I came to Marigo in March to see whether a teacher could learn to manage a bank branch. It wasn’t as though I was looking to change my stripes. I wanted to manage it as a teacher.

In my work as a classroom teacher I have tried to sustain a commitment to student-centered learning. I see my primary responsibility as helping students see the situations they find themselves in as questions they can work together to answer or problems they can work together to solve. The heart of the matter in Marigo is whether I can learn to be part of a bank branch’s staff that succeeds by looking at its situation in that same way.

But it’s hard to face problems when we’re buried in them. When we’re feeling overwhelmed, it can be hard even to know where to start. And the Marigo office was really in bad shape.

The heart of the problem was the result of two years of fast, careless growth. The Marigo branch opened in 2006, and from the outset it easily and quickly recruited new borrowers. But even at its period of high growth, member turnover was high. Borrowers would drop out of Fonkoze’s lending programs only to be replaced by new ones. As many borrowers who were lost would be more than replaced by new ones, however, and by early 2008, the office was serving between 1800 and 2000 borrowers.

The quality of the Marigo loan portfolio was deteriorating, however, so Fonkoze’s central office made an important decision: Marigo would have to stop recruiting new borrowers until it could learn to serve those already in the fold. Learning to serve them better would involve a lot of hard, disciplined work in the field, work that only grew more difficult in the face of an economic crisis brought on by the world-wide spike in food and gas prices and then more difficult still when three hurricanes and a tropical storm ravaged the Marigo region, and much of Haiti, last August and September.

Since putting an end to recruiting did not affect the dropout rate, the total number of borrowers started to decline. By the time I joined the office in March, there were about 1200.

What’s worse: It’s not as though they dropped out in a convenient, well-organized way. For its 2000 borrowers, the office had five credit agents serving them in 56 credit centers. When we were down to 1200 borrowers, we still had five agents and 56 centers. Our cost per loan was, necessarily, spiraling upward.

So the first charge I had from the central office was to lay off two credit agents. An ugly enough task in a country where unemployment is so high that employment in the formal sector can scarcely be said to exist. What made it worse was that we would need to plan how to serve borrowers from the same 56 centers with the three agents who remained.

I decided that we simply couldn’t. So we closed some of the very worst centers, ones we felt we could not save.

Most of these had become unsalvageable for one or another version of the same reason: A self-appointed local leader, almost always a man, had managed to interfere sufficiently with Fonkoze’s close relationship with borrowers that we no longer could make the borrowers feel genuinely responsible for their loans. For example, in Kajak a local leader had, among other things, taken it upon himself to collect reimbursement payments, but hadn’t been turning them over to Fonkoze. The women believe they’ve repaid, but the money never has gotten to us. And there’s no paper trail to pursue. In Nan Zèb, most of the loans went to ghost borrowers, women who did not have businesses at all, who signed for the money and then handed it over to the local leaders who put them up to it. Ironically, those leaders then used the money as loan capital that they then let out at truly exorbitant rates. They are loan sharks, exactly the thing Fonkoze is most anxious to combat.

We continue to pursue the money that is in those men’s hands, but our options are limited. And even if we succeed in collecting it, we have to admit that we cannot be confident of developing direct relations with borrowers that these men can’t undermine. So we just have to give up on the neighborhoods for now.

We also combined some centers that were close to each other and not too large. We thus managed to get the number of credit centers down to 42.

These changes involved a lot of chaos. On one hand, we had to change some center meeting days and times, which confused both our office and our borrowers. On the other, we had to make significant changes in the assignment of centers to credit agents. This interfered with the familiarity between the agents and their borrowers that credit depends on.

So because of the chaos, we had weeks of not really knowing where we stood. We were regularly seeing small signs of progress or of trouble, without feeling very clear about the overall state of things. Under the circumstances, it was been hard for me to do the kind of teaching I like to do. The questions and problems haven’t been presenting themselves clearly enough for the staff to face.

In this context, Wednesday morning’s staff meeting was a real step forward. The credit agents gave a short report about each of their centers: the centers in pretty good shape, the ones that are improving, the ones that are still weak, and the ones for which we have no real hope. Each of the three agents seems to see the situation he is facing in his centers with increasing clarity, and though there are a couple more centers that they think that we will lose, and a couple that we just don’t know about, the core of well-functioning centers seems to be growing both larger and stronger.

The clearest indicator of that improving core, in fact the key to it, is the increasing cooperation we are getting from our center chiefs, the women whom borrowers elect to lead them.

There is the center in Lilankou, for example, which was created when a small center already in existence there was joined by members of neighboring centers in Nan Aka and Savann Dibwa. The combined center now has very good attendance, though better among members from the non-Lilankou groups than from those from closest by.

And one has the sense when one visits the center that its chief is really in charge. I watched our credit agent, Jean Bellande, make some loans the last time I was there, and was encouraged to see how he went through the process of counting out the money and handing it over to the center chief, who then handed it to the borrowers. The chief was careful to say that Fonkoze had lent her the money and that she, in turn, was lending it to the borrowers who signed for the loan. It was the best evidence I had seen thus far that at least one center chief was ready to take the full measure of responsibility that we ask chiefs to take.

On Saturday, I was in Lagad, the home of one of our oldest centers, in the mountains beneath Segen, with another agent, Bob. He wouldn’t normally go on Saturday, but the center’s chief had told him that it would be a day of neighborhood celebrations. The locals who live in Port au Prince would be home for the weekend, and so it would be a good day for the two of them to walk door to door, talking with delinquent borrowers who aren’t usually around. And it was exciting to see that she led the way to most of the houses and that she did most of the talking as well. Her center is a little behind the one in Lilankou in terms of financial and attendance figures, but the sense of responsibility she demonstrated gives very good reason for hope.

A final example: Tuesday I went with Bob to Koray Lamòt, the most distant credit center we serve by far. It’s in the mountains above Belans, almost three hours away by motorcycle. We never should have opened the center, but we did. And we’re a little loathe to give up on it.

But it remains weak, and one of the main issues is attendance. Tuesday, its chief offered a clear explanation: Only very few of the poorest members of the centers live and have their businesses in Koray Lamòt. Most buy merchandise in the area every couple of weeks, and then take it by truck to Port au Prince. That is where they live and do most of their business. They like coming home to see their families and to buy the coffee, charcoal, and agricultural products they principally sell.

So we suggested a compromise. We would adopt the pattern we are already using for women who live in Marigo but work in Ansapit. They would need to come only once per month. Tuesday afternoon would be their center meeting and Wednesday would be their reimbursement. The credit agent would stay overnight, probably as the center chief’s guest. We can’t be certain that this plan will work, but we might as well be optimistic.

As we get a grip here in Marigo, the look of things really changes. We begin to see a finite number of problems, some big and some small, where once we saw nothing but one giant, indeterminate mess. That feels like progress.