How Agribusinesses Can Measure Rural Women's Participation in Their Supply Chains

FAO reviewed agriculture and rural development policy documents from 68 countries. Three quarters of them recognised women's roles or the challenges women face. Nineteen per cent contained a gender-related policy goal.

That is the gap this article is about, and it reproduces itself almost exactly in corporate supply chain reporting. Mentioning rural women is close to universal. Measuring whether anything changed for them is rare, and the reason is that most of what gets counted is activity rather than outcome.

The Day And What It Marks

Today is the International Day of Rural Women, observed on 15 October, and the 2026 theme is "Rural Women Farmers Caring for Mother Nature," honouring the work rural women do conserving seeds, protecting biodiversity, managing water and restoring land.

The structural figures behind the day are worth having to hand. Women make up close to 40 per cent of the agricultural labour force, and more than half in parts of Africa and Asia, while owning around 15 per cent of agricultural land. In wage employment in agriculture, FAO finds women earn 82 cents for every dollar men earn. On farms of equal size, women farmers show a 24 per cent productivity gap against men, which FAO attributes to unequal access to inputs, services and secure tenure rather than to farming ability. Closing the productivity and wage gaps together would raise global GDP by nearly a trillion dollars and reduce the number of food-insecure people by 45 million.

For a buyer, the operative part is the productivity gap. It says the women in your supply base are producing below their capability because of constraints you can partly see in your own data.

Activity Versus Outcome

The distinction is old and routinely collapsed. An activity is something your programme did. An output is what it produced. An outcome is a change in the condition of the person it was aimed at.

"Four thousand women trained" is an activity. So are women reached, groups formed, sessions delivered, kits distributed and awareness raised. They are all perfectly legitimate things to track for programme management, and none of them tells you whether a single woman ended the year with more money, more say or more security than she started with.

Outcome measures have an awkward property that explains their scarcity: they can go down. A training count only ever rises. The share of your supply contracts held in a woman's own name can fall when you add a new sourcing region, and that is precisely the information you wanted.

Four domains cover most of what matters, and each has a familiar activity measure sitting in front of a harder outcome measure.

Access: Supplier Registration

Start with the registration question, because it quietly determines everything downstream.

In a great many sourcing models the registered farmer is a man and the woman doing a large share of the field labour appears in the system, if at all, as a household member. She can attend every training you run and still not hold the contract, the delivery receipt, the input credit or the payment. Count her attendance and you will report rising participation on an unchanged structure.

The outcome measure is the share of supplier registrations, delivery records and input credit agreements in a woman's own name, tracked over time and by region. Expect it to be low. Men hold greater ownership or more secure tenure over agricultural land than women in 40 of the 46 countries reporting on SDG indicator 5.a.1, and registration systems frequently key off land documents, so the number reflects a real constraint rather than a reporting failure. Knowing its level is the first honest thing you can say about access.

A second measure worth adding: among households supplying you, how many have more than one registered supplier? Joint or dual registration is one of the few interventions that changes the structure rather than decorating it.

Payment: Account Ownership

Payment is where measurement is easiest and most often avoided.

The Global Findex data for 2024 shows 73 per cent of women in low- and middle-income economies holding a financial account, up seven percentage points since 2021, with around 700 million women worldwide still without one. In sub-Saharan Africa women remain 12 percentage points less likely than men to hold any account, though about 80 per cent of women in developing economies own a mobile phone.

One Findex finding should change how buyers think about this. Around 60 per cent of women with a bank or similar account opened it in order to receive a digital payment from a government or an employer. The payment is not a reward for inclusion. The payment is the mechanism of inclusion.

So the activity measure is that you digitised payments. The outcome measures are the share of your payments made to an account in the woman supplier's own name, and the share of women suppliers who report deciding how that money is used. The second question requires asking her, separately, which is the part most programmes skip.

Roles: Horizontal And Vertical

Headcount in the field is a horizontal measure. It tells you how many women are present and nothing about where.

The vertical measure is the share of women in the roles that carry discretion and pay: lead farmer, aggregator, grader, quality inspector, extension agent, warehouse supervisor, cooperative board member, and the buying positions inside your own company that set prices and allocate contracts. A supply chain can be 60 per cent women by headcount and have no women anywhere a decision is made.

The useful technique here comes from the project-level Women's Empowerment in Agriculture Index, developed by IFPRI and partners across 13 agricultural development projects in Africa and South Asia. Pro-WEAI organises its indicators into three domains of agency: intrinsic agency, or power within; instrumental agency, or power to; and collective agency, or power with. Its most borrowable feature is the gender parity index, which compares a woman's empowerment score with that of the man in her own household rather than with a national average.

That household comparison is the measurement idea most worth stealing, because it removes the escape route. Regional averages let a programme look good while the women inside each supplying household remain exactly where they were.

Grievance: Channel Use

A grievance channel that exists and is never used by women is not evidence of an absence of grievances.

The eight effectiveness criteria under Principle 31 of the UN Guiding Principles set the standard: legitimate, accessible, predictable, equitable, transparent, rights-compatible, a source of continuous learning, and based on engagement and dialogue. OHCHR guidance on meeting them is specific about what blocks access, and the list reads like a description of rural women's circumstances: cost and lost wages, childcare difficulties, literacy and digital skills gaps, language barriers, restrictive time limits, fear of retaliation, and lack of confidence tied to age or social status.

The same guidance is equally specific about how to tell whether a mechanism works. Publish the number and types of grievances received, how many were rejected and why, how many completed, the outcomes, the average duration, and stakeholder satisfaction. Keep disaggregated records of patterns and causes, and evaluate whether the remedies actually worked.

Disaggregate every one of those by sex and the outcome measures appear by themselves. Grievances raised by women as a share of women's share of your supply base. Median time to resolution, women against men. Remedy acceptance rate. Repeat grievances from the same site, which is the clearest signal that a remedy did not hold.

A Measurement Table

The right-hand column has a common property. Every measure in it is a ratio or a median that can move in either direction, and each one is computable from records you either already hold or could hold without a survey. Only two of the eight require asking a woman a question directly, and those two are the ones worth paying for.

Three Cautions

Disaggregation is not analysis. Splitting a figure by sex and printing both numbers is the first step. The analysis is the comparison, the trend and the explanation of why the gap is the size it is, and a report that stops at the split has done the easy half.

Do not infer empowerment from a proxy you have not tested. Account ownership is not control of income. Attendance is not adoption. Board membership is not influence. Each of those links is an assumption, and in this field the assumptions fail often enough that treating them as established is how programmes come to believe things that are not happening.

Sample from the register, not from the meeting. If you verify your numbers by interviewing women who turn up to a scheduled session, you have sampled the women who can leave the house, which is a filter correlated with almost everything you are trying to measure. Draw the sample from your supplier list and go to them.

Where To Start

If none of this exists yet, the sequence that produces the fastest honest answer is short.

Pull your supplier register and compute, once, the share of registrations held by women. That single number is usually enough to settle an internal argument about whether there is a problem, and it costs an afternoon.

Then look at payments and establish whose account the money lands in, by value rather than by count, because the two diverge sharply when larger consignments come from registered men.

Then disaggregate the grievance log by sex, including the rejected cases, and compare the result against women's share of the supply base. If women are 45 per cent of your suppliers and 4 per cent of your grievances, you have learned something about your channel rather than about their experience.

Nothing in that sequence requires a new framework. It requires counting what you already have in a way that can embarrass you, which is the practical difference between measuring participation and reporting it.

General guidance only. Statistics cited reflect the sources named and the years indicated, and figures on women's share of agricultural labour and land ownership vary considerably by source, definition and measurement method; the agrifood employment shares and agricultural labour force shares quoted use different denominators and are not interchangeable. Findex figures describe low- and middle-income economies in aggregate and should not be applied to a specific sourcing geography without local data. Pro-WEAI is a survey instrument requiring trained enumeration and ethical review; the household comparison described here should not be attempted informally. Collecting data on women's income control, household decision-making or grievances can expose respondents to risk, and any such data collection needs a do-no-harm assessment, informed consent and confidentiality protections appropriate to the context. Confirm legal obligations on supply chain due diligence and reporting with your own advisers.

Sources

United Nations, UN Women, Food and Agriculture Organization of the United Nations, Sustainable Development Goal indicator 5.a.1, World Bank, Global Findex Database, International Food Policy Research Institute, CGIAR GENDER Impact Platform, International Labour Organization

This article is intended for general professional information and does not constitute legal, financial, or investment advice.