There is an increasing need for indicators that can track the impacts of agricultural policies and technologies upon gender inequalities at the national and international levels. A recent working paper commissioned by the CGIAR Gender and Agricultural Research Network reviews the body of published research that uses such indicators and recommends a set of robust indicators that can help measure these impacts, either using data that already exist, or data that could be collected through relatively simple additions to existing national and international surveys. The goal is not to measure empowerment specifically – that is done in the Women’s Empowerment in Agriculture Index – but to track changes with regard to two specific outcomes: 1) control over key agricultural resources, and 2) decision making about labor, income, and within groups or collective bodies. Since agricultural interventions are often targeted at a particular point in the value chain, the recommended indicators are disaggregated by resource type, such as land, livestock, or common pool resources.
One of the challenges in writing this paper was clarifying criteria for selecting the indicators. Such criteria relate to both conceptual and measurement issues. For example, if we want to measure how a project affected women’s access to land, we first need to answer the question “how do we define access to land?” (conceptual issue) and then we can ask “are data collected from interviewing only heads of household sufficient?” (measurement issue).
Five conceptual and five measurement-related criteria emerge as particularly significant (see Box 1). Although many of the recommended indicators do not meet all of these criteria, foregrounding the criteria could help us be more aware of the strengths and weaknesses of the indicators we use, as well as help us work on improving them.
Box 1. Criteria for evaluating indicators of gender gaps in control over productive resources
Conceptual issues: Does the indicator of a gender gap…
- capture the differences between control and use rights: where rights to control (to manage, exclude others, or alienate the resource) may not be the same as rights to use (to access the resource and withdraw the output)?
- capture de facto rights in preference to de jure rights?
- measure joint rights (shared among individuals but not necessarily equal) as well as sole rights?
- disaggregate rights and decision making by sex of the respondent rather than by sex of the household head?
- measure the value of resources as well as (or instead of) the quantity of resources?
Measurement issues: Is the indicator of a gender gap based on…
- quantitative data, with qualitative validation?
- data that can be readily collected at several different points in time?
- data that are easily replicable across sites, countries, or regions?
- data that are not resource intensive to collect because they are already being collected at scale or could feasibly be included in an existing large-scale survey?
- short- or medium-term outcomes (3-5 years) in preference to long term?
The paper explores each criterion in more depth, but two are worth highlighting here. Of the conceptual criteria, the first attempts to recognize the different layers of rights that individuals may exercise over resources (Schlager and Ostrom 1992). We recommend separate indicators of use and control to enable a more fine-grained understanding of rights claimed by men and women. Separate indicators of use and control also help us generate indicators that may change in the short to medium term (measurement criterion #5), in time to course-correct policy interventions. Patterns of use fit this criterion better than patterns of control, although the latter may be more definitively linked to gender equity.
The term “ownership” refers to varying combinations of these rights in different contexts. Thus, where possible, it would be better to use indicators of the specific rights we are interested in (the right to alienate, for example) rather than using the term “ownership”.
The review of the literature also suggests that project interventions more often affect the value of the resources such as land rather than their quantity. Our fifth conceptual criterion thus seeks to identify whether the indicator measures the value of resources owned and controlled (based on sale or rental price) in addition to measuring pure quantity of the resources.
Even these relatively simple criteria are hard to meet with data currently available. The good news is that small additions to existing surveys could help fill these gaps, underlining the need to continue advocating for more sex-disaggregated data.
Recording of the webinar with Smriti Rao discussing the paper is available here>>
About the author:
Smriti Rao is a development economist who studies the causes and consequences of gender and economic inequality in labor and credit markets. She has a particular interest in rural, agrarian economies, with a regional focus on India, where she has conducted fieldwork.
She teaches courses on Development, International Trade and Finance, and the Economic Systems of China and India. Smriti received her Master’s and PhD in economics from the University of Massachusetts, Amherst. She received her bachelor’s degree from the Birla Institute of Technology and Science, Pilani, India.
Currently, she is an Associate Professor of Economics and Global Studies at Assumption College, MA and Resident Scholar at the Women’s Studies Research Center, Brandeis University, MA.
This post is part of EnGendering Data, a blog on collecting and analyzing sex-disaggregated data to improve the knowledge base on the role of gender in agriculture and food security, maintained by the CGIAR Research Program on Policies, Institutions, and Markets (PIM).
Featured image credit: IFPRI/Yifei Liu, Flickr