Thirteen
per cent

A programme can train a woman, equip her, and generate income for her. If she has nowhere to put the money, most of what was built leaks away within a year.

The number, and the gap inside it

World Bank Global Findex data puts account ownership in Pakistan at roughly 13% of women against 34% of men — a gap of about 21 percentage points, and one of the wider ones recorded.12

The comparison that makes it land is not with wealthy economies. It is with Pakistan's own peer group. Across low- and middle-income economies, about 73% of women held a financial account as of 2024, up seven percentage points since 2021.3 Roughly 700 million women worldwide still have no account, and a disproportionate share of them are in a small number of countries.

Figure 1

Account ownership: Pakistan against the low- and middle-income average

Women, low- & middle-income average73%

Men, Pakistan34%

Women, Pakistan13%

Why the usual explanations are incomplete

The standard account says the barriers are income, education and employment: women have less of all three, so they have fewer accounts. That is true and it is not sufficient.

Research on the determinants of financial inclusion in Pakistan finds that education, income and employment all support inclusion — but that their effect is weaker for women than for men, which points to barriers operating beneath the socio-economic variables rather than through them.4

That finding matters enormously for programme design, because it means the intuitive intervention does not work as advertised. Raise a woman's income and her likelihood of holding an account rises by less than the same income increase would deliver for a man. Documentation requirements, mobility constraints, branch environments, and who in a household is permitted to hold an asset are doing work that an income variable does not capture.

If income alone closed the gap, the gap would already be closing faster than it is.

What this does to a livelihoods programme

Consider the standard design. Train a group of women in a skill. Provide equipment or inputs. Connect them to a market. Measure income generated. Report success.

Now add the constraint. Income arrives as cash. Cash held at home is visible, and a visible asset in a household where the woman is not the financial decision-maker is frequently not hers in practice. It cannot be saved securely, cannot build a credit history, cannot be used as collateral, and cannot survive a shock — which, in flood-exposed districts, is not a hypothetical.

The programme reports income generated. Twelve months later the capital has dispersed and the participant is roughly where she started. Nobody lied. The indicator was simply measuring the wrong moment.

The same error as reach, in a different costume

There is a corresponding error on the financial inclusion side, and it is worth naming because it is committed by people trying to solve exactly this problem.

Accounts opened is not financial inclusion. It is the account-ownership version of reporting reach instead of outcome. An account opened during a registration drive and never used again is a number in a report and nothing in a life. Global Findex work has been consistent on this: ownership has risen faster than equal access and use.3

The indicator that means something is active use at a defined interval — an account with transactions in it six or twelve months after opening, controlled by the woman whose name is on it.

How we design around it

We are not a financial institution and we do not open accounts. What we can do is stop designing programmes that assume the financial system is available to participants when it is not.

• Treat account access as a baseline variable, not an outcome. We record it at intake for every livelihoods-adjacent programme, because it determines what the programme can realistically claim.
• Report retained income, not income generated. Generated income measures the intervention. Retained income at a fixed interval measures whether it mattered — and the gap between the two is where the financial-access constraint shows up as a number.
• Partner rather than improvise. Where account access is the binding constraint, the answer is a financial institution or a mobile money provider with the licence and the branch network, not an NGO inventing a savings scheme.
• Publish the constraint when it caps the result. If a programme's outcome was limited by participants having nowhere to bank, that belongs in the programme review. It is the single most useful thing the next designer can read.

The wider point

The unmet credit need of micro, small and medium enterprises in developing economies runs to trillions of dollars.5 Numbers at that scale invite the conclusion that this is a problem for banks and regulators, and largely it is.

But the part that belongs to programme designers is small, specific and entirely within reach: knowing whether the people in your programme can hold money safely, measuring it, and being honest in your reporting about what it caps. That costs one question at intake and one follow-up at endline.