Is there a case for support for smallholder agriculture? A response to Palmer and Sender

The paper by Kim Palmer and John Sender makes the following argument: Claim (i): Rural households are currently not benefiting much from agriculture. Claim (ii): The current situation cannot be improved by policy changes (for example, in agricultural policy). Conclusion: We should direct resources away from agriculture, to non-farm souces of income (for example, wages, social grants). Palmer and Sender set their argument against the advocates of land reform and agrarian change who argue that land reform is important to promote rural livelihoods. They argue that the availability of more land will not improve the livelihoods of poor rural households. The evidence is primarily statistical (which is on the whole very well presented). But the basis of their evidence does not lead to such a policy conclusion, because there is a missing premise. There are many other kinds of policy changes apart from land reform which could improve the effectiveness of agriculture. For example, the provision of productive rural infrastructure (farm roads, fences, dams, boreholes, chearing sheds, bridges); the provision of rural credit; more improved extension services; and improved marketing facilities and systems. Palmer and Sender's conclusion, therefore, does not follow from their statistical evidence. They suggest that social grants wxould be a much better intervention than agrarian reform. But this is to make a false comparison. The social grant system makes R61 billion available (2006-7 budget), to be increased to R66 billion (2007-8). Compared to this, the entire department of agriculture (national and provincial) get 51.?9 billion (2006-7 budget), to be increrased to R2.1 billion.