Perhaps predictably, I find it hard to disagree with Steve’s arguments. There is a pro-smallholder and pro-science – even pro-GM! – position, drawing on a strong empirical record, that Paul completely misses in his attempt to slay the giants of romanticism. I will, therefore, confine myself to two main points:

The first augments Steve’s points about the comparative advantage of smallholder vs large-scale commercial agriculture. In low income economies, replacing labour with capital is often not efficient. This is true for many agricultural production tasks. Moreover, smallholder family labour is often better motivated and hence more efficient than the hired labour that large-scale farms have to rely on. In general, therefore, there are few economies of scale in agricultural production in Africa, although there may be in processing and marketing. That said, there are supply chains – most notably, export horticulture – where significant capital investments at farm level are unavoidable.

There are also economies of scale in traceability and other aspects of quality assurance. In such supply chains, the advantages of large farm organisation may outweigh the labour benefits of smallholder production. In a recent review of commercial agriculture in Africa for the World Bank (http://go.worldbank.org/XSRUM2ZXM0), we found that large-scale production had outperformed smallholder systems in export horticulture, sugar and flue-cured tobacco, but that smallholder production systems had outperformed large-scale in cotton and cashew, with strong performance under both forms in tea. The current debate has been prompted by the high food prices observed in 2008. Notably, food crop production in Africa remains dominated by smallholders.

The high costs of accessing and defending large landholdings in much of Africa may contribute to this. However, in a low income economy there are no obvious scale advantages in maize production and poor consumers are a long way from demanding the traceability and food safety assurance that could tip the balance in favour of large producers. Tellingly, where large farms do exist, they often choose to produce higher value crops than maize and other staples. Paul argues that “allowing commercial organizations to replace peasant agriculture gradually would raise global food supply in the medium term”. However, as Prabhu Pingali and others have shown for East Asia, market forces will tend to produce farm consolidation only when real wages in an economy rise well above levels seen in most of Africa today. When this happens, replacing labour with capital will make increasing sense and increasingly large plots will be necessary to generate an income for the owner comparable to that which could be obtained in an (attainable) off-farm job.

My second point augments one of Paul’s points. We can point to plenty of evidence showing that, where smallholders are supported through public or private delivery of support services (accessible input supply, seasonal finance, technical advice etc), they can compete strongly with large-scale farms in low income economies. However, large-scale farms do possess an important advantage: they can access such support services themselves (e.g. direct contact with commercial banks), whereas smallholders are heavily dependent on services being brought close to their farmgate. As Steve notes (not altogether approvingly), large-scale farms can even lobby for public infrastructure provision, something that smallholders have rarely been able to do.

The case for large-scale farms, therefore, looks stronger where states completely fail to provide or to encourage support services to smallholder producers. Without such service provision, smallholders are indeed more likely to be trapped in chronic poverty than to be drivers of agricultural growth. In recent years there have been encouraging commitments from African governments to increase their investment in the agricultural sectors of their countries. This is critical if smallholder production is to supply the ever-rising demand for food on the continent.

Colin Poulton, School of Oriental and African Studies, University of London