The great AMI debate: Are the critics clueless?

Some think a new income formula would make affordable housing cheaper

City skyline encircled by a measuring tape
(Illustration by Ilya Hourie for The Real Deal with Getty)

Numbers can be stubborn. Converting the speed limit to kilometers per hour doesn’t get you out of a speeding ticket. In football, close losses don’t get a team into the playoffs. As Bill Parcells famously put it, “You are your record.”

In housing development, numbers can’t be coaxed, nudged or bullied, and they have a way of building invisible cages around a project. They cut a building’s height, limit its density and kneecap its profitability.

But of all the numbers developers deal with, area median income has become the most misunderstood.

Municipalities generally decide which applicants qualify for affordable housing by comparing their pay to the AMI — the typical household’s earnings in that housing market. It’s a measurement, just like miles for distance and pounds for weight. But critics argue that New York’s AMI is the wrong yardstick for affordability.

Here’s the problem: The rents of affordable units built in New York City are linked to the AMI not of the neighborhood but of the entire city plus the northern suburban counties of Westchester, Putnam and Rockland.

These housing markets are in some ways connected, but Westchester has a lot more cul-de-sacs than bodegas, and no one would mistake Mott Haven for the Upper East Side. Some 83 percent of Putnam families own their homes; just 38 percent of New York City households do. Incomes in adjacent neighborhoods, let alone counties, can vary drastically.

The dissonance has brought the same chorus to rezoning fights across the city: AMI might be the median income for this huge sweep of land, but it’s not accurate for this neighborhood. The critics’ fix is simple: Use a hyperlocal AMI to set the rents, so they will be lower in poorer areas.

But here’s the speed trap: Construction costs would not also go down.

“There’s no magic here,” said Mark Willis, a senior policy fellow at NYU’s Furman Center. “You can’t just say, all right, ‘I want the units to be more affordable. Now, developer, go ahead and do it.’ We’ve got to look at what the costs are and what’s required to write down those costs.”

After five years of single-digit increases, New York’s AMI jumped 16 percent this year. Recent wage increases have been concentrated among higher earners , and inflation has particularly strapped low-income households, which spend higher percentages of their money on inflation-ravaged categories such as food and housing. The standard percentages of AMI used to determine income limits for affordable housing have never been less reflective of a working family’s ability to afford an apartment.

But it’s also become more expensive to build one.

“Whatever the AMI is, it doesn’t affect fundamentals of affordable housing development,” said James Lloyd, director of policy for the New York State Association for Affordable Housing.

“You’re still going to have to generate a certain amount of revenue from rent, and you’re going to have to make up the difference between what market-rate rent would be and what those subsidized renters are paying.”

Many developers will say they’re happy to set aside units for lower AMI bands, so long as enough revenue is added elsewhere for the project to pencil out. But subsidies are limited, and adding profit-generating market-rate units is a nonstarter for locals, who generally want that number reduced.

Instead, activists assume developers can simply cut their profit margins to deliver more low-rent units.

“If you wanted to make it all fully affordable, you could still make a profit,” community board member and City Council aide Doreen Muhammad said late last month to developers proposing a 2,845-unit Queens megaproject by Silverstein Properties and others.

Sign Up for the undefined Newsletter

This disconnect has already killed some developments in their cribs.

Council member Kristen Richardson Jordan scuttled a Bruce Teitelbaum project in her Harlem district, saying its proposed affordability “does not apply to the humans who actually live in our district.” A typical Harlem household, she said, earned barely 30 percent of AMI, and therefore apartments priced beyond their reach were not truly affordable.

At 5 World Trade Center, Silverstein Properties and Brookfield Properties plan 1,325 apartments, 25 percent of them affordable. A group backed by a number of local political figures says nothing short of 100 percent affordability will do.

Silverstein estimates that it would need a $670 million subsidy to make even half of the project’s apartments affordable to renters earning between 40 and 60 percent of the area median income — $2.3 million for each additional unit.

Because of the high cost of building a skyscraper — in the Financial District, no less — even housing renters earning 165 percent of AMI would require a $300 million subsidy. That’s three or four times the income limit that affordability advocates typically demand. And calls for 100 percent of units to be affordable are growing.

“When a developer says, ‘Okay, I’ll go to 50 percent affordable,’ what they’re saying is, ‘I’m committing to going to HPD and trying to make a deal somehow to get a subsidy,” Lloyd said, referring to the city’s primary housing agency.

Read more

An affordable rent is defined as no more than 30 percent of the tenant’s income; hence, the lower the AMI that is used, the lower the rent. That, in turn, reduces a project’s expected return and makes it harder for developers to raise and borrow money to build it. Virtually all projects depend heavily on investors and lenders.

While some developers can afford to cut profits a bit in the name of affordability, they still have to be high enough to justify the risk of high-stakes development, which can be disastrous. The more plausible approach is to fill the gaps on a project’s balance sheet with city subsidies.

“AMIs provide a framework, but all projects are negotiated in order to meet the needs of the communities and the broader city,” said William Fowler, a spokesperson for HPD.

The group focused on the World Trade Center site, 100% Affordable 5WTC, hasn’t decided yet which AMI bands would be acceptable. But it argues that area median income fails to capture the range of identities the state should consider when it decides who qualifies for low-rent apartments.

“The politics of AMI are very confusing. People tend to think more in terms of identity,” said Todd Fine, one of the coalition members. “A lazy, trust-fund kid might have no income. So should he qualify for an affordable apartment just because he doesn’t want to work that much?”

Trust-fund kids have on at least one occasion gamed the system to score subsidized housing, but that was an obscure, poorly designed program. Affordable units are typically awarded by lottery. And in any case, the challenge of changing the AMI to create ultra-low rents is getting the project done at all. Investors and lenders don’t fund money-losing projects.

The Furman Center argues that focusing on the minutiae of AMI distracts from more important discussions, like how to divvy up the state’s limited pool of subsidies.

Whatever approach the city might take to increasing affordability, it is clear that switching to a local AMI would not change projects’ other numbers — the ones they need to pencil out.

“There’s a lot of pressure put on all of these developments to be that Goldilocks, just ever so perfect so it’s politically possible and economically feasible,” Lloyd said. “It can be a vanishing point.”