Sources of crop productivity in Pakistan, 1960-1990

It analyzes aggregate time series data (1947-90), as well as cross-sectional household data from large and small samples; desegregates the time trends into discrete epochs and the national trends into provincial ones; decomposes output changes into area, yield, and multiple effects; and reviews crop-specific research reports as well as reports that look more broadly at farming systems and the agriculture sector as a whole. At a heuristic level, the study systematically relates changes in crop acreage and yields to changes in profitability, policy, irrigation, and technological change. Combined results of aggregate, time-series, and farmer-level, cross-sectional analyses imply the following conclusions. (1) While periods of high output growth for wheat and rice appear to be dominated by yield rather than area effects, the area effect for cotton (1960 to 1980) appears to have been a more important source of major production increases. (2) There is considerable variation in the relative importance of the various sources of productivity over time, across cropping systems, and between crops. (3) However, price policy effects, input market deregulation (especially for agrochemicals), technological change (and the research that generates it), and water availability have had powerful and enduring influence on all crops over time. Extension is not seen to have had a positive impact and the impact of water management, and thus far soil quality, on productivity has not been adequately researched at the farm level. (4) There is reason to believe that the uptake of available technology in Pakistan has been slow, and that, in particular, small farmers have been slower than others in adopting new technology, although the difference between large and small farmers has tended to disappear over time with the development of markets and the spread of relevant knowledge. (5) Only in the case of wheat (by virtue of the Pakistan Agriculture Research Council-Centro Internacional de Mejoramiento de Maiz y Trigo Program) do recent field-level data exist that integrate agronomic and economic variables in analyzing the constraints on productivity -- although these studies do not incorporate irrigation variables. (Author abstract)