation, and New Haven, CT: Yale University (mimeo). Svedberg, P. 1990. Undernutrition in sub-Saharan Africa: Is there a sex bias? Journal of Development Studies 26: _. Winikoff, B. and M. A. Castle. 1987. "The Maternal Depletion Syndrome: Clinical Diagnosis or Eco-Demographic Condition?" Paper presented to the International Conference for Better Health for Women and Children through Family Planning. Nairobi, October 1987. World Bank. 1992. World Development Report. Washington DC: World Bank. World Health Organization. 1985. Energy and Protein Requirements: Report of a Joint FAO/WHO/UNU Expert Consultadon. WHO Technical Report Series No. 724. * ~ ~~~ ~~ ~ ~ ~ *.a * * * * - *. * . -. 26 TABLE 1: Definitions of Major Variables Variable Definition BMr Body mass index: Wt (kg)/Ht2 (m) Age Reported age in years Height Measured height in centimeters EUation Dummies Primary Equals one if highest grade raepondent reports completing is in prmary school, and zero otherwise. Secondary Equals one if highest grsde respondent reports completing is in secondary school, and zero otherwise. Post-secondary Equals one if highest grade respondent reports completing is beyond secondary school, and zero otherwise. Household Size Number of members of respondent's household. Female-headed Equals one if respondent resides in a household headed by a woman, and zero o*-erwise. - Forest Zone Equals one if respondent's household is located in the Forest agro-ecological zone, and zero otherwise. Savannah Zone Equals one if respondent's household is located in the Savannah agro- ecological zone, and zero otherwise. 2ad Quarter Equals one if respondent was interviewed during the second quarter of the year (April-June) and zero otherwise. 3rd Quarter Equals one if respondent was interviewed during the third quarter of the year (July-September), and zero otherwise. 4th Quartr Equals one if respondent was interviewed dunng the fourth quarter of the year (October-December), and zero otherwise. Urban Equals one if respondent's household is located in an urban area, and zero otherwise. Semi-urban Equals one if respondent's household is located in a semi-urban area (population 2,000 or more), and zero otherwise. 27 TABLE 1: (continued) Variable Definition Accm Equals one if respondent's household is located in Accra (the national capital), and zero otherwise. Kumasi Equals one if respondent's household is located in Kumasi (capita of Ashanti Region), and zero otherwise. Log per cap- The average total calories available to the household ita calories during the two weeks prior to interview, divided by available household size, in natural logarithms. Log per cap- The total expenditures (cedis) reported on all goods its household and services in the two weeks prior to interview by expenditures all household members, including imputed value of home-produced foods, divided by household size, in natural logarithms. Number of days The number of days respondent reported being ill in the il in past four weeks prior to interview. 4 weeks Number of days The number of days respondent reported being suffi- icpitated, ciently il or injured as to prevent the performance pma 4 wecs -of her n6rm1 ac3tivitie. * - - ' *' Log health care The amount (cedis) spent by household for curative and expenditures, preventative health care goods and services for the past 4 weeks! respondent during the four weeks prior to interview, in natural logarithms. Parity The number of pregnancies carried to term by the respondent, including stillbirths. Agricultural The number of hours reported spent by respondent in houm, last the seven days prior to interview on agricultural 7 days' and forestry activities. Nonagricul- The number of hours reported spent by respondent in twal hours, the seven days prior to interview on all non- last 7 days' agricultural activities other than non-market oriented labor at home. Home labor The number of hours reported spent by respondent in hours, ast the seven days prior to interview on non-market 7 days oriented labor at home. * Endogenous variables. 28 TABLE 2: Means and Standard Deviationt of Major Variables Rund Urban Pooled Variable (na=1,977) (n-1,153) (n=3,130) BMI 20.888 23.026 21.673 (3.204) (4.951) Age 38.861 36.607 38.033 (15.989) (14.885) (15.628) Height 156.998 158.306 157.481 (6.267) (6.357) (6.330) Education Primary 0.097 0.101 0.099 Seondary 0.217 0.401 0.284 Post-secondary 0.009 0.061 0.028 Household Size 6.591 5.904 6.338 (3.837) (3.381) (3.690) Female-headed 0.328 0.410 0.358 .~~~~~~~ . - - Forest Zone 0.461 0.328 0.412 Savannah Zone 0.305 0.134 0.242 2nd Quarter 0.212 0.263 0.231 3rd Quarter 0.257 0.201 0.237 4th Quarter 0.258 0.159 0.222 Urban - - 0.368 Semi-urban 0.316 - 0.120 Accm - 0.256 0.094 Kumasi - 0.094 0.035 Log per cap- ita calories 7.656 7.555 7.619 available (0.675) (0.675) (0.677) Continued 29 TABLE 2 (Continued) Rural Urban Pooled Variable (n=1,977) (n'1,153) (n=3,132) Log per cap its household 10.647 11.102 10.815 expenditures (0.596) (0.623) (0.644) Number of days ill in past 3.388 3.S60 3.451 4 weeks (6.127) (6.218) (6.158) Number of days incapacit, 1.893 1.479 1.740 past 4 weeks (4.252) (3.S60) (4.014) Health care expenditures, 586.584 1,010.460 742.329 past 4 weeks (1,928.430) (2,830.070) (2,309.960) Parity 4.836 3.824 4.461 (3.475) (3.253) (3.430) Agriculral hours, last 16.172 3.345 11.447 7 days (15.896) (8.962) (15.081) Nonagicul- tal hours, 7.676 19.607 12.071 last 7 days (16.483) (24.658) (20.702) Home labor hours, last 20.686 21.111 20.843 7 days (11.867) (13.462) (12.479) Standard devions apear in parentheses below means. 30 TABLE 3: Distribution of Nutritional Status (Chronic Energy Deficiency') (Nonpregnant, Noilactating Ghanaian Women Age 18 years and Older) Rural . Urban BMI Range Classification (n= 1,977) (n= 1,153) (Percent) BMI > 23.0 Overweight 19.1 38.1 18.5 5 BMI ! 23.0 Normal energy reserves 61.2 50.5 17.0 s BMI < 18.5 Normal-to-mild CED 12.9 7.8 16.0 S BDM < 17.0 Mild-to-moderate CED 4.7 2.2 DM1 < 16.0 Severe CED 2.1 1.5 a See James, Feffo-Luzzi, and Waterlow. Classifications are somewhat imprecise for the mild and moderate CED levels because of normal variations observed in healthy populations (ibOa). 31 ALE 4 Distribution of BMI, Weight and Height by Income Decile') (Nonpregnant, Nonlactating Ghanaian Women Age 18 years and Older) RURAL URBAN Decile Weight Height BMI n Weight Height BMI n 1. 51.9 157.6 20.9 255 59.3 159.4 23.3 58 2. 51.0 157.1 20.7 263 54.6 158.5 21.6 50 3. 51.3 157.0 20.8 268 55.4 158.1 22.2 45 4. 50.7 156.9 20.6 244 54,1 157.8 21.8 70 5. 50.7 156.1 20.7 228 57.6 157.7 23.2 85 6. 51.5 156.9 20.9 207 57.7 157.1 23.3 106 7. 53.1 157.2 21.4 196 56.6 158.4 22.5 118 8. 51.3 156.8 20.8 155 55.8 157.9 22.4 159 9. 52.5 157.0 21.3 109 59.8 158.0 23.9 204 10. 56.1 158.5 22.3 54 59.9 159.5 23.5 258 a Weights and heights expressed in kilograms and centimeters, respectively. Rank ordering of individuals performed over the entire sample, rather than within rural and urban subsamples separately. Ranking was on the basis of predicted per capita household expenditures, an exogenous income instrument, where log of per capita expenditure was regressed on household assets, demo- graphic, community and regional characteristics, and quarterly indicators. 32 li2l]LEL£: Nutritional Satus Rgrehssionwtb (Womem a!es 18 yoas and older) Variable (1) (2) (3) (4) Intecept 13.132 IS.336 16.005 10.426 (2.798) (3.478) (3.663) (3.666) Age (years) 0.223 0.192 0.372 0.298 (0.022) (0.027) (0.061) (0.058) AgV squared -0.002 40.002 -0.004 -0.003 (0.0002) (0.0003) (0.0006) (0.0005) Height (cm.) O.0S -0.013 -0.018 4.019 (0.011) (0.014) (0.015) (0.014) Educaton Prinmay 0.860 0.466 0.269 0.449 (0.247) (0.325) (0.347) (0.324) Seondazy 0.643 0.536 0.202 -.01S (0.185) (0.227) (0.259) (0.243) Postsecondary I.S28 1.199 0.491 0.514 (0.432) (0.530) (0.596) (0.549) Household Sizo 0.104 0.125 0.156 0.190 (0.024) (0.030) (0.033) (0.032) Female-headde 0.221 0.271 0.174 0.361 (0.156) (0.214) (0.227) (0.232) Fores Zoe -1.050 -0.715 -0.S75 40.030 (0.164) (0.218) (0.233) (0.239) Savannah Zone -1.327 O.S91 O.S98 -0.230 (0.201) (0.291) (0.306) (0.305) 2ad Quater 40.246 -0.427 -0.443 -0.171 (0.188) (0.232) (0.244) (0.242) 3rd Quarter 0.398 0.147 0.103 0.314 (0.188) (0.236) (0.248) (0.241) 4th Quarter 0.002 40.037 40.106 0.231 (0.192) (0.234) (0.247) (0.251) Continued 33 T fle 5: (oondnued) Vaiable (1) (2) (3) (4) Urban 1.854 1.559 1.302 0.250 (0.174) (0.221) (0.244) (0.284) Sami-urban 0.532 0.488 0.411 0.084 (0.191) (0.236) (0.249) (0.240) Log per capita calories 0.306 0.322 *0.002 0.830 available (0.276) (0.348) (0.379) (0.402) Number of days ill in past -0.020 0.044 -0.078 4 weeks' (0.085) (0.090) (0.088) Log heath cr expenditures, -0.120' 0.129' 0.102' pat 4 week (0.028) (0.030) (0.028) arity- --*0.398 4.247 (0.119) (0.116) Agricultur hours, last - 0.064 7 days' (0.020) Nonagricul- - 0.058 twal huxv, * (0.018)* - * last 7 days' Home labor - 0.027 hours, last (0.019) 7 days R2 0.129 0.097 0.092 0.119 n 3,130 3,130 3,130 3,130 Endogenous RHS variable. 'Standard errors are located in parenhse beow estima. They, and the R2 ttics, are correted to account for the distbances in the first-stoa re ions (soe Kmenta, 684). Te asymptotic standard erors of the edinmat in tho probit regwesions ar not corrected, however. b Dependent variable is BMI. Sample mean of dependent varible is 21.67. 'Pregnant and lating wormn we excluded from samplo. dtimes 10.2 Maximum likelihood probit estmaton. Dependat variable equal one when BMI c 18.5, nd zeo ohewis. 522 cases equal one, and 2,608 equal zero. 34 TABLE: Conditional Nutrition tgrgiofAnb (Women ages 18 yea and older)' Variable (1) (2) (3) (4) Intercept 0.573 4.642 7.148 10.191 (3.115) (3.873) (4.209) (4.241) Age (year.) 0.215 0.192 0.312 0.294 (0.022) (0.026) (0.060) (0.058) Age squared -0.002 -0.002 -0.003 4.003 (0.0002) (0.0003) (0.0005) (0.005) Hoight (cm.) 4.002 -.015 -0.019 -0.019 (0.011) (0.013) (0.014) (0.013) Education Primary 0.754 0.540 0.402 0.403 (0.246) (0.303) (0.324) (0.309) Secondary 0.403 0.334 0.165 -0.057 (0.189) (0.219) (0.242) (0.233) Post-secondary 0.904 0.757 0.412 0.338 (0.444) (0.508) (0.555) (0.527) Household Size 0.176 0.182 0.193 0.180 (6.025) (0.029) ' (0.031r (0.-30) Female-headed 0.425 0.508 0.394 0.372 (0.159) (0.208) (0.233) Forest Zone -0.822 -0.562 -O.501 -0.069 (0.168) (0.208) (0.219) (0.228) Savannah Zone -0.878 4.392 -0.456 -O.151 (0.214) (0.277) (0.292) (0.295) 2nd Quarter -0.325 -0.453 -0.458 -0.226 (0.188) (0.218) (0.228) (0.230) 3rd Quarter 0.432 0.266 0.202 0.336 (0.187) (0.223) (0.236) (0.236) 4th Quarter 0.226 0.159 0.055 0.282 (0.195) (0.224) (0.239) (0.254) Urban 1.373 1.239 1.155 0.145 (0.184) (0.306) (0.226) (0.276) 35 TABLE: (continued) Variable (1) (2) (3) (4) Sani-ura P.395 0.412 0.394 0.036 (0.190) (0.219) (0.249) (0.230) Log per capita 1.450 1.239 0.889 0.624 expendi;r' (0.247) (0.306) (0.357) (0.368) Number of days ill in past -0.057 -0.027 -0.065 4 weeks' (0.080) (0.085) (0.084) Log health care expenditures, - 0.100d 0.113' 0.090 past 4 weeks' (0.027) (0.028) (0.028) PaOity- -0.267 -0.236 (0.120) (0.119) Agricultural hours, last - -0.053 7 days. (0.018) Nonagricul- - 0.054 lural hours, (0.018) ast 7 days- Home labor -- 0.0.23 hours, last (0.018) 7 days- R2 0.137 0.113 0.105 0.127 n 3,130 3,130 3,130 3,130 Endogenous RHS variables. 'Standard errors are located in parentheses below estimates. They, and the R2 stastis,. are cornretd to acoount for the disturbance, in the first-stage regressions (see Kmenta, 684). To asymptotic standard errors of the estimates in the probit regressions are not corrected, however. b Dependent variable is BMI. Sample mean of dependent variable is 21.67. O Pregnant and lactating women were excluded from sample. 'time 10.2 * Maximum likelihood probit estimation. Dependent variable equals one when BMI < 18.5, and zo othawise. 522 cases equal one, and 2,608 equal zero. 36 TABLE Al: OLS Nutrition Regressions l - Ignoring Endogeneity (Women ages 18 years and older)r Variable (1) (2) Intrcet 7.949 13.992 (2.210) (1.965) Age (years) 0.204 0.210 (0.025) (0.026 Age squared 4.002 4.002 (0.0003) (0.0003) Height (cm.) 4.001 0.002 (0.011) (0.011) Edwcad on Prinaiy 0.752 0.799 (0.185) (0.24S) Secondary 0.409 O.S23 (0.185) (0.18M ) Post-seondary 1.077 1.388 (0.432) (0.431) Houshol Size 0.139 0.106 (0.021) (0.020) Femnle-headed 0.364 0.2S7 (0.15S) (O.ISS) Fonr Zone 4.716 -0.817 (0.165) (0.165) Savannah Zone 4.85S5 -1.077 (0.204) (0.202) 2nd Quarter 4.194 4.1S2 (0.18S) (^.187) 3rd Quarter 0.475 0.448 (0.185) (0.186) 4th Quarter 0.270 0.151 (0.192) (0.192) Usban 1.092 1.326 (0.182) (0.180) 37 TLE Al: Contiued Varible () (2) Semi-urban 0.290 0.357 (0.188) (0.189) Log per capita oalories - 0.255 avalable (0.104) Log per capita .0.783 expenditur (0.127) Number of days ill in puat 0.019 -0.015 4 weeks (0.012) (0.012) Log Health care 0.007' 0.007' expenditures, (0.003) (0.003) past 4 wes Prty 0.033 0.025 (0.029) (0.029) hours, lea 4.028 -0.029 7 days (0.005) (0.005) Nonagric- tural hours, 0.015 0.017 lat 7 days (0.004) (0.004) Home abor hour., last 0.011 0.013 7 days (0.006) (0.006) p2 0.159 0.150 n 3,130 3,130 * Standard errors ae in parenthes bdow estimates. i Dependet variable: BMI. Sample -n of depedent vable ui 21.67. Models (1) and (2) ar equivalent to model (4) in Tabla 8 and 9, respectively. * Pregnant and ltaing wom_ w ecoluded from sampl. 'time 10.2 38 TABLE A2: Probit Regresions of Undernutrition Risk .b (Wonen ages 18 years and older) Variable Under Under- Under- Under- numtitiond nuuitiond Nutritiond Nutritiond (1) (2) (3) (4) Intercqpt -0.207 -0.499 0.566 -1.026 (1.130) (1.178) (1.214) (1.317) Age (yeasn) 40.017 40.028 40.016 -0.030 (0.009) (0.019) (0.009) (0.019) AVe squared 0.040, 0.02ff 0.040S 0.041 (0.209) (0.003) (0.009) (0.016) Height (cm.) 0.002 0.003 0.003 0.003 (0.004) (0.004) (0.004) (0.004) Educaion Pimary -0.099 0.007 -.092 0.012 (0.104) (0.111) (0.105) (0.111) Secondary -0.172 4.063 4.156 4.065 (0.810) (0.087) (0.082) (0.087) Post-sesondry 4.131 0.086 -0.095 0.080 (0.202) (0.214) (0.204) (0.214) Hou_ehold Sie -0.036 -0.035 -0.043 4.031 (0.010) (0.011) (0.011) (0.012) Penale-headed -0.065 4.036 -0.087 -0.020 (0.065) (0.074) (0.067) (0.080) Poret Zone 0.192 0.062 0.173 0.065 (0.068) (0.075) (0.069) (0.076) Savanimh Zone 0.355 0.197 0.304 0.202 (0.081) (0.092) (0.085) (0.092) 2nd Quarter 0.037 0.058 0.045 0.058 (0.077) (0.082) (0.077) (0.082) 3rd Quaer -0.093 -0.076 4.097 -0.071 (0.078) (0.081) (0.078) (0.082) 4th Quantr 0.016 -4,051 0.012 -0.044 (0.07) (0.082) (0.079) (0.084) Uban -0.285 0.005 -0.223 0.002 (0.072) (0.091) (0.074) (0-090) 39 1ADEJflJ2: Contnued Variable Undr Under- Under- Under- nutitiond nuition Nutrition' Nutitiod (1) (2) (3) (4) Semi-urbmn 40.131 40.065 4.108 0.066 (0.076) (0.079) (0.075) (0.079) Log per capita calories -.116 4.016 - - available (0.111) (0.122) - - Log per capita - - 4.152 0.039 expendiures - - (0.089) (0.402) Number of days ill in paa - 0.003 - 0.001 4 weeks - (0.018) - (0.018) Health care expenditume, - -0.025' - (-0.026' pant 4 wfs' - (0.007) - (0.008) Pai - 0.048 - 0.054 - (0.039) - (0.040) hours, lag - 0.009 - 0.008 7 days- - (0.004) - (0.004) Nonagricul- tuml houn, - 4.008 - -0.008 lt 7 daysW - (0.004) - (0.004) Home labor houn, la - 0.013 - 4.013 7 days' - (0.005) - (0.005) Log Lakli- 213.4 254.9 215.3 255.0 hood ration n 3,130 3,130 3,130 3,130 * Endogenous RHS variable. * Maxinum ikUelhood probk estians. Asymptoi sandard enr amr bcated in panthe beow esinates. lTmes am not corretod to aount for first-stae prediction eron. * Pnt and actatng women wee eluded fiom sample. Depndent variable equals one if BM1 < 18.S and ze othwine. 523 cea equal unity. * times 10' 40 TABLLh: Ptobk Rersson of Ovevwe Risk (Women aS yar and olderr Variablb Ovor- Over- Over- Over- weighd wgJe Wyighe we (1) (2) (3) (4) tntert -6.283 6.090 -7.325 4.091 (1.058) (1.096) (1.118) (1.317) AV (yam) 0.090 0.105 0.090 0.104 (0.009) (0.019) (0.009) (0.020) AgV squared -0.0w 4.100 40.09P 4.102? (0.011) (0.018) (0.011) (0.018) Hoig (an.) 0.005 0.004 0.004 0.004 (0.004) (0.004) (0.004) (0.004) Ecd~on Prima 0.241 0.213 0.231 0.217 (0.089) (0.095) (0.089) (0.095) Soeonday 4.107 4.012 0.077 -0.016 (0.069) (0.075) (0.070) (0.075) Post-secondasy 0.477 0.301 0.417 0.294 (0.148) (0.160) (0.150) (0.160) Household Size 0.045 0.048 0.056 0.051 (0.009) (0.010) (0.010) (0.011) Famalebeaded 0.094 0.113 0.142 0.139 (0.057) (0.066) (0.060) (0.071) Forndt zone 4.217 4.029 0.178 0.016 (0.059) (0.065) (0.061) (0.066) Savannah Zone 4.421 0.228 4.308 4.165 (0.076) (0.084) (0.079) (0.086) 2nd Quarter 4.121 -0.110 4.141 4.121 (0.070) (0.074) (0.070) (0.074) 3rd Quare 0.087 0.104 0.090 0.108 (0.068) (0.081) (0.068) (0.072) 4th Quatuer 0.086 0.022 4.025 0.058 (0.072) (0.076) (0.073) (0.077 Umban 0.604 0.178 0.452 0.087 (0.06) (0.084) (0.066 (0.082) 41 TL& A: Continued Variable Over- Over- Over. Ovar wO wegi wO wowP (1) (2) (3) (4) Semi-urban 0.285 0.168 0.234 0.136 (0.072) (0.076) (0.071) (0.075) Log per capita caories 0.337 0.283 availabe (0.103) (0.112) - - Log per capita - - 0.333 0.203 oxpenitures- - - (0.080) (0.094) Number of days DU in pas - 0.033 - 0.03S 4 wrgks - (0.016) - (0.017) Health care expenditurs, - 0.013 - 0.013 paat 4 weeb? - (0.006) - (0.006) Party - -0.41 - -0.037 - (0.037) - (0.038) Agriculual hours, la - -0.021 - -0.021 7 daye - (0.004) - (0.004) Nonagricul- tul hours, - -0.012 - 40.012 las 7 days - (0.003) - (0.003) Home labor hours, lut - -0.008 - 0.008 7 days . - (0.005) - (O.005) Log LDbdi- 330.9 416.9 337.5 41S.1 hood radon n 3,130 3,130 3,130 3,130 - Endogenous RHS variabls. * Maxmnum lkelhood probit estimaes. Asymptoti standard errors ae located in parentes below estimat. hsc ae not corretd to account for first-stage predicton erros. o Pregnant and lacting women wern excluded from mple. d Dependent variabe equals one if BM1 > 23.0 and zero othewin. 816 case equal unit. * dmes 102 Policy Research Working Paper Serles Contact Title Author Date for paper WPS992 Regional Integration in Sub-Saharan Faezeh Foroutan October 1992 S. Fallon Africa: Experience and Prospects 37947 WPS993 An Economic Analysis of Capital S. ibi Ajayi October 1992 N. Lopez Flight from Nigeria 34555 WPS994 Textiles and Apparel in NAFTA: Geoffrey Bannister October 1992 A. Daruwala A Case of Constrained Liberalization Patrick Low 33713 WPS995 Recent Experience with Commercial Stijn Claessens October 1992 Rose Vo Bank Debt Reduction Ishac Diwan 33722 Eduardo Fernandez-Arias WPS996 Strategic Management of Population Michael H. Bernhart October 1992 0. Nadora Programs 31091 WPS997 How Financial Liberalization in John R. Harris October 1992 W. Pitayatonakarn Indonesia Affected Firms' Capital Fabio Schiantarelli 37664 Structure and Investment Decisions Miranda G. Siregar WPS998 What Determines Demand for Freight Esra Bennathan October 1992 B. Gregory Transport? Julie Fraser 33744 Louis S. Thompson WPS999 Stopping Three Big Inflations Miguel A. Kiguel October 1992 R. Luz (Argentina, Brazil, and Peru) Nissan Liviatan 34303 WPS1 000 Why Structural Adjustment Has Not Ibrahim A. Elbadawi October 1992 A. Maranon Succeeded in Sub-Saharan Africa Dhaneshwar Ghura 39074 Gilbert Uwujaren WPS1001 Have World Bank-Supported Ibrahim A. Elbadawi October 1992 A. Maranon Adjustment Programs Improved Economic 39074 Performance in Sub-Saharan Africa? WPS1002 World Energy Subsidies and Global Bjorn Larsen October 1992 WDR Office Carbon Emissions AnwarShah 31393 WPS1003 Rent-Sharing in the Multi-Fibre Kala Krishna October 1992 M. T. Sanchez Arrangement: Evidence from U.S.- Ling Hui Tan 33731 Hong Kong Trade in Apparel WPS1004 Family Planning Programs in Sub- Regina McNamara October 1992 0. Nadora Saharan Africa: Case Studies from Therese McGinn 31091 Ghana, Rwanda, and the Sudan Donald Lauro John Ross WPS1 005 An Approach to the Economic Laszlo Lovei October 1992 M. Dhokai Analysis of Water Supply Projects 33970 Policy Research Working Paper Series Contact Title Author Date for paper WPS1 006 Preparing Multiyear Railway Jorge M. Rebelo October 1992 A. Turner Investment Plans: A Market-Oriented 30933 Approach WPS1007 Global Estimates and Projections Rodolfo A. Bulatao October 1992 0. Nadora of Mortality by Cause, 1970-2015 Patience W. Stephens 31091 WPS1 008 Do the Poor Insure? A Synthesis of Harold Alderman October 1992 C. Spooner the Literature on Risk and Christina H. Paxson 32116 Consumption in Developing Countries WPS1009 Labor and Women's Nutrition: Paul A. Higgins October 1992 C. Spooner A Study of Energy Expenditure, Harold Alderman 32116 Fertility, and Nutritional Status in Ghana