A meta-regression analysis including 167
farm level technical efficiency (TE) studies of developing
and developed countries was undertaken. The
econometric results suggest that stochastic frontier
models generate lower mean TE (MTE) estimates than
non-parametric deterministic models, while parametric
deterministic frontier models yield lower estimates than
the stochastic approach. The primal approach is the
most common technological representation. In addition,
frontier models based on cross-sectional data produce
lower estimates than those based on panel data whereas
the relationship between functional form and MTE is
inconclusive. On average, studies for animal production
show a higher MTE than crop farming. The results also
suggest that the studies for countries in Western Europe
and Oceania present, on average, the highest levels of
MTE among all regions after accounting for various
methodological features. In contrast, studies for Eastern
European countries exhibit the lowest estimate followed
by those from Asian, African, Latin American, and
North American countries. Additional analysis reveals
that MTEs are positively and significantly related to the
average income of the countries in the data set but this
pattern is broken by the upper middle income group
which displays the lowest MTE.