Are small farms still efficient?

The seminal work by Nobel laureate Amartya Sen on farm size and productivity triggered debates among academics regarding farm size and productivity (output per unit of land). In the Indian context, possibly in the 1960s, the findings from the field-level data sparked the debate on an empirical plane. The noted economist observed that an inverse relationship tends to exist between farm size and productivity - the larger the size, the lower is the output and vice versa. The symptom suggests that the yield rate of small farms exceeded that of the larger ones due to a number of reasons.

In the context of Bangladesh, late Dr Mahabub Hossain's findings corroborated Sen's observations. It may be mentioned here that the proponents of drastic and redistributive land reforms drive home their justification on the basis of these empirics.

However, after a long period and the hypothesis having been generally accepted across academics, Keijiro Otsuka, Yanyan Liu and Futoshi Yamauchi have introduced a lively debate. They argue that it is not the inverse but a positive correlation that exists between farm size and productivity. The authors opine that in the light of declining farm size, Asia is faced with weakened inverse relationship between farm size and productivity. In other words, the emergence of a positive relationship is apparently in evidence, indicating an increasing advantage of large farms.

In order to maintain efficiency of agriculture in Asia, so runs the recipe by Keijiro et al., farm size expansion must take place notwithstanding distorted land markets in many Asian countries including Japan, China and India. But first look at the statistics on land size that show that during the last few decades, the average farm size in Bangladesh declined from 1.4 ha  to 0.3 ha; in India from 2.3 ha to 1.3 ha; in Indonesia from 1.8 ha to 0.8 ha and in the Philippines from 3.6 ha to 2.0 ha. Japan, however, witnessed a rise from 1.0 to 1.8 ha. The authors reckon that family farms dominate agriculture almost throughout the world due to difficulty in monitoring hired labour.

In low-income (wage) economy, small farms with family labour-intensive production method are optimum; in high-income (wage) economy, large farms with labour-saving, capital-using and land-using production method are optimum. In other words, the long-run average cost curve of low-wage economy would turn flatter to steeper with the passage of time while that of the high-wage economy would run steeper to flatter. As output increases, average cost would rise fast with low-wage economy and vice versa.

The inverse relationship tends to be observed in low-wage economies when efficient small farms and inefficient large farms coexist. The positive relationship tends to be observed in high-wage economies, when inefficient small farms and efficient large farms coexist. "It is worth emphasising that when land markets function efficiently, only efficient farms exist so that neither the inverse nor positive relationship is observed. Indeed, the inverse relationship was often observed in South Asia but not in Southeast Asia except in the Philippines. This argument is particularly valid for grain production. Since the main reason for the inverse relationship between farm size and yield per hectare is the higher labour intensity of smaller farms, the inverse relationship will be lessened if profit per hectare is used in lieu of yield per hectare."

Keijiro et al lead us to a few case studies where the possibility of a positive relationship looms large. In Indonesia, for example, wage increase resulted in substitution of labour by machines particularly by large farmers. This augmented farm-size expansion by renting land. This has contributed to higher growth of income by large mechanised farms, and emergence of positive relationship between farm size and crop yield per hectare. Almost a similar trend was found in Vietnam. The wage growth led to increased machine renting particularly by large farms at the behest of farm-size expansion by land renting (e.g., from 1 ha to 1.4 ha in South). All these have lessened the inverse relationship between farm size and rice yield; the elasticity of crop yield with respect to farm size changed from -0.16 in 1992/98 to -0.06 in 2006/08.

In China, average farm size is only 0.6 ha and it increased by only 0.05 ha per year in the recent decade. There also wage increase resulted in out-migration plus land-renting plus machine rental. The higher growth in income among large farms in the 2000s speaks of the emergence of scale advantages; consequently there is in evidence the emergence of positive relationship between farm size and productivity. The development of active machine rental markets in China, however, helped reduce handicap for small farms. The authors are of the view that the efficiency of large farms will continue to increase in India in coming years.

Despite the emergence of scale economies due to mechanisation, small farms continue to dominate due to imperfect and regulated land markets. The implications are far-reaching: cost of food production increases due to the continued rise of wage rate, leading to the loss of comparative advantage in agriculture. Food demand exceeds supply, resulting in massive import of food grains in Asia as a whole and the hike of world food prices. It is thus no wonder that the poor throughout the world will starve if Asia as a whole becomes a massive importer of food grains, as food prices would shoot up. In order to reduce undue reliance on grain imports, institutional innovations are needed in India, China and other Asian countries to facilitate farm size expansion.

The emergent farm features, as portrayed by the authors, seem to apply to Bangladesh also. The wage rate is rising, mechanisation is taking place very fast (mostly two-wheeler), and the farm size is declining. Yet, till now, there is little indication that large farms have more productivity than smaller ones at least in terms of yield per unit of land. Whether increase in farm size is needed to arrest declining productivity is a new hypothesis that requires to be tested on an empirical plane.   The proponents of the new hypothesis deserve special thanks for providing us with a 'wake-up call'.

The writer is a former Professor of Economics at Jahangirnagar University.

abdul.bayes@brac.net