Farm size and productivity
Abdul Bayes | Published:
November 30, 2015 19:23:00
October 23, 2017 15:55:58
'Small Farms: Changing Structures and Roles in Economic Development' is a study undertaken by Joachim von Braun and Alisher Mirzabaev of the Center for Development Research (ZEEF), Bonn. Drawing on various empirical studies, and as paraphrased below at times, the authors make a surprising observation. It is that the negative relationship between farm size and productivity - that we assumed in the past through rigorous research in Bangladesh and in India - may not hold water any more. It may be mentioned here that the negative relationship in a predominantly labour-intensive, less mechanised agriculture owes much to labour market imperfections where small farms apply excessive amounts of family labour than under optimum conditions. Negatively association also occurs between crop yields and on-farm capital investments (such as in China) as that between maize yields and plot size due almost to the same reason.
In Bangladesh, for example, expansion of migration, non-farm activities, remittance etc. are putting pressure on agricultural labour market for which the benefits of family labour force on small farms fade. Again, the growing mechanisation in agricultural practices has also reduced the demand for labour thus seizing upon small farms' yield advantage over larger ones. Joachim von Braun and Alisher Mirzabaev also observed the similar pattern through a global lens. The emerging converse positive relation happens, they argue, as agricultural production gets more capital-intensive. This supports the contention that recent data from field survey in Bangladesh that farming is becoming capital-intensive for which the traditional edge of small farms over larger ones is diminishing in terms of productivity. Researchers indicate two factors that might be behind the reversal of the inverse relationship: "Excessive land fragmentation might have made some farms too small for efficient operation and recent technological advances, including in information and communication technologies, might have made it easier to supervise labour, thus cutting down on the transaction costs of hired labour. Yet another strand of studies finds the relationship between farm size and crop yields to be U-shaped. The conclusion may be that the findings on the inverse relationship might have been influenced by small variations among farm sizes in many countries (where all farms are relatively small), but may emerge at much larger sizes (out of sample for many studies focusing on developing countries)".
Several studies found small farms to have higher land productivity than bigger farms due to higher incentives and productivity of family labour, especially in Asia where labour is more abundant than land. For example, decreasing returns to scale in agricultural production were found in East Java, Indonesia and in Pakistan. However, researchers have also observed that in certain cases, once the varying degrees of soil fertility and land potential (irrigated vs. rain-fed) are taken into account, the diseconomies of scale in land productivity between small and large farms may disappear. Moreover, there are numerous evidences that small farms could be less efficient in terms of labour productivity conclude that distinct advantages of small farms are present in cases when the main agricultural input is family labour and there is very little use of external inputs. The production is chiefly for home consumption with whatever surpluses sold to small-scale traders.
Macro-level cross-country comparisons in Asia show that average farm sizes are positively associated with agricultural value added per worker. This relationship is even stronger in countries of Sub-Saharan Africa (SSA). The comparison of agricultural growth rates with changes in farm sizes does not show a consistent picture for Asian countries. The overall trend seems to show that increase in farm size is associated with faster agricultural growth, but this seems heavily influenced by just a few countries (Tajikistan, Uzbekistan, South Korea and Vietnam). The results for SSA countries are showing little association with farm size changes and the rates of agricultural growth.
Most of the Asian and SSA countries have very small average farm sizes. So passing from say 0.4 ha to 0.8 ha may not necessarily have any strong qualitative change that may influence agricultural growth rates while passing from, say average 5-10 ha to 20-25 ha may have bigger impacts. This is also in line with the hypothesis of U-shaped relationship between farms size and farm productivity indicated above.
Many small farms are no longer producing primarily for home consumption, but are integrated, to various degrees, into labour market and agricultural value chains. They also use a variety of external inputs, and are increasingly faced with ever more concentrated supply chains. However, small farms are persisting despite many of these changing factors. Small farm households operate in various markets - labour, land, capital, crops, animal products - and that may give degrees of freedom to risk management. The wider adoption of information and communication technologies (ICTs) can greatly help access to market information and knowledge by small farmers which is also important for resilience.
In conclusion, the researchers reckon that there is no longer a clear-cut answer to the question whether small farms perform more productively, and if so, in what circumstances. Certainly, the performance of small farms is affected by a variety of accompanying factors related to policy, institutional, market and agro-ecological conditions. In view of the rising role of innovation (total factor productivity, TFP) in agricultural growth, we would expect that agricultural production grows more and better if accompanied with a strong knowledge society. The comparison of average farm size and knowledge economy index shows positive association globally. However, the relationship seems weaker in Asia.
The writer is Professor of Economics at Jahangirnagar University.