Land Quality and Agricultural Productivity: A Distance Function Approach
Increases in agricultural productivity are one of the main drivers of increases in agricultural
growth. Agricultural growth in developing countries makes major contributions to overall
economic growth and reductions in hunger and poverty. In more developed countries, increases
in agricultural productivity have served to keep food inexpensive for consumers and have kept
the agricultural sector competitive with other countries.
Economists have a long-standing interest in the study of agricultural productivity.
Researchers study cross-country differences in agricultural productivity to gain a better
understanding of the factors that are most effective in increasing agricultural productivity.
Although analysts have long recognized that land quality plays an important role in agricultural
productivity, land quality has been difficult to quantify and include in productivity models due to
data limitations.
An important component of productivity measurement is technical efficiency—the
measure of how well producers convert inputs to outputs when compared to a given standard.
Typically, when comparing producers (such as countries), efficiency is measured relative to a
common technology frontier that represents the maximum level of output that can be produced
with a given level of input(s). However, differences in production structure may limit the ability
of a producer to achieve efficiency relative to this common frontier. Land quality is one such
factor that affects production structure and thus the ability to achieve efficiency. For example, a
country must take its soil and climate as given and uncontrollable, at least in the short term, even
though they contribute greatly to total agricultural output.
Previous studies have attempted to incorporate the impact of land quality on agricultural
productivity by including a land quality proxy as an explanatory variable in an econometrically
estimated agricultural production function. For example, Hayami and Ruttan (1985) tried to