![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
Abstract
African agriculture is often dominated by smallholder farms composed of multiple plots. In policy circles it is often assumed that fragmentation coupled with small farm size is an impediment to increasing yields and thus decreasing poverty and food insecurity. There also exists an influential literature that explores the inverse relationship between farm size and yield. While there are many studies internationally, few have been conducted in Africa. Using an Ethiopian national survey, we explore the relationship between yield, farm size and fragmentation. We find an inverse relationship between farm size and yield, and a positive association between yield and land fragmentation.
Acknowledgements
For comments on the manuscript the authors would like to thank three anonymous reviewers for this journal, as well as James Boyce, Michael Ash, and Anita Milman.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 This appeared in a special issue of the Journal of Agrarian Studies (2004): volume 4, issue 1–2.
2 Due to labor market imperfections, the household applies labor beyond the marginal product at the prevailing regional market wage. This is because the market wage is often not available to all members of the household.
3 This is an econometric technique that allows for comparison, for instance, within villages. The idea is that by comparing like to like (within villages for example) the potential for omitted variable bias is reduced.
4 Since measuring land quality is challenging and expensive, using an instrumental variable is an econometric technique used to estimate, or stand in for, land quality while remaining uncorrelated with the error term. The instrumental variable is a variable that is correlated with the actual variable the researcher is interested in – in our case, for example, land quality.
5 This is not typically the case due to the expense involved in large-scale soil sampling schemes that go along with the rural survey. In order to deal with this gap in the data, we use fixed effects in the model.
6 One of the frequent goals of land reforms is the consolidation of plots for the sake of ‘efficiency’.
7 The government established this limit, concluding 10 hectares was the maximum a household could cultivate efficiently with a pair of oxen.
8 The PAs were seen by many as supporting the villagization process promoted by the Derg.
9 There are nine Killils across Ethiopia. These are ethno-linguistically based regional states.
10 These are lower levels of regional governments.
11 Landless households account for 10 percent of our sample, though we believe this number is underrepresented since the EEA/EEPRI used PAs to identify households. To be a member of the PA, one needs to hold land. Thus, the landless tend to be bypassed by the PAs.
12 The use of specific crops in ceremonies and rituals, regional dishes, medicine, etc. may drive the variation between market and shadow price. In other words, culture and tradition affect the shadow price of many crops and therefore influence peasants’ resource allocation.
13 Since we have multiple observations for each agro-ecological zone, the fixed-effects model allows us to compare farms within the same agro-ecological zone, thus removing the potential omitted variable bias.
14 Data is normally distributed.
15 In this instance, we square the number of plots to see if the effect diminishes with a larger number of plots.
16 We would expect that the farther away a plot is the more likely it is to be of different characteristics than plots closer by; therefore, the distance to a plot may represent variation in the plot’s quality and type of soil, thus allowing the peasant to improve diversification on the farm.
17 We are careful not to claim that the IR represents increases in labor efficiency. On the contrary, yield increases could be due to increase in actual labor performed. In this respect, Johnston and Le Roux (Citation2007) raise important questions about the unitary household model and whether the labor from other household members may be coerced, meaning the increase in production may be at the expense of the individual welfare of members of the household and further distribution of gains may not be equitable. Given our data we are unable to examine intra-household allocation of labor. As we stated earlier, we do believe that in the case of some countries, such as Ethiopia, there is available labor in the rural areas due to the inability of the non-farm labor market to absorb the labor. The evidence of this is decreasing farm sizes over time as households divide up land among multiple heirs. If conditions changed and there was an increased labor absorption by the non-farm sector, we would expect that the ability of small farms to produce more per unit of land would be severely diminished.
Additional information
Notes on contributors
Mark Paul
Mark Paul is a postdoctoral associate at the Samuel DuBois Cook Center on Social Equity at Duke University. He holds a PhD in economics from the University of Massachusetts Amherst. His research focuses on inequality by race, class, ethnicity and gender; environment economics; development; political economy; and agricultural economics.
Mwangi wa Gĩthĩnji
Mwangi wa Gĩthĩnji is the author of Ten millionaires and ten million beggars (Ashgate Press, 2000) and co-author (with Robert Pollin and James Heintz) of An employment-targeted economic program for Kenya (Edward Elgar, 2008), as well as numerous articles and book chapters. His research interests are in the political economy of Africa with a particular interest in inequality. Mwangi was Chair of Africana Studies and an associate professor of economics at Gettysburg College in Pennsylvania before moving to the University of Massachusetts Amherst in 2006, where he is an associate professor of economics. Email: [email protected]