Last August, I obtained from a colleague dozens of internal documents from the New York City affiliate of Habitat for Humanity. The story a source had in mind focused on the possibility that the nonprofit had become unwittingly entangled in a massive money-laundering operation. Former employees had already voiced similar suspicions to their superiors and Habitat had asked a friendly law firm to look into the matter. The lawyers had found nothing of the sort, though, according to a one-page report they prepared.

When ProPublica and the New York Daily News finally published my story eight months later, it didn’t focus on money laundering at all. Instead, the story detailed how the charity’s management of an ambitious housing initiative had led to the displacement of low-income families in Brooklyn, families who had been forced out of their homes just days or weeks before Habitat moved to buy their former apartment buildings, which the charity described as “long-vacant” structures. Worse, federal money that was meant to fight the foreclosure crisis helped fund the project that displaced these people.

The lesson I learned as I navigated the reporting process is that even if one avenue of reporting doesn’t pan out, another might. I looked into the money laundering story and concluded that, for a number of reasons, it didn’t check out. But the memos and emails that had made their way to me suggested that Habitat was involved in something troubling in a different way. One employee had complained to Habitat’s whistleblower tip line that the New York affiliate was spending “federal money to throw low-income New Yorkers out of buildings.”

By coincidence, at the same time as I dug through Habitat’s documents, I was also working on an investigation on the failures of New York City’s rental protections with a ProPublica colleague, Cezary Podkul. It was through this reporting on the intricacies of the rental market that I saw the whistleblower’s allegations through a different lens.

Habitat’s project seemed to have a design flaw. In an effort to spend funds efficiently, the charity had prioritized buying vacant multi-family buildings in a poor — but gentrifying — neighborhood of Brooklyn.

However, a project focused on finding vacant multi-family rental buildings is no simple endeavor in this city.

Most of these buildings are covered by the state’s strongest rental protections — called the rent-stabilization law — which afford tenants the right to renew their leases in perpetuity, no matter how many new owners come in or how many mortgages they default on. Checking the buildings’ tax records showed that these protections covered all but one of the buildings acquired by Habitat. And the neighborhood Habitat picked was filled with stabilized buildings. In other words, Habitat had been looking for vacant buildings in an area where buildings rarely, if ever, go vacant. Supporting this, the charity’s struggles to find vacant buildings were well documented in the records I had reviewed.

As a result, it seemed that a project that ultimately targeted rent-stabilized multi-family buildings was a recipe for problems. Housing experts told me as much. “You don’t find multi-family buildings 100 percent vacant unless somebody has done something really, really, really bad,” one expert told me. Another one concurred. “Empty multi-family buildings were tough, if not impossible, to come by anywhere in New York City, even at the depths of the recession,” he said.

With all this background information, I really just needed to figure out two questions to prove or disprove the story idea.

How do you find out how long the buildings were occupied?

How do you find out when Habitat for Humanity first expressed interest in purchasing any particular building?

I first went to Brooklyn housing court. If any of the building owners who sold to Habitat had been trying to get rid of their tenants, then they would most likely have filed eviction cases. The cases I found gave me a list of former tenants to begin tracking down.

It’s important to see not just when a case began but also how long it dragged out. Throughout the length of a case, it was reasonable to think that a tenant would still be living in the apartment. My first findings from housing court showed that a couple of buildings had had tenants in them just months before Habitat closed on them. This was a good start. However, in any real estate transaction, the closing is only the culmination of a negotiation process whose length is impossible to predict simply from the closing date. Finding when those negotiations had started was trickier.

I then filed public records requests with the city, asking if tenants at any of the buildings Habitat ultimately purchased had ever complained about issues such as lack of heat, water, rats, noise and any other maintenance complaints. More than the issues themselves, I was concerned about the dates, since this would prove somebody was living in the building. The records turned out to be extremely helpful, and, while they mentioned no names, they mentioned specific apartments within the buildings. At this point, using Nexis and other public records databases, I could figure out who used to live in each particular unit.

Not knowing exactly what I expected to find, I also filed Freedom of Information Act requests with the federal Department of Housing and Urban Development. I was just hopeful that, buried in the paperwork, a Habitat employee or a federal official would have noted the exact dates when negotiations began on specific buildings. I didn’t expect, however, that HUD was aware that the buildings had been recently occupied. A senior official at HUD had told me early during my reporting process that he wasn’t aware of any irregularities in Habitat’s paperwork.

While filing FOIAs can be a dead end with many federal agencies, HUD promises to respond within a few months. I filed two separate requests. One was for the “entire file” involving Habitat for Humanity’s New York City project. The other was for any communications between Habitat-NYC and HUD. The logic behind filing this way is that dividing your request into smaller units can help you avoid “complex tracks” — FOIA-lingo for being stuck in the slow waiting line.

Indeed, HUD emailed me a couple months later saying they had identified close to 1,000 pages of relevant records to fulfill both of my requests. However, there was a surprise: HUD had decided to show Habitat the records it intended to disclose to give the charity a chance to challenge their disclosure. The move was frustrating — although it was a judgment call on the part of the agency that was ultimately grounded in FOIA law. But HUD’s decision helped drag out my request for an additional four months. To my surprise, I got all the records in the end, either because Habitat did not oppose the release of the records or because HUD overrode those concerns.

Records-wise, I based most of my reporting on court, city and HUD reports, which I then compared with Habitat’s internal documents. From there it was straightforward enough to create a timeline. HUD’s records proved to have what I most needed: signed documents attesting to the date Habitat had “first visited” the properties it sought to acquire. In one case, HUD records showed Habitat first visited a property 15 days after the last recorded tenant filed a complaint about the building with the city. In another case, Habitat was already negotiating a purchase 20 days after the seller had sued the last remaining tenant, trying to evict her.

As I tracked down tenants using Nexis or through their relatives, I had really just one question for them: When did you move out? Talking to them almost five years after their move, I was concerned they would not remember or provide me with contradictory information. I made sure during the interviews not to offer the dates I had independently found for fear of influencing their answers. In the end, all their answers matched the documents I had obtained.

In response to the story, Habitat said that while it was “unequivocally unaware of any coercive tactics” used to push out the tenants, it stood ready to help any of the “seven displaced families” find affordable housing.