One of the formidable challenges at the center of our extreme poverty program is ensuring we select the right families. We have a process that always involves three steps: community input, initial screening, and final verification. But sometimes there’s also a fourth step. Fonkoze’s Social Impact Management Unit – which is our research and evaluation team – takes 25% of the families that the CLM staff selects for the program and re-evaluates them. They then follow the families through the program so that we have a way to independently confirm the work that our case managers do.

Earlier in the month, we got an initial report on the 360 new members we are currently working with. The data it included about the first 64 members whom they will follow confirmed a lot about our selection. But there was one problem: One member, Elsie Fénélon, appeared to be completely unqualified for the program. Social Impact’s survey found that she owed three goats, a horse, and a cow before she even joined the program. While that would hardly qualify Elsie as wealthy, it would disqualify her for CLM, especially since we knew she was managing a small commerce and had been able to send her kids to school last year.

I asked her case manager to follow the report up. He’s an especially smart and experienced man named Christian, and I’ve learned to value his judgment greatly. I wanted him to find out about the livestock and to let me know his overall opinion of the household.

This is what he reported: There were in fact three goats, a mother and her two bucks. They had not yet been weaned. She had been given the mother by Mercy Corps, an NGO that had been working in the Mirebalais area at the time. She also had purchased a horse, really just a colt. A younger brother had offered her one for 2000 gourds because he wanted to help her out. She didn’t have that much, but she had 1500 in a savings club that she was in with other market women. So she paid that money down, and her brother agreed to wait for the rest. The cow belongs to her husband. Right around the time she was selected for the program, he finished a job for a neighbor who planned to pay him by selling a young bull. When the neighbor couldn’t get the price he was seeking in the local market, the husband agreed to take the bull in lieu of payment. Unfortunately, however, Elsie is not the husband’s only partner, and he does not contribute dependably to the household.

So she has more in assets than a CLM member typically would at the start, a lot more. And with her existing small commerce, she might seem like an excellent example of an extremely poor woman who could be served by our Ti Kredi program, rather than by CLM. Her business – she sells powered laundry detergent by the cup full – is 100% dependent on credit. She has none of her own money invested in it all at. She buys a sack of detergent at a time with no money down, and has to sell it off to pay for it before she can get another.

Our team occasionally refers to “original” CLM members. These are the ones who are the worst off, the ones who have nothing at all when we find them: No children in school, no assets of any kind, and no economic activity beyond day labor and begging. We find a lot of families like that, and Elsie was not one of them. We knew that her business was working for her – it had enabled her to pay into the 100-gourd-per-week savings club that made it possible for her to buy the horse – and, even if we discounted the cow as an asset she couldn’t depend on, the goats would give her something to fall back on if her commerce took a bad turn.

But not all of our members are “original.” Some are slightly better off, even though they are still extremely poor. Christian still wanted to keep Elsie in the program. He felt that she was really struggling to feed her family, and that her lack of decent housing in a neighborhood where housing is relatively expensive would put her at great risk.

I needed to take his view seriously, very seriously. I can’t pretend to judge more astutely than he. But I also know that he tends toward including marginal cases. It is a predictable consequence of seeing as many miserably poor people as we do: You want to help them all, and the weight of their poverty tends to play with your sense of who it is who is poor enough. Deciding, even correctly, to exclude a hungry household from the program because it is not quite poor enough becomes harder and harder as you go along.

So I went to talk to her myself. I wanted to figure out if I could why we had missed the assets when we first interviewed her. Had she simply lied? And needed to decide, more importantly, whether she belongs in CLM.

I went by to see her early in the morning, hoping to catch her before she went out. By the time I got to Labasti, where she lives, she was already out in the street, trying to sell the last couple of scoops of detergent so she’d be able to pay for it and get a new sack the next day. But she hadn’t gone far, and when word got to her that I was there, she came running home. It took her a few minutes to catch her breath before we could start to chat.

I already knew that her small commerce was on the edge. I had seen her the previous day in the small market in Trianon, trying to sell the same last bit of detergent she was walking with. The Trianon market is failing. Very little is sold there. But a few desperate merchants sit around, hoping that something will happen. If her sales were thriving, she wouldn’t have wasted a day there. So things weren’t going well. The day I saw her she was on her way to Terre Rouge, a long uphill hike from her home. She thought she’d be able to sell out her product if she went to the various fried-foods merchants who set up there for passing traffic. They don’t have time to do much shopping, so are often glad when the things they need come to them. It would be a long day, with low potential for return, so she must really feel the need for a few gourds of additional sales.

We talked about her goats. She told me that she had received one as a gift from Mercy Corps, but that she had offered it to her brother when she could not figure out how to pay her kids’ tuition bill when the school year was coming to an end. The brother offered her the 2000 gourds she needed, but then said he’d rather she kept the mother goat. He would take the goat’s first two young, instead. When the goat had a litter of two, he turned the kids over to Elsie’s oldest boy to take care of. He has now just sent for them from Port au Prince. It is also true that she has a horse, though it is not all hers and it’s much too small still to bring her anything but the extra work of taking care of it.

But as I talked to her, and looked at both her and her young children, I just thought “CLM.” Their house is a wreck, and the boys were scavenging something to eat in their grandmother’s corn. The ears are dry, ready for harvest, but the boys grill the whole kernels and then count on their good teeth to snack on them. Anything like a real meal would have to wait until the afternoon, probably late because the trip to Terre Rouge and back would take Elsie all day.

I could be wrong. Maybe Elsie would be a good bet even with Ti Kredi. I know that I’m subject to the same pressures that I see acting on case managers like Christian. But Elsie’s household is struggling by any measure, I know that CLM can help, and I’m not certain that Ti Kredi is really an option.

So I decided to make what was certainly the easy decision, the one to keep her in the program. I’m pretty certain that she will succeed. A lot of the entrepreneurship we seek to teach seems already there.