Farm size and productivity

Abdul Bayes This article invokes the debate of farm size and productivity in the light of available information on the most recent changes. Some researches have already reached the conclusion that, there exists an inverse relationship between farm size and productivity. We can now draw upon the data set that we have for recent years to accept or reject the hypothesis. And to this effect, we shall take the help of the results of the surveys conducted in 1988 and 2008. The hypothesis that we are going to test is: small farms are more efficient than large farms. Interest in the relationship between farm size and productivity (output per unit of land) is not new. It dates back to the early days of development economics. Since that time, an avalanche of empirical research on this has showed a negative relationship between these two variables. The majority of the researches opined that, small farmers are ahead of large farmers in terms of yield or production per unit of land. In other words, on a plot of land, small farmers produce more than large farmers. Later, especially in the context of land reform and rural development, the idea emerged with important ramifications. It is being argued that, the negative relationship between farm size and productivity is a genuine argument in support of drastic land reform. That means, if the size of farm could be equalised, then efficiency and equity will increase and the attained success would impart dynamism in rural economy. And in support of this contention, examples of land reforms are drawn from the experiences of Japan, South Korea and Taiwan. It is opined that, land reform-led economic transformation of those countries contributed to creating agricultural surplus, consumer markets and political stability which are preconditions of sustainable development. In the context of Bangladesh, a maiden attempt has been made with the research on farm size and productivity. According to Mahabub Hossain, the low opportunity costs of family labour enable small farmers to use them intensively and for this reason, they receive more output (yield) than the farms of big sizes. He reckons that, due to the high costs of monitoring and supervising labour force, large farmers use labour less intensively. However, K.A. Toufique has drawn a reverse conclusion. According to him, because of low transaction costs of time in high potential area (such as Madhupur), large farmers become efficient. On the other hand, small farmers remain more efficient in low potential area (such as Chandina). Therefore, Toufique argues, the relationship between farm size and productivity may not be negative everywhere. It appears that, crop-specific information on yield levels that we collected, do not consistently support the negative relationship in all cases. But we have to accept the hypothesis only for all varieties of paddy. May be, the conventional notion failed to get support due to some techno-ecological considerations that were, possibly, dominating the production of specific crops. Let us elaborate the observation citing a few examples. In 1988, in the case of the yield of TV aman, small farmers had marginal edge over large farmers. But the inverse relation was not consistent as the in-between farms had more yield than large or small ones. But by 2008, the yield per unit of land was the largest for the functionally landless farms, followed by other farm-size groups. The negative syndrome emerged quite visibly in this case. For MV aman in 1988, however, a positive relationship between farm size and productivity could be noticed. But in 2008, we observed just the opposite; the negative relationship was not consistent for all types of paddy. For example, in 2008, small farms had 6.0 per cent higher yield than large farms (4.18/ton vs. 3.94 ton/ha) and the difference with other groups was even more. Again, in the base year, the difference was very distinct when small farms had 2.68 ton/ha compared to large farms at 2.08 ton/ha - a difference by about 29 per cent. The observations, thus, go to confirm the conclusions reached by M.A Taslim. We can accept the arguments of both, although the respective conclusions were drawn from different angles. It is true that, small farms tend to have low opportunity costs of family labour and hence can be used intensively. On the other hand, the cost of monitoring and supervision of labour is relatively high for large farms to discourage them from using labour intensively. We also reckon that, the desperation of small farms for food security from tiny parcels always makes them vigilant over crop management and this helps to raise productivity. The famous observation of Nobel-winning economist Theodore Schultz might be true for Bangladesh: small farmers are rational and efficient. Therefore, it is not unlikely that these rational and efficient farmers give everything of their lives to get the maximum production from the given land. For others, the economic boat may not be rocked by production shortfalls. But the same is not true for small farmers. A small crop failure can become a huge curse for them. Another debate on farm size and productivity is centred on the tenurial status of the farms. Following Adam Smith and Alfred Marshall, many argue that, the yield level of the tenants are likely to be lower than owner cultivators. The main reason in this case is incentive problem. Since, the tenants have to shed one-half of the incremental output; they lose interest in producing an additional amount. Besides that, a farmer is not that much careful for others' land as he is for his own land. When all types of paddy are considered together, the output per unit of land of the share-croppers is found to be the highest, compared to that of owner operators. Share-croppers have topped the list in the case of boro and MV aman in 2008. Apparently, our findings reject the contentions of Marshall and Adam Smith. That is, we observe no shortage of productivity, even after paying half of the output to the owners. Possibly, hypothesis of Cheung did work in this case: if the owner shares some of the inputs with share-croppers then, share cropping could also be efficient. The other reason could be that (and it is really so), in the case of adoption of MV paddy, share-croppers are not lagging behind the owner operators. And finally, low opportunity costs of labour and high costs of supervision and monitoring might have upset the conventional impressions. The same results also follow for non-paddy crops. Especially for wheat, we notice a discernible positive relationship between farm size and productivity. That is, large farms tend to have more yields per unit of land. This could be due to the fact that, the crops are mostly grown in Dinajpur district where the average size of land is relatively very large. Added to this is another factor: wheat is considered to be relatively a risky crop (e.g. depends on the duration of winter) that could become expensive for small and tenant farmers. But for potato and pulses, we observe an inverse relationship between farm size and productivity. Again, especially for tenant farmers, the output level of potato is higher than that of owner-farmers. Abdul Bayes is a Professor of Economics at Jahangirnagar University. [email protected]