In light of the current controversy regarding the siting of a 200-bed homeless shelter on the Barron Center campus in the Nason’s Corner neighborhood of Portland, I decided to use my training in research to conduct a web-based study of homeless shelters in other New England cities comparable to Portland in population (most are somewhat higher).

The cities are Burlington, Vermont (population 42,561); Manchester, New Hampshire (111,099); Worcester, Massachusetts (187,104); Springfield, Massachusetts (157,247); Providence, Rhode Island (178,710), and Waterbury, Connecticut (113,692). Also included are the Maine cities of Lewiston, Augusta and Bangor. Listed below are my conclusions from examining the data, which have led me to support having several smaller shelters around the city rather than a consolidated facility in one neighborhood.

All of these cities, except for Portland, had their shelters run by nonprofit organizations or, in a few cases, commercial companies. None had city-run shelters. This suggests that these entities have the experience and efficiency in managing these shelters that city government lacks. This model is used by the city of Burlington, a liberal and progressive area. It must feel that this approach better serves the homeless.

Having different nonprofits running the shelters results in scattered sites throughout the city, although almost all were downtown, with exceptions being facilities for veterans battling drug addiction (in Worcester) or for persons with chronic mental illness (in Waterbury), where the sites were in isolated regions with little commercial development or private homes. Nonprofit sites in Bangor are a few blocks from downtown on Main Street or near the airport in a corporate commercial area.

None of the cities has a single shelter that houses 200 individuals, although two cities (Providence and Worcester) have 2.7 and 2.8 times more residents, respectively, than Portland does. The warehousing model is losing favor, and managing a large number of homeless in one location creates safety issues for staff, clients and the community, as recently reported by the NBC affiliate in Missoula, Montana, where residents of the city’s west side have seen an increase in crime since a homeless shelter relocated there.

The model Portland wants to build – a shelter and service center – most closely resembles the Friends of the Homeless facility in Springfield. However, this complex is in an industrial area, with the city police headquarters less than a block away.


None of the emergency homeless shelters for adult men and women in the cities studied were in suburban residential neighborhoods. These larger shelters, transitional housing and/or service centers are in a downtown industrial area because they need space and the scope of the complex is disruptive to a neighborhood with a large number of owner-occupied houses and nearby schools.

Harrington Hall, the largest men’s shelter in the state of Rhode Island (112 beds), is next to the state prison in Cranston, a city of about 81,000 that is 5 miles from Providence. Apparently, Cranston doesn’t think it stigmatizes the homeless to place the shelter in such a location.

Portland concluded in 2015 that a scattered sites model for emergency shelters would cost an estimated $4.6 million per year to operate. However, Amos House in Providence has a campus of 15 buildings, including a 90-day transitional housing program emphasizing addiction recovery; a community center; permanent supportive housing (six rooming houses); a dining hall (the largest soup kitchen in Rhode Island); three apartment buildings; culinary education and carpentry and maintenance training; a literacy program; a mother-child reunification program, and social services.

Amos House has a staff of over 50, yet its budget in 2017 was less than $4 million a year. These facts reinforce the conclusion from my first bullet point: that non-government entities can manage complex issues involved in helping the homeless in a more cost-effective manner.

I determined that Portland has 4.5 beds in homeless shelters per 1,000 city residents, much higher than in any other city in the study. It is also higher than Boston, which has 1.8 beds per 1,000 residents. The Portland data correlate with the fact that 74 percent of the clients housed in the city-run Oxford Street Shelter in 2017 were not city residents. There were more out-of-state clients than Portland clients. Therefore, Oxford Street, with its low entrance barrier, acts as a magnet to attract the homeless from other areas.

This model is unsustainable. Opportunities exist for a public-private partnership to pursue the scattered sites model, including transitional housing. These creative endeavors would better serve the homeless and ease the financial burden on the city of Portland.

Only subscribers are eligible to post comments. Please subscribe or login first for digital access. Here’s why.

Use the form below to reset your password. When you've submitted your account email, we will send an email with a reset code.