Summary

Rural income growth from increased crop production can have multiplicative effects on a region when that income is respent on local goods and services that would not otherwise have had a market outlet. These spin-off effects on local activities from the spending of increased farm incomes are called "agricultural growth linkages," and they were shown to be an important element in the creation of rural industry in Asia following the Green Revolution in cereals production. Yet it has been hard to demonstrate the existence of such spin-offs from crop growth in Africa, since additions to farm income have typically been spent on goods that are considered to be either imports to rural localities or displacements of potential exports from them. Thus, until now, the extra effects on production of rural income growth in Africa were thought to be lost (from the standpoint of local employment) to imports, or thought to displace production that would otherwise have been exported from the local region.

After reviewing the literature on agricultural growth linkages in Africa, this report examines the mix of farm and nonfarm goods and services that rural Africans purchase, and the implications of these expenditures for rural economic growth in five African countries: Burkina Faso, Niger, Senegal, Zambia, and Zimbabwe. In the West African countries, in addition to farm and nonfarm sectors, individual commodities are sorted into tradable and nontradable categories, and by geographic zones of interest: local, national, and multicountry regional. The same process is followed in Southern Africa for Zambia, but no assumptions are made on regional tradability because of lack of data. Fully comparable data were not available for Zimbabwe, but some similarities in rural consumption patterns could be detected.

The classification of goods as tradable or nontradable is based on the judgment of those who collected the data, extensive field inquiry into what was actually consumed, judgment as to where products consumed typically originated, and whether tradable substitutes (in the sense that their price movements were in tandem) were available locally. Food was a big item in household expenditures, and, as it turned out, many foods consumed were nontradables. Earlier studies also noted the high propensity to spend increments to income on food in Africa, but mistakenly classified virtually all important foods as tradables, following the assumptions made in Asia. Thus demand for additional food in the earlier studies was considered a "leakage": spending increments to income on tradable goods (including food) was thought to either decrease the quantity of goods available for local export or increase the amount the region spent on imports. It is not surprising that previous estimates of rural growth multipliers in Africa have generally been very low.

The results of this report are much more optimistic, largely because the underlying assumptions about tradability follow African conditions more closely, but also because of the unusually detailed data used on the flows of consumption expenditure over the year. These were from weekly or biweekly repeated interviews, which captured both food consumption patterns and total expenditures (a proxy for household income) especially well.

The way rural people spend increments to income is measured in the report as average budget shares - the percentage of total household expenditures going to a given group of goods and services - and marginal budget shares (MBS) - the percentage of the last unit of income earned that is allocated to the goods or services in question. MBSs are estimated econometrically and show the direct impact of unit income changes on consumption. The MBS for nontradables was 47 percent in Niger and about 33 percent in Senegal, suggesting that an ample share of rising incomes will be spent on items that would not otherwise have a market outlet and conversely no alternative source of supply. Examples are processed and unprocessed foods, inputs to agriculture, and services. If the new demand for these goods cannot be met because of supply rigidities, hefty price increases could result.

The MBS for nontradables as a group is the single most important determinant of the magnitude of estimated growth multipliers. In the four-sector, modified, semi-input-output model used, this was decomposed into farm and nonfarm items. Other determinants were technical coefficients of input use, value-added shares by sector, and the savings ratio. Multipliers were calculated by solving the set of regional equations that balanced consumption and availability of goods and services.

The report finds that the farm sector in Africa is better able to propagate income growth than previously thought. Growth in household income that comes from increases in agricultural production, perhaps spurred by new technology or changes in export prices, is largely spent on farm and nonfarm items that are nontradable, such as perishable foods, services, and locally produced nonfarm goods. Overall the report finds that adding US$1.00 of new farm income potentially increases total income in the local economy - beyond the initial $1.00 - by an additional $1.88 in Burkina Faso, by $ 1.48 in Zambia, by $ 1.24 to $ 1.48 in two locations in Senegal, and $0.96 in Niger.

Given the methodology used, these are upper bounds of the potential gains. Actual gains may be as much as 30 percent less, due to possible rigidities in the supply responsiveness of nontradables to price rises under African conditions. Even so, the results are substantial, suggesting that $1.00 of initial growth in rural agricultural incomes leads to an additional $1.00 on average of income from production of rural nontradables. This implies that the overall benefit of finding a way to boost rural incomes (from additional exports, say) on the supply side is probably twice as high as the immediate return from the activity that was promoted in the first place.