The average family living in the Portland metro area brings home $139,100 each year, according to the federal government.

In Lewiston/Auburn, it’s $93,900, and in Kennebec County, $96,000.

The number isn’t merely symbolic. It’s what’s known as the area median income and it serves as the base for nearly every affordable housing program in the country, helping to determine eligibility for assistance and set rent.

Housing experts also say it’s an imperfect measure.

There are ongoing arguments about the accuracy and fairness of using AMI, which lumps together the incomes of renters and homeowners across what can be a wide geographic area with disparate housing markets.

Yet without a viable alternative to help guide thousands of state-level housing agencies serving millions of people, the same experts agree it’s the best tool they have.

“It’s central to everything we do,” said Scott Thistle, spokesperson for Maine State Housing Authority.

Here’s a closer look at how AMI is set, how it’s used and whether there’s room for improvement.

HOW IS AMI CALCULATED?

Area median income is set by the U.S. Department of Housing and Urban Development using data from the U.S. Census Bureau’s most recent American Community Survey (currently 2024) and adjusting for inflation.

The country is divided into metro areas (largely determined by the Office of Management and the Budget based on economic and commuting patterns) or counties.

An area’s general AMI is based on a family of four, but in practice, agencies take family size into account. For example, while the Portland area’s AMI is $139,100, that drops to $97,370 for a single person. In Lewiston/Auburn, AMI for a single person is $65,730, and in Kennebec County it’s $67,200.

Maine has 19 designated areas, including four metro areas: Portland, Lewiston-Auburn, Bangor and York/Kittery/South Berwick. The remaining 15 are the counties (minus the towns in the metro areas) with the exception of Androscoggin County. All the Androscoggin County towns are included in the Lewiston-Auburn metro area.

Portland’s metro area includes: Portland, Cape Elizabeth, Casco, Chebeague Island, Cumberland, Falmouth, Freeport, Frye Island, Gorham, Gray, Long Island, North Yarmouth, Raymond, Scarborough, South Portland, Standish, Westbrook, Windham, Yarmouth, Buxton, Hollis, Limington and Old Orchard Beach.

WHY IS IT IMPORTANT?

AMI isn’t simply a handy number to have. It “plays a foundational role in every affordable housing program in the whole country,” said Thistle, at MaineHousing.

Established in 1937, it has been widely used to set income limits (eligibility) for affordable housing programs at the state and federal level.

Families earning 80%, 50% or 30% of AMI are considered low-income, very low-income and extremely-low income, respectively. And different housing programs and projects target different income levels.

Low-income families can qualify for the federal Section 8 Housing Choice voucher program, but the majority (75%) of new admissions to the program need to be extremely low-income, said John Concannon, executive director of Westbrook Housing and former director of the U.S. Department of Housing and Urban Development.

Meanwhile, the state’s Affordable Home Ownership Program, which builds homes for more “middle-income” families making at or below 120% AMI. In the Portland metro area, that would be about $116,850 for a single person or $166,900 for a family of four.

Those income limits are also used to set rents for apartments in many different housing programs, including the Low Income Housing Tax Credit program, the largest affordable housing program in the country.

In that program, rents are determined by the number of bedrooms and a percentage (around 30%) of the income limit, rather than the tenant’s actual income. Rents can change as AMI changes, but will not change in tandem with a tenant’s income.

By dictating rents, AMI is also a critical number for affordable housing developers, said Jennifer Hawkins, executive director of Avesta Housing.

“AMI drives the viability of projects,” she said.

In rural communities with lower AMIs, for example, lower rents mean lower revenue to cover operating costs for the affordable housing nonprofit.

That can also point developers toward higher AMI areas.

“It makes your project need more subsidy,” Hawkins said. “When there’s a scarcity of subsidy that becomes really, really hard.”

IS IT ACCURATE?

It’s hard to say definitively how accurate AMI is, said Dan Emmanuel, director of research for the National Low Income Housing Coalition.

“There are real limitations to the American Community Survey,” he said.

The “metro area” can be a challenging metric because many towns have vastly different economic conditions and varying incomes within them. The lump sum AMI might be too high or too low for some towns, and that can have real consequences.

But the American Community Survey already has a relatively small sample size, so drilling down town-by-town reduces that even more.

“There is a balancing act there, between having the most granular, most accurate and true estimate for a given community versus that estimate being statistically reliable and actually meaningful,” Emmanuel said.

Even if town-by-town data were reliable, it would create a mountain of work for already overworked and underfunded state housing agencies.

“It’s easy to criticize AMI as a benchmark, but if you tried to bring it down to the microscale (by neighborhood) it would be impossible to administer,” said Thistle at MaineHousing.

IS IT FAIR?

It can be hard to reconcile an area median income of almost $100,000 for a single person in Greater Portland, when a minimum-wage earner working full time is making less than $35,000 before taxes.

But some of that is due to a mix of incomes from such a wide swath of towns.

Old Orchard Beach, for example, has a median family income of about $60,000, while Cumberland’s is $170,000. Both are in the Portland metro area, though, which means the $139,100 median is the benchmark for both communities.

Thistle said that disparity “can definitely create a disconnect with what’s really happening on the ground.”

But the misalignment likely also stems from the fact that the median includes the incomes of both renters and owners. In the Portland metro area, the median household income for homeowners was $111,610 in 2024. For renters it was just $59,697, according to Census data.

“In my mind, it’s not so much the reliability of the estimates … it’s how they’re being used,” Emmanuel said.

AMI is used to determine eligibility requirements for a host of programs that are overwhelmingly geared toward renters. At first glance, a higher AMI could then be a boon — more people could qualify for help.

But if rents are also based on a percentage of an AMI that is artificially high, it could push housing that is supposed to be affordable for lower-income families out of reach.

For example, for a low-income family of four making $106,800 a year (80% AMI), the maximum rent they could be charged for a three-bedroom apartment in Portland is $2,403.

Even if affordable housing agencies feel rents are too high and want to change the formula — by charging 60% AMI rents but accepting people making up to 80% AMI, for example — they can’t, said Hawkins, at Avesta.

“We’re beholden to this, even though we recognize it’s flawed,” she said.

ARE THERE OTHER OPTIONS?

Any changes to how AMI is calculated or used would need to come from Congress.

Emmanuel suggested that income limit thresholds for rental housing programs could be based on median renter family incomes, rather than just blanket median incomes.

Rents, however, shouldn’t be set based on AMI at all, he said.

“Generally speaking, it’s better to base your rents in a program on actual rents in a community,” he said. “They should be tied to the market and not tied to people’s incomes.”

To change AMI as a whole, Emmanuel said that beyond creating an entirely new formula, the only real way would be to increase funding for the American Community Survey.

“The degree to which the ACS can provide more granular estimates is entirely a function of how many households the Census Bureau can sample,” he said.

“But I think we’re stuck with them for the moment because they’re just so entwined (in affordable housing),” he said.