Population land use and the environment in developing countries: what can we learn from cross-national data?

Environmental degradation is the key word for the 1990s. This review of the ways in which population change can influence land use patterns was an initial attempt to advance the limited knowledge base on environmental issues. Cross country and world regional data were used to exemplify the linkages between land use and environmental change for rural areas of developing countries. Data quality was best for demographic measures less so for land use data and much weaker for environmental data; quality even for demographic data still varied by country. Land use data may be dated and based on limited agricultural census. Desertification and soil erosion figures were non existent and deforestation figures were variable by source. Population data included rural population growth from the 1960s to the 1980s and population density. Land use variables included total land area devoted to agricultural uses in 1965 and 1985 median changes land per agricultural worker the 1987 per capital Gross National Product and the average kilograms of fertilizer used per hectare of cropland for 1975-77 and 1985-87 (agricultural intensification). World Resources Institute data were used for the average annual loss of closed canopy and open canopy forests. Other variables included land and labor productivity and ungraphed variables: the Gini coefficients for land distribution and income equality the population-supporting capacity of land equality pesticide and tractor usage food production and total fertility rates. Analytical problems other than poor data quality and missing data also exist for determining how to measure some concepts and what measures of what variables should be compared with what measures of other variables. Shortcomings were identified as unavailability of the change in rural population density data lack of specification of quality of land or whether land is in use or fallow lack of specification of excessive fertilizer use or type of fertilizer lack of data on losses in irrigation by cause and inadequate and missing environmental data. Results of simple correlations for example indicated a positive relationship between population growth and land use but it was weaker than expected and conditional. If Brazil is included there is weak inverse correlation between rural population growth and loss of forests but it is positive and insignificant if Brazil and countries reporting zero closed forest loss are excluded.