Social Services, Water Access, and Agricultural Productivity in Sub-Saharan Africa

Abstract

In sub-Saharan Africa, public resource allocation in social service sectors can be particular important in that low levels of development can make private services inaccessible for much of the population. In addition, short-term social services needs often compete with longer-term growth enhancing investments for limited government funds. However, the analysis and understanding of the agricultural productivity effects from the provision of social services has been severely limited by data constraints and the need to control for a variety of related factors. Given the importance of agriculture for rural incomes, a better understanding of the effects on agricultural productivity from the provision of a variety of social services is important for allocating spending that will not only support short-term needs but also longer-term growth in rural areas. Despite the limited data availability, previous analyses have been conducted regarding the role of public expenditures on rural poverty. Often, this is limited to one sector or overall rural incomes rather than agricultural productivity. In addition, these analyses often assume homogeneous production technology and do not account for climatic variations driving production. Expanding on the work that has been done previously, we analyze multiple expenditure sectors, allowing heterogeneity in the underlying socio-economic and agro-climatic environment. This research estimates the impact of public expenditures on agricultural productivity across countries of sub-Saharan Africa and then, specifically in the case of Tanzania. Furthermore, given the limited funding available for the water sector in Tanzania, it looks specifically at water constraints at the household level and how this may constrain agricultural productivity. In an effort to estimate the impacts of public expenditures on agricultural productivity, the first part of the analysis explores the existing data for sub-Saharan Africa, using a cross-country regression framework. It exploits multiple analytical options while including a newly-compiled dataset on annual precipitation for agricultural land. While our conclusions are substantially limited by the data constraints, this analysis conducts an efficiency analysis for the group of sub-Saharan African countries using health and educational outcomes as indicators of social service availability. It also implements a latent variable structural equation framework for a subset of countries, allowing not only for country-specific heterogeneity but also more direct estimation of the role of social service expenditures covering multiple sectors. Overall, the results provide evidence that public service expenditures (especially on health and education) can influence input productivity and efficiency in agriculture. Country-specific heterogeneity and climate-related variables appear to be a significant consideration for agricultural production in sub-Saharan Africa that should not be ignored. In the remaining sections, we focus on Tanzania in particular. We begin with an investigation of the impacts of district-level health and education expenditures on marginal productivities of agricultural inputs and overall production. The results of our latent variable, covariance structural model confirm the significance of government social expenditures in human capital formation as measured through health and education indicators. The results also show the effects of these health and educational outcomes on agricultural productivity. We find that the marginal productivities of inputs (labor in particular) respond significantly and positively to health and education outcomes, especially when health and education are considered jointly. The impacts also seem to be a function of the type of health constraint, with short-term health factors such as malaria and diarrhea impacting productivity of seeds and fertilizer. In contrast, longer-term health problems (i.e., chronic diseases) appear to have greater impacts on labor quality and land productivity. The results also suggest that there is a minimum level of expenditures per capita needed to see an impact on educational or health outcomes. The results confirm the importance of considering intra-country heterogeneity as well as climate-related constraints, as the results demonstrate that annual precipitation has a significant impact on production for all specifications. For the final part of the research, we focus on the water sector in Tanzania, given that public spending in this sector is very limited. For this analysis, we rely upon primary household survey data collected in four regions of Tanzania to investigate the impact of drinking water access on agricultural productivity at the household level. As in many parts of sub-Saharan Africa, a lack of sufficient and safe drinking water is a health constraint in rural Tanzania. Although the water sector has been prioritized in poverty-reduction efforts, water collection time has been overlooked in many studies addressing productivity and incomes. Using a production function approach, we estimate the impact of drinking water collection times on agricultural labor availability, labor productivity, and yields. The results show that additional time required to access drinking water significantly reduces household labor productivity and yields for households that have to spend more than the median time (20 minutes) collecting water. In these households, the time needed to collect water appears to serve as a more significant constraint to labor productivity than the quality of the water at the source. This effect remains consistent even after controlling for heterogeneity among households, crop diversity, and districts. The relative importance of other productive inputs also changes with increased time needed for water collection. Our results support further consideration of the time required for water collection when planning rural development initiatives. Overall, this research has highlighted the important data limitations and the confounding constraints that can hinder analysis of the impacts of public expenditures on agricultural productivity. In order for governments to efficiently target limited funds, a more comprehensive understanding of the direct and indirect impacts of these expenditures on farm-level decisions and constraints will be necessary. Further data sharing among agencies and working across ministries is needed in order to o meet common goals for long-term growth while addressing more immediate social needs. The findings here suggest that social expenditures can also positively impact agricultural productivity and these impacts should be further explored in making funding decisions. In addition to emphasizing the need for additional data, this research has other important policy implications. It points to the importance of providing sufficient levels of expenditures in order to influence health and education outcomes and the importance of recognizing context-specific needs given the heterogeneity between countries even within a country. In addition, this research has provided strong evidence of the importance of water-related constraints (both precipitation and household access to water) that seem to support not only consideration of these constraints but also funding of measures that may reduce vulnerability.