A Better Conversation about Area Median Income (AMI)

Research & Policy | September 10th 2018

NYU Furman Center Logo

There is a widespread misconception that the methodology used to calculate AMI prevents the poorest households from accessing affordable housing. That's an intuitive response to affordable housing lotteries where the eligibility is too high for many neighborhood residents but these outcomes are the result of policy choices at the state and local level, not the methodology used to calculate AMI.

When people question how AMI could be so high relative to what seems like the obvious reality of a borough or neighborhood, they often blame the geography used by HUD's methodology. There are two problems with this argument. First, local policymakers are free to lower the income levels subsidized housing will serve. AMI is just a benchmark policymakers use to define eligibility for subsidized housing. State and local policymakers can (and do) choose to serve lower-income households. This decision can be a function of many factors, including the costs. They can for example, target households at 30% or 50% of AMI instead of the 60% maximum attached to federal tax credits. The tradeoff that comes with such choices is a greater subsidy needed per unit, meaning potentially fewer units built if overall resources devoted to subsidized housing are fixed. 

The second misconception is because HUD uses Westchester, Rockland, and Putnam counties in its calculation of NYC's AMI, many assume that these counties' more affluent areas are pulling affordable housing beyond the reach of the neediest households in the five boroughs. But removing Westchester, Putnam, and Rockland Counties from HUD's AMI calculation would not significantly change the metro-wide result. 

HUD's methodology includes an adjustment for "High Housing Cost Areas." Unsurprisingly, these days NYC definitely qualifies, regardless of whether Westchester, Rockland, or Putnam Counties are included. Once an area crosses the threshold into one HUD considers "high-cost," the final AMI number is tied to Census data on rent levels, NOT incomes. 

The difference between median rents in NYC and its surrounding counties is much smaller than the difference in incomes: only $100 separates the five boroughs ($1294) from Westchester ($1394) according to 2016 ACS data. Rockland's median is $1367. Putnam's is $1307. In fact, for the last two years, HUD’s Studio, 1BR, 2BR, and 3BR Fair Market Rents for the entire Metro-Area (NYC, Westchester, Rockland, and Putnam) were higher than those for Westchester County alone. Accounting for NYC's much larger population, the effect of these suburban county differences is likely extremely small or nonexistent on the final AMI number, though one can't calculate a precise impact without the custom data HUD uses from Census.

In summary, AMI does not preclude state and local policymakers from serving lower-income households, which would require more subsidy per unit built; and removing suburban counties wouldn't significantly change HUD's AMI result for NYC. Continued focus on the technical details of AMI distracts from more important conversations about the level of public resources devoted to housing, and the tradeoffs involved in choosing how to allocate them.

« Previous | The Stoop | Next »