A few months ago, I came across a reference to a review of the literature on contract farming by Otsuka et al. (2016) in a paper I was handling at Food Policy. Given how much work I have done on contract farming so far, I made sure to make time to read Otsuka et al.’s review.
One of the things that grabbed me in their review was the part where Otsuka and his coauthors write:
[i]t is less clear … how far [contract farming] improves farmers’ welfare. Although many empirical studies found positive effects of [contract farming] on the income from contracted crops, such evidence is not conclusive, because crops and products under [contract farming] are usually labor-intensive so that income from other crop production or nonfarm activities might be sacrificed … [I]ncome from other sources should be analyzed along with income from contracted production to identify the net income gain and the degree to which [contract farming] sacrifices other income … To our knowledge, such a study is lacking … (p.369)
When I read that, I realized that I could actually answer that question using the data I used in my 2012 World Development article and in my 2017 American Journal of Agricultural Economics article with Lindsey Novak, and that I could do so rather quickly.
So as soon as I finished reading Otsuka et al.’s review, I set to work looking at whether participation in contract farming had spillover effects on various sources of income. The data disaggregates income in:
- Income from contract farming,
- Income from livestock,
- Income from agricultural sources other than contract farming and livestock,
- Income from labor-market participation, and
- Income from non-farm enterprises.
Obviously, participation in contract farming will have a direct effect on the first category above, so I looked at sources of income 2 to 5. In the space of a weekend, I ran the estimations and drafted the paper.
Last week, I was in the UK to present at the annual conference of the Centre for the Study of African Economies at Oxford (I was presenting another paper on contract farming; that topic is the gift that keeps on giving!) I exploited the insomnia caused by jet-lag to finish and polish the paper.
In that new working paper, I report my findings. As it turns out, participation in contract farming appears to have positive spillovers on agricultural income sources (i.e., income from livestock, and income from agricultural sources other than livestock or contract farming), but negative spillovers on non-agricultural income sources (i.e., income from labor-market participation, and income from non-farm enterprises).
Here is the abstract:
An important literature has established that participation in contract farming leads to higher incomes and has a number of other beneficial effects on the welfare of participating households. Yet no one has looked at the opportunity cost of and the various trade-offs involved in participating in contract farming. I look at the relationship between participation from contract farming and income from (i) labor markets, (ii) non-farm enterprises, (iii) livestock, and (iv) agricultural sources other than livestock and contract farming. Using data from Madagascar, I find that participation in contract farming is associated with a 46 percent decrease in how much income per capita it derives from labor markets and a 23 percent decrease in how much income per capita it derives from non-farm enterprises, but also with a 25 percent increase in how much income per capita it derives from livestock and a 24 percent increase in how much income per capita it derives from agricultural sources other than livestock and contract farming.
What I find interesting in all this is that this seems to contradict some structural transformation narratives, according to which the development of a more industrial agricultural sector (say, via contract farming and agricultural value chains) will free up workers to work outside the family farm and will lead to the development of businesses. My findings seem to suggest that before this can happen, households seem to turn inward, in some kind of agricultural involution process due to specialization involving low levels of agricultural technology.