Gender & Agricultural Data and Methodology Previous studies looking at the gender gap in Productivity in Malawi sub-Saharan Africa rely largely on data from small-scale surveys, and are limited in terms of By Talip Kilic, Amparo Palacios-Lopez geographic coverage, topic, or attention to intra- and Markus Goldstein household dynamics (or, in some cases, all three). This study uses data from the Third Integrated Agriculture is critically important to both Household Survey (IHS3), collected from March Malawi’s economy and to its social fabric. The 2010 to March 2011 by the Malawi National sector accounts for 31 percent of gross domestic Statistical Office, with support from the LSMS-ISA product, and 84 percent of Malawian households initiative. The IHS3 data covers 12,271 own and/or cultivate land. The majority of households. The full sample consists of 16,372 farming households practice subsistence plots, 26 percent of which are managed by agriculture, and their living standards are directly women. affected by the inconsistent agricultural performance that Malawi has seen over the last Our econometric approach applies a two decades. These impacts are especially acute decomposition methodology that has been for the poorest households. widely used in labor economics, starting with studies by Oaxaca (1973) and Blinder (1973). To Agricultural growth has been shown to our knowledge, this is the first time that this particularly benefit the poorest in the developing method has been used to understand the gender world. Poverty in Malawi is especially widespread gap in agricultural productivity. We look at the among female-headed households, suggesting average difference in agricultural productivity that investing in agricultural growth has benefits (defined as gross value of output per hectare) on both for poverty reduction and for gender male- and female-managed plots, and seek to equality. Yet systematic gender differences determine how much of the gender gap is driven persist in agricultural productivity across sub- by differences in: Saharan Africa, mostly due to differences in: (i) access to and use of agricultural inputs, including i. Levels of observable inputs or attributes, improved technologies; (ii) tenure security and such as the education level of the plot manager, related investments in land; (iii) market and credit the amount of inorganic fertilizer application, etc. access; (iv) human and physical capital; and (v) We refer to the impact of the collection of these informal institutional constraints affecting factors as the endowment effect. farm/plot management and the marketing of agricultural produce. Addressing these gender ii. Returns to observable inputs or differences could result in tremendous attributes, such as the monetary return that a productivity gains. The FAO reports that if female farmer earns form applying one kilogram of farmers had the same access to productive fertilizer per hectare. We refer to the impact of resources as men, they could increase yields by the collection of these factors as the structure 20-30 percent, which could increase total effect. agricultural output in developing countries by 2.5 to 4 percent and lift 100 to 150 million people Thus, we seek to quantify the mean gender gap out of hunger. as well as the relative contributions of key inputs Living Standards Measurement Study Brief Series www.worldbank.org/lsms-isa and returns to these inputs. Identifying the gender lines may account for a relatively less factors driving the gender gap in this manner is efficient usage of fertilizer by female farmers. crucial for informing policy interventions aimed at addressing the gap at its roots. Finally, the gender gap widens as agricultural productivity increases. While the gender gap in A second contribution of our study is to move Malawi is 25 percent at mean productivity, it beyond the “average” effects, and to break down ranges from 22 percent at the 10th percentile of how the key factors drive the gender gap at the agricultural productivity distribution, to 37 different points in the agricultural productivity percent at the 90th percentile. At the same time, distribution. Since farmers at different levels of the returns to key inputs decrease progressively productivity may face different constraints – or for female farmers but not for male farmers, similar constraints but at varying levels – we try meaning that the structure effect increasingly to tease out the contributions of key factors explains the gender gap as one moves up in the towards the gender gap at the low-, mid- and agricultural productivity distribution. One high-level of agricultural productivity. possible explanation is that even as female farmers use higher levels of productive inputs, Results they are less effective in achieving the On average, we find that female-managed plots combinations of inputs that result in the greatest in Malawi are 25 percent less productive than yields, and thus experience lower returns than plots that are managed by males. The their male counterparts. endowment effect explains 82 percent of this gender gap. In particular, female-managed plots Conclusions are constrained by lower use of inorganic Our findings suggest that a large and significant fertilizer, lower use of household adult male difference in the levels of inputs is the central labor, lower production of high-value export factor behind the gender gap, particularly for crops, and restricted access to agricultural tools. farmers at lower levels of agricultural productivity. On male-managed plots, higher Female plot managers try to compensate for levels of household adult male labor and area these deficiencies with higher levels of under export crop cultivation widen the gender household adult female, household child, and gap, while household and childcare exchange labor, but this is not enough to responsibilities restrict the time that female plot overcome the differences in productivity. managers can dedicate to farming. Ensuring that female plot managers have similar years of The remaining 18 percent of the gender gap – schooling as men and apply similar levels of non- the structure effect – is driven by differences in labor agricultural inputs could reduce the mean returns to the use of household adult male labor, gender gap by 50 percent. Future research will and the application of inorganic fertilizer. explore why inequalities in time use, as well as access and returns to agricultural inputs, Not only do adult males in the household spend continue to persist. This will be a first step less time on female-managed plots, but the time towards informing policies that are designed to that they do devote is less productive than when alleviate the gender gap at its roots. they work on male-managed plots. One reason for this may be that female plot managers are This brief is based on: Kilic, Talip, Palacios-Lopez, able to provide less supervision due to other Amparo, and Goldstein, Markus (2013). Caught in a household responsibilities. Indeed, our study Productivity Trap: A Distributional Perspective on found that a greater child dependency ratio Gender Differences in Malawian Agriculture. World decreases the productivity of female-managed Bank Policy Research Working Paper No. 6381, plots but has no effect on male-managed plots. Washington, D.C.: The World Bank. This points to childcare responsibilities falling primarily on women, preventing them from For more information, please visit: providing as much labor supervision as male plot managers. www.worldbank.org/lsms-isa In terms of inorganic fertilizer use, female Or contact us at lsms@worldbank.org farmers not only apply lower levels of this input, but the fertilizer that they do apply does not yield as many benefits. A knowledge gap along Living Standards Measurement Study Brief Series www.worldbank.org/lsms-isa