Search

The economic impacts of COVID-19 lockdowns on poor and vulnerable households living in rural areas of developing countries are not well understood due to a lack of detailed micro-survey data at the household level Utilizing weekly financial transaction data collected from households residing in a rural region of India, we estimate the impacts of India's COVID-19 lockdown on household income, food security, welfare, and access to local loan markets A large portion of households living in our study region is reliant on remittances from migrants to sustain their livelihoods Our analysis reveals that in the month immediately after India's lockdown announcement, weekly household local income fell by INR 1,022 (US$ 13 5), an 88% drop compared to the long-term average with another 63% reduction in remittance In response to the massive loss in earnings, households substantially reduced meal portions and consumed fewer food items Impacts were heterogeneous;households in lower income quantiles lost a higher percentage of their income and expenditures, but government food aid slightly mitigated the negative impacts We also find an increase in the effective interest rate of local borrowing in cash and a higher demand for in-kind loans, which are likely to have an adverse effect on households who rely on such services The results from this paper have immediate relevance to policymakers considering additional lockdowns as the COVID-19 pandemic resurges around the globe and to governments thinking about responses to future pandemics that may occur © 2020 Agricultural and Applied Economics Association

We estimate the effects of changes in cotton adoption on children’s schooling and child labor in rural Burkina Faso. Using time and spatial variations, we find evidence that expansion of cotton farming has led to an increase in enrollment and to a reduction of participation in child labor for girls. There are, however, no detectable effects on boys. In theory, cotton adoption could increase household income, leading to increased demand for schooling and reduced child labor. On the other hand, because children are productive on cotton farms, adoption of cotton could increase the opportunity cost of child time and the demand for child labor. We provide suggestive evidence showing that boys are more productive than girls on cotton farms. Taken together, the results suggest that the income effect from cotton adoption might have been larger than the wage effect for girls, hence the overall positive impacts on school enrollment for girls.

This comment discusses the robustness of the policy implications of Bellemare, Barrett, and Just's paper, “The Welfare Impacts of Commodity Price Volatility: Evidence from Rural Ethiopia” (2013). Bellemare, Barrett, and Just present a theoretical and empirical approach to the estimation of willingness to pay for food price stabilization that accounts for the covolatility of prices, making a significant contribution to the literature. However, in the course of applying their model to data from rural Ethiopia, the authors make an empirical assumption in the treatment of zero-valued income households that produces a distortion in the distribution of household budget shares. This comment identifies the consequences of this assumption for the estimated relationship of poor and wealthy households' willingness to pay for food commodity price stabilization, and shows the results one would obtain under a different, distribution-preserving treatment of zero-valued income. The key finding is that the distributional benefit incidence of food price stabilization found in Bellemare, Barrett, and Just (2013) is reversed when the budget share of marketable surplus is calculated over observed, as opposed to mean, household income where available.

In many developing countries, supermarkets are expanding rapidly. This affects farmers’ marketing options. Previous studies have analyzed welfare effects of smallholder participation in supermarket channels from a static perspective, using cross‐section data. We develop a conceptual framework and use panel data to better understand participation and impact dynamics. The analysis focuses on vegetable producers in Kenya. Participation in supermarket channels is associated with income gains. However, many farmers have dropped out of the supermarket channel due to various constraints. The initial income gains cannot be sustained when returning to the traditional market. Organizational support may be needed to avoid widening income disparities.

The yield potential of a set of improved rice management practices, known as the System of Rice Intensification (SRI), has attracted much attention. Yet we know surprisingly little about SRI's socio‐economic impact. Using data from Indonesia in 2009, this study assesses the impact of SRI on household incomes and child schooling. We find that SRI generates significant estimated yield gains. However, because SRI induces a reallocation of family labor from non‐farm to farm, SRI users enjoy no household income gains. Despite the increased labor demand for farming, we find no evidence that SRI has a child labor effect.

Using the 2007 National Survey of Children's Health data, we find a statistically and economically significant effect of neighborhood parks and playgrounds on childhood obesity based on covariate matching estimators. The park/playground effect depends on gender, age, race, household income, neighborhood safety, and other neighborhood amenities. The results suggest that adding a neighborhood park/playground may reduce the obesity rate and make children more fit, but relevant interventions must consider socioeconomic status of the targeted children as well as other neighborhood amenities.

Item non response is a pervasive problem in survey data, and a large literature in statistics has developed to address missing data due to item non response. Government statistical agencies regularly impute for missing values in the microdata that they make available to researchers. Unfortunately, while the imputation methods typically used by government statistical agencies may be suitable for the main objectives of the agency—usually, publishing aggregated descriptive statistics—these imputations are usually unsuitable for multivariate analysis of microdata. Using USDA’s Agricultural Resource Management Survey (ARMS), we apply a new imputation method to missing 2008ARMS data on federal government farm commodity payments and off-farm household income items. We find that the choice of imputation method can have a noticeable effect on important economic indicators, even when the rate of missingness is small.

Thisarticleexploresintrahouseholddietarydiversityallocationwithinpastoralisthouseholdsineastern Africa.We estimate income elasticities of dietary diversity for demographic cohorts allowing asymmetric behavior depending on household circumstances.We find that household heads disproportionately bear the nutritional burden when household income is below mean,while other cohorts disproportionately enjoy the nutritional gains when it is above mean.The commonly assumed symmetric framework masks this intrahousehold behavior and produces biased elasticity estimates. Stochastic dominance tests show adult daughters as better-off than other household members in their dietary diversity, sons as worse off, and little difference between male heads and their wives. Nutritional welfare is an individual, not household, characteristic. While the average nutrient intake and dietary quality of the household can be one of the largest determinants of the well-being of its members, nutritional well-being is individually experienced. Therefore, development policies targeting nutrition are typically more concerned with their possible implications for individual welfare than for household-, community-, or national-level aggregates. The household, though, may be one of the most important institutions influencing individual nutritional welfare. Consequently, in order to evaluate income generation and pricing policies, it is essential to understand both how households allocate resources to individual members as well as how allocations

Farm households are economic agents whose income is derived from farm, off-farm, and government sources. This article uses farm-level data from the Agricultural Resource Management Survey (ARMS) and recent advances in the econometric theory of dynamic pseudo-panels to show that farm households consume various sources of income differently at the margin. Particular attention is given to a specific type of lump-sum government transfer payment intended to be decoupled from (independent of) farm production decisions. The results suggest that relatively decoupled government subsidies have a greater marginal effect on farm household consumption than subsidies that are tied to market conditions. Farm households often receive income from off-farm employment and government support payments in addition to farm production revenues. If income is fungible, then marginal consumption should not vary by income source. However, Carriker et al. (1993) show that farm household consumption does vary at the margin by income source. They find that off-farm income and income from government subsidies are consumed at a higher marginal rate than income from farm production revenues. If farm household consumption varies at the margin by income source, different types of government payments may affect consumption differently. Closely following Carriker et al. (1993), this analysis further disaggregates sources of farm household income and compares how different types of government payments affect farm household consumption. It is shown that government payments that are relatively “decoupled” from farm production decisions are consumed at a much higher marginal rate than government payments that

The inverse relationship between land productivity and farm size is an old and puzzling empirical regularity. Most explanations for this relationship rely on market imperfections that jointly determine the farm size and the household's shadow price of some productive inputs. We use plot-level data from the ICRISAT/VLS to assess whether these household-specific theories can explain the puzzle. The data exhibit plots of different sizes being simultaneously cropped by the same household. The inverse relationship is shown to hold true with the same magnitude across the plots of each household, thus cross-household heterogeneity does not suffice to explain the puzzle. Copyright 2007, Oxford University Press.

Farm families with incomes below the poverty line are far less likely than wealthier farmers to receive farm support payments. Using data from the 1989‐2004 Current Population Survey, we find that poor farm families are also not participating in other assistance programs. Controlling for other factors, eligible farm families have substantially lower participation rates in the Food Stamp Program and in Medicaid than eligible nonfarm families. Removing farm safety net program payments would increase the number of farmers eligible for these programs but, in the absence of behavioral changes, would only lead to small increases in the number of recipients. The genesis of a strong federal presence in agricultural commodity markets was the persistence and pervasiveness of rural poverty. In 1930, 44% of Americans lived in rural areas, and 30 million of these 54 million people lived on farms (U.S. Department of Commerce 1933). Farm household income was less than half that for nonfarm households (U.S. Department of Commerce 1975). The high percentage of Americans in farming and their low average incomes presented an important challenge to achieving national prosperity. In response, as part of the New Deal, the federal government established a series of farm safety net

The extent to which farm families engage inoff-farm employment activities is a topic ofcentralimportancetounderstandingthefinan-cial well-being of farm households. The degreeof financial stress experienced by U.S. farmfamilies is always a concern to policy makers.Yet,policyconsiderationsoftenignorethesub-stantial involvement of most farm householdsin nonfarm labor markets. It has been shownthat,whentheentirefarmhouseholdisconsid-ered, the “average” U.S. farm household hasa higher total income and much greater assetholdings than is the case for nonfarm house-holds (see, e.g., Goodwin). Mishra et al. pre-sented data for 2000 suggesting that, when allfarms are considered, 92% of total householdincomecamefromnonfarmsources.Ofcourse,such figures depend on how one chooses to de-fineafarm.Whenoneconsiderslargeandverylarge commercial farms, the share of farm in-come in total household income ranges from50 to 75%.

The adoption of more efficient farming practices and technologies that enhance agricultural productivity and improve environmental sustainability is instrumental for achieving economic growth, food security, and poverty alleviation in sub-Saharan Africa. Our research examines the interaction between public investments, community health, and adoption of productivity and land enhancing technologies by households in the northern Ethiopian state of Tigray. Agricultural technology adoption decisions are modeled as a sequential process where the timing of choices can matter. We find that time spent sick and opportunity costs of caring for sick family members are significant factors in adoption. Sickness, through its impact on household income and labor allocation decisions for healthcare and other activities, significantly reduces the likelihood of technology adoption. Our findings suggest that agencies working to improve agricultural productivity and land resource conservation should consider not only the financial status of potential adopters, but also their related health situation. Copyright 2004, Oxford University Press.

Introduction This brief article focuses on how supermarkets in Latin America are imposing standards of quality and safety on producers of fresh fruits and vegetables (FFV) mainly for the domestic market. It is a synthesis of recent case studies from a range of Latin American countries. The range includes Brazil, Costa Rica, Guatemala, Honduras, and Nicaragua – the range being in decreasing order of household income, share of supermarkets in overall food retail, and from strongest to weakest domestic public health standards. The research is based on fieldwork in Brazil in 2002 and in Central America in March/April 2003, including rapid reconnaissance surveys of supermarket chains, wholesalers, and suppliers. The questions focused on procurement practices and application of standards, including private enforcement of public standards, and application of private standards. The article begins with a short section analyzing the “trade bias” in the literature on FFV standards and argues for the need for agricultural development and trade economists to pay close attention to the role of supermarkets in setting quality and safety standards for domestic suppliers. This role is both a neglected area of research and an important one because we show supermarkets in Latin America buy far more from local producers than is exported from Latin America to the rest of the world. The next section focuses on changing

Data from the third National Health and Nutrition Examination Survey are used to analyze the effect of the Women, Infants, and Children (WIC) Program and other factors on the health of U.S. preschool children. Ordered probit equations are estimated for the physician's overall evaluation of the child's health. The WIC Program has a significant positive impact on the overall health of children. In particular, children in households participating in WIC are significantly more likely to be in excellent health. Increased household income also improves their health. Copyright 2003, Oxford University Press.

This article deals with the problem of item nonresponse in contingent valuation surveys using a payment-card method, by applying a grouped-data sample-selection estimation technique that is capable of imputing the missing values conditional upon a respondent's decision to answer a willingness-to-pay question. The advantage of the technique lies in its ability to utilize all of the information in the sample, permitting a more efficient estimation in the presence of item nonresponse bias. The major determinants of willingness to pay appear to be household income, number of children, education, perception of existing water quality, and identification with environmental issues. Copyright 2003, Oxford University Press.

The common thread in the three papers (Jones; Adelaja, Nayga, and Wall; McDowell, AllenSmith, and McLean-Meyinsse, all in this issue) is that they stratify their respective data sets along household income lines and examine how either the quantity of food or nutrients or food expenditures vary by income. In two of the papers, income is matched with the household where the food or nutrient consumption occurs. These studies look at food expenditure and quantity of nutrients consumed by various income groups. In the third paper, income is measured by the median family income in a 5-mile radius around a store and by total expenditure on cereals being purchased in the same store. This study measures differences in price and expenditure elasticity of demand among consumers in highand low-income neighborhoods.

The literature on risk management by agricultural households has made important contributions to our understanding of household-level economic decisions in developing countries. Because incomes from particular farming enterprises fluctuate widely depending on weather, crop prices, and macroeconomic policies, households adopt a number of strategies to smooth their consumption. One approach is to diversify the sources of income for the household. The members of peasant households may engage in farming, trading and other business activities, and wage labor (Reardon, Delgado, and Matlon). When household members are involved in multiple activities and pool their income, total household income may be less subject to extreme variation. Morduch notes "mitigating risk through production choices can be costly since typically expected profits must be sacrificed for lower risk" (p. 104). This approach to risk management may be costly for the household. Over the past three decades, economists have used theories of household risk management and consumption smoothing to explain behavior observed in peasant households. For example, the purchase and sale of assets are frequently used as a means of savings and consumption smoothing when there are limited formal financial institutions. Marriage decisions may establish kin relationships with people in areas that experience different weather and production shocks (Rosenzweig and Stark). Although it may seem inefficient for households to farm numerous small plots that are widely scattered, this strategy provides some protection against location-specific crop failures from bad weather or pests (McClosky, Walker and Ryan). The extensive literature on contracting and sharecropping arrangements focuses on the role that these institutions play in minimizing a household's exposure to risk (Hoff, Braverman, and Stiglitz). All of these theories assume that households

Nutrition information and intake data for a sample of U.S. household meal planners are used to estimate the effects of fiber-specific information on dietary fiber intake. The information variables are measured using survey questions on fiber content of foods, attitude toward consuming fiber-rich foods, and awareness of fiber-health links. The evidence confirms the influence of nutrition information on fiber intake. Household income, meal planner age, smoking status, vegetarian status, race, and ethnicity are major determinants of both information and intake. Education exerts a sizable intake effect by enhancing the information level. Direct and indirect effects of exogenous variables are reported. Copyright 1996, Oxford University Press.

be targeted for credit and small enterprise programs not only because their income boosts household income, but also because it meets global societal objectives such as increased spending on food and children's goods. Some researchers, however, have speculated that the observed impact of female income share on household expenditure patterns does not reflect gender-differentiated preferences, but rather reflects the differential "lumpiness" (seasonality) of male and female income flows. That is, women may earn income in flows that ensure a higher propensity to purchase certain items.

Farm household surveys in the 1970s and 1980s showed that nonfarm activities provide an important share of household income, contrasting with the conventional image of rural African households as deriving their food entitlement almost exclusively from the land (e.g., Matlon; Collier and Lal; Low; Reardon et al.; Haggblade, Hazell, and Brown). Much of the literature on nonfarm income has concentrated

become increasingly recognized. Earlier empirical studies have indicated that they provide 20% to 45% of full-time employment and 30% to 50% of rural household income (Chuta and Liedholm; Haggblade, Hazell, and Brown). This is also reflected in the agricultural growth linkage literature (e.g., Haggblade and Hazell), where rural enterprise growth is typically a demand-driven spinoff of agricultural growth, as well as in the rural household studies that have

A rural-urban interregional computable general equilibrium (CGE) model is constructed to simulate the effects of terminating farm subsidies on household incomes, employment rates, farm and non-farm sectoral activity, regional costs of living, and other economic indicators. The magnitudes of the effects depend on regional factor and goods market segmentation. Robust short-run implications are that ceasing farm subsidies would cause rural nonfarm (particularly household service sector) employment to fall and lead to lower household income. On the other hand, rural manufacturing activity expands and the cost of living falls relative to urban. Urban employment, household income, and land rents rise. Although termination of farm subsidies causes a decline in rural real product, the urban real product gain outweighs rural losses.

has been more pronounced in the noncommercial and small commercial farm classes during this period. The decline is drastic even after considering the changes in the definition of a farm used by the Bureau of the Census and the Department of Agriculture. Large farms have shown an almost steady increase in numbers during this period (Carr). Even with the drastic decline in the number of small farms, they still account for a large proportion of the nation's total farms. According to the 1982 Census of Agriculture, about 76% of the nation's farms are small. The South, one of the regions with the largest number of small farms, has in excess of one million farms, and over 80% of these farms are defined as small. Georgia has approximately 49,627 farms; at least 67% of its total are small farms (Brown). Limited resource farmer (LRF) is a term used to describe the traditional small, part-time farmer throughout the United States. The term limited resource farmer apparently took its origin from the financial assistance program developed by the Farm Home Administration and approved by Congress in 1978 via the Agricultural Credit Act. However, the concept of limited resource farmer pertains to the lack of adequate inputs for the normal operation of the farm business. Additionally, the term limited resource farmer is sometimes used interchangeably with low-equity fa m, low-income farmer, small-scale farme , marginal farm, and small farm (Carlin and Crecink, Chidebelu, Tweeten). Previous studies by Saupe, Chidebelu, Tweeten (1982), Faust, Beauford, Williams, and Yeboah and Strickland address production, financing, and structure of agricultural issues applicable to the small farm issue. Recently, more concern has been directed to the survival issues of the limit d resource farmer, especially as it relates to structural changes that are occurring. Beauford alluded to the fact that a major difference exists between the income-generating strategies, patterns of household income, racial make-up, and geographic location of small farms compared to that of larger farmers. She concluded that, while differences exist between farm-income-dependent and nonfarm-income-dependent households, their economic well-being remains relatively the same. Limited resource farmers have been, and continue to be, important in terms of their numbers and the resources that they control. There are indications that beliefs about small farms could

The marginal propensity to consume (MPC) for a sample of eighteen Illinois farms over the 1979–86 period is determined. Four consumption models were estimated using disposable household income plus depreciation as the measure of income. Estimated short-run MPCs ranged from 0.007 to 0.020, while long-run MPCs varied between 0.143 to 0.381. These results indicate farm family consumption responded little to changes in income and that the life cycle hypothesis model explains consumption significantly better than the other models. Robustness of the results is demonstrated using a larger sample of farms for 1986&87.

The depressed thirties brought forth side by side the paradox of food surpluses and obvious need. Federally sponsored food programs simultaneously were aimed at raising farm income and improving diets (Gold, Hoffman, and Waugh, p. 1; Southworth and Klayman, p. 1). These pre-World War II objectives are embodied in current legislation, although emphasis has shifted from farm income support toward improved diets. From their beginning, economists have recognized that an injection of funds by means of food assistance programs yields a different pattern of resource allocation than an equal injection by other means. For example, an injection to provide food assistance results in a different allocation of resources than an equal injection to support national defense. In 1961 federal assistance to the food stamp, food distribution, and school lunch programs amounted to $466.5 million.' For 1975 the preliminary estimate is $6,139.1 million (USDA 1976). The 1975 figure is a 13.2 multiple of the 1961 statistic. Programs of such magnitude deserve quantitative as well as philosophic analysis. The focus of this paper is on the quantitative. Its purpose is to measure the economic effects of these programs in terms of net changes in business receipts, household income, jobs, and gross national product (GNP). The analysis is based upon the assumption that the funding for food bonus stamps and the cash contribution to the schools is paid for by a tax increase.2 The taxpayers' disposable income and, hence, expenditures were reduced prior to making the income transfers to the recipients. The results of the analysis of food stamps, assuming a transfer from taxpayers to food stamp households in the form of bonus stamps, are reported in net terms because both the output increases associated with the income transfer to the participant household s ctor and the decreases in output linked to the increased personal income taxation of the nonparticipant household sector have been taken into account. The results of the analysis of school lunches, when the transfer is assumed to be in the form of a cash contribution, are also in net terms; however, the increases in output are linked to the federal cash transfer to the schools and not to the participant household sector. Business receipt changes are reported on a sector-by-sector basis. As anticipated, some sectors were better off with than without the programs. Conversely, others would have The reactors for this session were Loren Geistfeld of Purdue University, Stephen J. Hiemstra of the Food and Nutrition Service, U.S. Department of Agriculture, and Wendell Primas of Georgetown University. Paul E. Nelson, Jr., is an agricultural economist with the Economic Research Service, U.S. Department of Agric lture, and John Perrin was with the Division of Coordination Planning, Office of the Governor, Austin, Texas. The following colleagues made constructive suggestions at various stages of manuscript development: Tom Carlin, Clark Edwards, Gar Forsht, Alden C. Manchester, Masao Matsumoto, Bob Reese, Gerald Schluter, and Abner Womack. Herbert Grubb of the Water Development Board, Texas, and A. C. Hoffman in Bull Shoals, Alabama, were also very helpful. These figures exclude special supplements to women, infants, and children (WIC), institutional distribution, and school milk activities. 2 Alternative assumptions that could have been adopted include (a) all federal contributions were funded by deficit financing or (b) the 1967 data already included the needed funds.

Economic analysis of income transfer programs requires knowledge of how households with different incomes alter expenditure patterns as their incomes change. Classical demand theory, however, provides little guidance for quantifying these income response parameters. Consequently, researchers have relied on statistical approaches to find functional forms which "best" describe the relationship between expenditures and income. The most commonly used specifications (linear, semilogarithmic, and double-logarithmic) imply that the Engel (expenditure) elasticities are monotonically increasing, decreasing, or constant across the income spectrum. However, serious errors may occur if actual elasticities do not adhere to one of these patterns. This paper utilizes a nontraditional approach, based on income and expenditure distributions, to estimate expenditure elasticities. The method places less emphasis on the mathematical specification of the Engel relation and allows for increased flexibility in elasticity patterns. The primary purposes of this article are to examine household income and expenditure distributions in order to deduce the underlying Engel relationship and to measure inequality within each distribution. When other relevant factors (e.g., prices) are held constant, the expenditure distribution on a commodity is directly related to the income distribution via the commodity's Engel function. A recently proposed technique for the estimation of Lorenz and concentration curves is employed (Kakwani and Podder). These curves are directly related to the cumulative distributions of income and expenditures. Expenditure elasticities and measures of income and expenditure inequality can be calculated from the estimated parameters of the Lorenz and concentration curves (Kakwani). Elasticities thus estimated are not constrained to be monotonic or constant across the income spectrum, and may vary across commodities. Related objectives are to (a) introduce a generalized functional form to describe the Lorenz and concentration curves, (b) compare estimated expenditure elasticities among three urbanizations for several food groups using U.S. data, (c) examine elasticity patterns within each urbanization, and (d) examine inequality of food expenditure and income distributions for each urbanization.

Farm wives allocate their time outside of the household to two work activities, farm and off-farm work. Although few studies have attempted to determine the value of the time of farm wives devoted to farm production, the value of their time is important for understanding the level and fluctuation of farm household income and the changing time allocation of farm wives. The study on which this paper is based focuses on assessing the value of time spent by U.S. farm wives at farm work. The data on wives' farm work were derived from the 1964 Census ofAgriculture (U.S. Dep. Commerce 1967, 1968). These data have deficiencies, but until surveys are revised to better measure the working time of farm wives, these and similar data are the best available. The

Household expenditures for at-home consumption on three species of shellfish and five species of finfish were analyzed. Factors explaining the variation of expenditures on seafood commodities were own price, household income, coupon value, household size, geographic region, urbanization, race, and seasonality. Own-price elasticities ranged from −0.45 (flounder/sole) to −1.13 (oysters). Expenditures on fishery products were more sensitive to changes in household size than to changes in household income. Cross-price effects of red meat and poultry were not statistically significant. For comparison purposes, estimates of own-price, income, and household size elasticities from the literature were made with this set of elasticities.

this paper is to identify selected studies which have contributed to the present knowledge on socioeconomic characteristics of households and dietary status. Studies of the socioeconomic factors influencing dietary status have undergone significant change in recent years because of developments in statistics and household decision theory. In the latter case, households utilize various inputs-for example, food as purchased and the labor of household members-to produce diets with varying features including nutrient compositions. As indicated by the work of Basiotis et al. (1983); Blanciforti, Green, and Lane; Davis et al. (1982b); Sanderson; and Sexauer; this framework provides a systematic rationale for the incorporation of socioeconomic variables into integrated statistical analyses of food consumption and dietary status. More specifically, the scope of traditional consumer demand theory has been significantly increased by household economic decision theory. In addition, permitting explicit incorporation of socioeconomic variables in Engel and full demand systems, this theory makes a plausible linkage between food intake and dietary status. Since household economics merely extends, rather than refutes, the intrinsic utility maximization properties of traditional demand theory, the Engel and demand systems relationships postulated linking socioeconomic characteristics, food expenditure, and dietary status are admitted by the theory (Stigler and B cker). As will be demonstrated, research results are accumulating on household size, household income and expenditure, nutrient levels in household food supplies and individual diets, efficiency of food use, away-from-home food consumption, and use of convenience foods. However, studies of the relationship between socioeconomic characteristics of households and/or individuals and nutrient availability and/or dietary intake levels have produced varyingly conclusive results. Among other factors, these varying results can be attributed to (a) different measures of diet quality and (b) differences in individual food intake versus food disappearance data. Further, diet quality has been assessed directly (units of nutrients available or consumed) as well as indirectly (expenditure levels on food). The dietary components studied most frequently are food energy, protein, calcium, iron, vitamins A and C, riboflavin, thiamin, and niacin. However, recently studies have given less attention to thiamin, riboflavin, niacin, and vitamin C and instead have investigated vitamin B6 and magnesium. In the near future, this list of target dietary components will probably include cholesterol, sodium, dietary fiber, zinc, and copper. This extension will occur for two reasons: (a) consumer interest in these dietary components and (b) availability of survey data on these dietary components, e.g., the U.S. Department of Agriculture's (USDA) 1985 Continuing Survey of Food Intake by Individuals and the 1987-88 Nationwide Food Consumption Sur

This paper presents estimates of net income gains to a sample of households in rural Mexico from sending migrants illegally to the United States, correcting for sample selectivity bias, and it examines the role of expected income gains in driving illegal Mexico—U.S. migration. There is no evidence that people who migrate illegally to the United States are above-average contributors to household income, either as workers in Mexico or as Mexico—U.S. migrants. However, other things being equal, villagers who are in the best position to contribute to household income as workers in Mexico are positively selected not to migrate to the United States.

An approach to measuring the magnitude and incidence of regional downstream effects of an irrigation project is described. The approach is applied to the Muda River project and surrounding region in Northwestern Malaysia. In aggregate, the Muda project's downstream effects are the same as its direct effects. That is, for every additional dollar of value added in paddy production generated by the project at maturity, approximately 75 cents of value added is generated by downstream effects. Also, each dollar of downstream value added is probably supported by just over a dollar of additional investment in plant and equipment spread over the sectors where project-related expansion occurred. The direct effects of the project did not worsen the distribution of income among farm households, but its downstream added value accrued mainly to the nonfarm households engaged in paddy milling and the production of nontradables. Although the spread in nonfarm incomes was wide, the majority of downstream income went to households that were better off than those engaged in paddy farming. Thus, while the project's downstream effects did much to boost the aggregate income of the Muda area, they worsened the intraregional distribution of income. 11 references.

Consistent with prior expectations based on household production theory, household income, time value, size and composition, and the environment in which production and consumption occurred were all important determinants of total household expenditures on food-away-from-home. However, the importance of these factors varied by type of food facility: conventional restaurants, fast-food facilities, and other commercial establishments. Decomposition of the tobit elasticities indicated the differential importance of market participation effects of household size, income, and time value by level of the variable and by type of food facility.

IN view of the multidimensional nature of the social welfare function, concern for the economically efficient provision of public goods and services should be coupled with equal concern for the equity effects of that provision [4]. More specifically, the personal income distribution effects of a project or program, as well as other equity considerations, must be analyzed if the planning process is to provide the information flows necessary for making public decisions consistent with the general welfare. Quantification of such effects make the trade-offs that exist between governmental objectives more explicit and point up the equity impact of programs undertaken wholly or partially for other reasons. The Study This paper reports on portions of an empirical study carried out with the use of fiscal 1968 data on New York State administered outdoor recreation expenditures [1]. Several different conceptual approaches were used and compared. The first utilizes a classical flow of funds form of analysis, tracing the redistribution effects of all state revenue devoted to outdoor recreation in a given year. In this approach, recreation expenditures were treated as gross transfer payments and, for each income class, were netted against tax and fee burdens imposed. The second approach analyzes the present value equity impact of incremental state investment in recreation facilities. Projected annual benefits, measured by a proxy for willingness to pay and converted to a present value basis, were used to estimate gross equity impacts and, for each income class, were netted against discounted costs. Empirical considerations dictated that the income base used be household income before taxes, as defined for the National Recreation Surveys (NRS) [2, 3], and that the class definitions be: under $3,000; $3,0005,999; $6,000-9,999; $10,000-14,999; $15,000 and over. The distribution of households within these classes was taken from the Current Population Survey. Data from the NRS were used to estimate the proportion of user days consumed by each income class for any given activity. These proportions, when multiplied by the total dollars spent by the state on that activity or LEONARD A. SHABMAN is a graduate research assistant and ROBERT J. KALTER is assistant professor of resource economics at Cornell University. 1516

We welcome the comment by Crawford and Lassiter on our earlier paper, which was based on fieldwork completed in 1977. As they point out, the labor requirement coefficients for maintaining a pair of oxen under eastern Burkina Faso conditions were guesstimates based on experience with range herds, subject to all the well-known problems of attributing herd size overhead on a per capita basis.' This was necessary because farmers did not use cattle for draft cultivation in Delgado's fieldwork zone, despite three decades of extension to the contrary. Since the time our article was written, the commentators have been major contributors to a growing literature on the impact of animal traction based on empirical work somewhat to the east of Delgado's research zone (Lassiter, Barrett et al.). They therefore have a potentially major contribution to make to our understanding of a complex, and very important, puzzle in Sahelian agricultural development. Generally, we accept two of Crawford and Lassiter's points. First, it appears on the basis of their empirical results that maintenance times for a pair of oxen are lower than we estimated originally. Second, other constraints, such as high overhead investment, long learning periods, liquidity constraints, risk, and failure of institutional support systems, are undoubtedly important issues in explaining low adoption rates. Of course the second point in no way contradicts our earlier assertion that lack of profitability within a reasonable time span is a sufficient, if not necessary, explanation of nonadoption. However, we have considerable difficulty with the gist of Crawford and Lassiter's other points and with their interpretation of the significance of their results generally. In fact, we still feel that the main message of our earlier article is correct. This was that the opportunity cost of farm resources tied up in oxen cultivation technology is prohibitive under the conditions of eastern Burkina Faso in the late 1970s. Thus, the main policy issues are the adequacy of technologies proposed and incentives, not the suitability of support institutions such as extension, as implied by the comment. First, lack of financial profitability over a fourto six-year period, as stated by Crawford and Lassiter, surely is a sufficient condition for nonadoption of a technology by Sahelian smallholders. Second, Crawford and Lassiter's comments, and the work on which they are based, only assess the financial returns to oxen cultivation in a multiyear perspective (Lassiter, Barrett et al.). Their contention is that, given enough time, the enterprise will become financially profitable. Presumably, this is offered as a reason for forebearance by farmers and subsidies by the state. However, the real issue, the opportunity cost of these investments, is not mentioned. It is regrettable that Crawford and Lassiter do not explore the financial opportunity cost at the household level of using oxen for cultivation. Our article hypothesized that this financial opportunity cost was incurred through diversion of labor away from cropping to team maintenance. Data in Lassiter and Barrett et al. suggest that this is not a problem. However, the true opportunity cost problem may be, in fact, only slightly different. Their detailed data on sources of farm household income show that for oxen cultivation, mean nonfarm income was nearly three times as high for hoe farm households as for oxen cultivation households (Barrett et al., p. 75). Thus, we believe that Crawford and Lassiter's implication that "ATO adopters in Upper Volta ... [had] proportionally more nonfarm cash income than nonadopters," because of the existence of relatively larger households among the adopters, is misleading. Furthermore, given that oxen cultivation households were 67% larger on average than hoe cultivation households (Barrett et al., p. 74), it appears that nonfarm income per person in the hoe households was nearly five times as high as in oxen cultivation households. Using Barrett et al.'s data for the twenty-four hoe farms and the forty-four oxen farms that are truly comparable, coming from the same locations, it can be shown that total household income per active worker for the hoe farmers was 12% above that for oxen workers (Delgado 1984). Furthermore, the source of this difference lay not in the direct financial costs of oxen maintenance but in the relatively much higher income from nonagricultural trading and artisanal activities by hoe farmers. Since these activities require both capital and dry season labor, as does maintenance of a plow team and equipment, this is strongly supportive of the prohibitive financial opportunity cost of resources hypothesis, although small sample size must be kept in mind. Christopher L. Delgado is Coordinator for African Research at the International Food Policy Research Institute, Washington DC. John McIntire is an economist at the International Livestock Center for Africa, Addis Ababa, Ethiopia. Views expressed are strictly personal. 1 The Republic of Upper Volta was renamed "Burkina Faso" on 4 Aug. 1984.