Repayment, Reconsidered

When I arrived I thought I knew what I was looking for. I had read about the Grameen model — small lending groups, regular repayments, the discipline that comes from both. The programme reported a repayment rate near ninety-nine per cent, and on paper the number seemed to settle the question of whether the model worked.

My days were spent interviewing borrowers and staff, taking field notes, and photographing the places where lending had become part of daily life. I met women using small loans to support businesses and households. I also saw the trust that group lending can create: members knew one another’s schedules, customers and family pressures in ways a bank officer rarely could. The model’s strengths were visible, and they were real.

The repayment rate became more interesting when I asked what happened during a difficult month. Several conversations pointed to work the headline number could not record. A payment might arrive on time because relatives supplied cash, because another group member covered a shortfall, or because a borrower postponed a different household expense. The loan would still appear as repaid. The effort required to produce that result would sit outside the spreadsheet.

The programme’s favourable terms mattered as well. Below-market interest rates and long repayment periods gave borrowers room that commercial products often do not. Those terms helped explain the strong performance — and they raised a question I could not yet answer: could the same design hold if a commercial lender had to price it?

I tried to answer that question too quickly. After the field visit I presented a plan for adapting the model to professionals from four banks. They asked about interest rates, repayment periods and operating costs. Their questions exposed the weakness in the proposal, and the criticism changed the direction of the work.

I went back to my notes and began separating two questions I had been running together: whether borrowers repay, and what repayment asks of them. The working paper that followed — Designed Visibility — looks at how selection, reporting conventions and the design of an evaluation shape what a programme’s numbers are able to observe in the first place, and argues that these are conditions under which difficulty becomes hard to see at all. It also proposes a practical fix: track early signs of household strain rather than waiting for a missed payment, which by then is a lagging indicator of something that began months earlier.

Development programmes are often assessed through figures that have to travel easily — between villages, banks, ministries and international organisations. Portability is what makes a figure useful, and compression is what makes it portable. Fieldwork asks us to stay with the people inside the data long enough to see what the summary has folded away.

Two weeks did not leave me with a verdict on microfinance. They left me with a better standard for asking questions. A repayment rate tells us whether money returned on schedule. It does not tell us what a household rearranged to make that possible, or whether the conditions behind one successful pilot can be reproduced somewhere else. Careful research begins where a convincing number no longer ends the conversation.

Zhiyuan (Archy) ZhangViews expressed in this field note are the researcher’s own.