urban institute nonprofit social and economic policy research

A New Look at Homelessness in America

Document date: February 01, 2000
Released online: February 01, 2000

Despite a booming economy and the lowest unemployment rate in a generation, homelessness remains a problem in communities across America. The extent of that problem and the status of services for the homeless was the subject of our First Tuesday's forum in February. Institute researcher Martha Burt, a leading expert on homelessness in America, presented her analysis of the most recent national survey of programs and services for the homeless. Burt estimates that at least 2.3 million people experience homelessness at some point each year, including nearly 1 million children. She stressed that those numbers are far higher than estimates from a previous national survey taken a decade earlier, but pointed out that services for the homeless have dramatically increased during the same period. Nan Roman, of the National Alliance to End Homelessness, and Sue Marshall, of the Community Partnership for Prevention of Homelessness discussed Burt's findings at the meeting.

Full Transcript of the Session | Questions from the Audience


Martha Burt, Urban Institute
"The question of numbers, of how many people are homeless, is a very tricky one. It implies that there is a stable population of 'the homeless,' and that we can find one number that will represent them. We shall see that neither of these things is true. This is the first time since 1987 that we have had new national data from which we can try to answer the question of how many people are homeless. And the first time ever that we have had an inventory of the whole homeless assistance network." Burt describes how she and her co-researchers derived their estimates.

Nan Roman, National Alliance to End Homelessness
"If the numbers [of homeless] have gone up, and we've got 40,000 [service] programs that are assisting homeless people, I think that's scary. That's a very frightening thing for us to have found. And, in fact, we have to look at why that would be." Roman contends that mainstream government programs are failing to help people avoid becoming homeless.

Sue Marshall, Community Partnership for Prevention of Homelessness
"On any given day there are literally 7,500 to 8,000 people in the District of Columbia who are homeless. If you take those figures and annualize them, on an annual basis we serve 12,500 people. What that means is that if you use a base of 530,000, roughly 2.3 percent of the District's entire population experiences homelessness at some point during the year." Marshall talks about the development of programs to help DC's homeless population.

Martha BurtMartha Burt, a principal research associate at the Urban Institute, directed the first national study of the urban homeless in 1987. Her current estimates on the numbers of homeless people are derived from analysis of new data collected in 1996 for the National Survey of Homeless Assistance Providers and Clients (NSHAPC).

"We present some estimates that we think are fairly startling, that over the course of a year at least 2.3 million and probably as many as 3.5 million people experience homelessness at least for a short period.

"We had to meet a number of underlying challenges to develop these estimates. The first thing that you need to remember is that the data, and therefore the estimates, are from [homeless] service users. I am talking consistently throughout this presentation about the number of homeless people who used homeless assistance programs in October and February 1996, or over the course of a year. There are obviously homeless people who do not use homeless assistance programs. So these are lower-bound estimates of the people who are homeless at any given time.

"Secondly, we have a lot of different kinds of programs that we visited to find homeless people. And many people used more than one of those programs. So we had to develop ways to avoid counting the same person more than once if we were going to get to an accurate estimate of the number of people who are homeless. Some of the information we used came from providers in terms of how many people they expected to serve. And providers almost always overestimate when they are asked that question, not because they are trying to tell lies or anything, but because the way you remember things is from the extremes. So you tend to report your highest-use level, not the average. And for food servers, it's even more difficult because they usually count meals, and so they prepare 100 meals, and if 67 people come and some people have seconds, they still report the 100 meals.

"There's also seasonal variation, as well as some programs that were open in February that were not open in October. The client diversity is also important, and we had to develop standard ways to decide this person is homeless, this person is formerly homeless, this person has never been homeless. I'm only going to be talking about the currently homeless people today.

"And finally, and very important for the annual projections, is the point that the homeless population is not static. There are a lot of people who are homeless for a relatively short period of time, as well as the more common image of somebody who is homeless for years at a time.

"When you look at the U.S. population in poverty, if you use the [lower] October estimate, you can see that 6.3 percent would have been homeless over the course of a year. But if you use the [higher] February estimate, it's almost 1 in 10 poor people, and about 9 percent of poor children who will have experienced a spell of homelessness over the course of a year.

"Now, these seem phenomenally high. But the other two sources of information that give us an idea of annual prevalence rates produce results that are not wildly different from those from this study. One of them is based on shelter use from New York and Philadelphia, where, on average, over the course of a year for every year since 1989, 1 percent of the total population of New York City [and] 1 percent of the total population of Philadelphia have spent at least one night in a homeless shelter.

"And the other source of information is from a household survey that also came up with something very close to 1 percent of the whole population for an estimate or a projection of the number of people who have experienced a homeless spell at least for some period of time within the course of a year.

"One reason that so many more people are using homeless assistance services, and therefore we can find them to count, is that there are a lot more of those services to use, and, to switch the mode slightly from the people to the services, we can see that there has been a huge increase in the shelter and other homeless assistance program capacity or use between the late '80s and 1996, when these data were collected.

"In 1988, in the winter, HUD estimated that there were 275,000 shelter beds in this country—virtually all emergency shelters and voucher distribution programs. In February '96, we have an expectation from service providers that 607,000 homeless people will use emergency shelters, transitional housing programs, and other programs in the homeless housing system on an average day in February. The growth in transitional and permanent housing is highly attributable to federal funding through the McKinney Act, as well as the local money that it stimulated in response.

"We also see an incredible growth in soup kitchen meals. We have soup kitchen meal estimates only from big cities for 1987, not for the country as a whole. And our estimate in 1987 was 97,000 soup kitchen meals available in big cities at that time over an average day. In 1996, that number is up to 382,000, almost a four-time increase. And there isn't a lot of federal money here. So both of these are in response to demand, of course, but the growth in the availability of prepared meals is a response to demand that isn't being in any way driven by easy availability of money. It's being driven by need."

Nan RomanNan Roman is president of the National Alliance to End Homelessness, an education and advocacy organization with over 2,000 nonprofit and public-sector agencies around the country.

"Why is it that when we have such a booming economy, a time when we would expect to be seeing this kind of problem get better, we are not seeing it get better? I think that definitely has to do with housing costs and the increase of housing costs, but there may be some other factors as well.

"We have a whole infrastructure of mainstream programs that's obviously doing a worse and worse job, if the people that these programs are supposed to be serving are falling into homelessness. So what are the implications of this? What is it that we should be doing?

"I think we have to look at the turnover rate among homeless people in terms of shoring things up. We have developed, essentially, a parallel system that's serving a whole level of extremely poor people that are falling into it, out of the mainstream poverty programs. And I think we have to be thinking about whether this a sensible approach. Is this parallel system of homeless assistance doing a good job of ending homelessness for people? I think that remains a large question, after reading the data in this survey.

"Another issue is the degree to which the people who are homeless have interacted with other systems that are supposed to be helping them. At the time of this study 52 percent of homeless families with children were on welfare. So presumably those families are going to be even further challenged.

"These folks are very involved in all kinds of other mainstream systems of assistance that are supposed to be helping poor people in this country not end up homeless, and yet they're ending up homeless. I think that's where we've got to look-we've got to look for what's broken."

Sue MarshallSue Marshall, is executive director of the Community Partnership for Prevention of Homelessness. As vice-chairperson of the D.C. Housing Financing Agency, she provides a unique perspective on homelessness in D.C.

"If I get one point [about the homeless population] across to you today, it is that the fastest-growing segment of that population is women with children. And we really do need to give some serious attention to children starting out their lives in situations like a homeless shelter.

"The estimate of 2.3 percent of the District's population experiencing homelessness during a year's time is more than twice the national average. That's because the District has a disproportionately high portion of its population that is poor. I believe that some very small fraction of this is also attributable to the fact that the District is one of the few places in the country that has a history of a right-to-shelter law. We no longer have such a law on the books, but the programs that were developed out of that right-to-shelter initiative remain.

"I want to talk about some of the experiences we've had with programming, and basically what we've done with programming is deepen the level of services that are available to poor people who are homeless. Homelessness in the District of Columbia, and I would also suggest in other major metropolitan areas and rural areas as well, is a function of chronic poverty.

"When we looked at approximately two years' worth of data in the largest programs we have in the District for single adults, 48 percent of the shelter users stay less than seven days. That's almost half of the people who stay less than seven days. Seventy-nine percent stayed less than 60 days in that entire almost two-year period. If you look at the other end of the data—I think this is one of the most telling statistics, and one that should drive the policies that we implement—the 15 percent of men and women who stayed more than 90 days in our shelter system used up 75 percent of the bed nights; 15 percent of the population used 75 percent of the bed nights! So I believe we need to begin to examine the efficacy of building shelter systems that respond to that small proportion of the population.

"I'd like to talk about characteristics of what we find to be good programs that help move people out of homelessness. And the first principle is housing first. Yes, it is true that deep services are needed, that services are needed to match the conditions that are present. Substance abuse, mental illness, parenting, and the other issues that homeless people face become much easier to solve if a family or an individual is housed."

Questions from the Audience

Steve Redburn, OMB

"Is it possible, given that your numbers show a four-fold growth in facilities in big cities between the two dates, that the 1987 number may have been an undercount, because there was less access by the homeless to those services in '87?"

    Ms. Burt: "There certainly were fewer programs to access, so how many people we might have missed is probably more than we missed this time. The actual '87 estimates are only from cities over 100,000 [people] and only for service users of shelters and soup kitchens."

James LoewenJames Loewen, author

"There are three groups that seem quite disproportionately over-represented: African Americans, Native Americans, and males, generally. And I wonder if you have any thoughts and explanations as to why those three groups are so over-represented among the homeless?"

    Ms. Burt: "I don't know enough about Native American situations to say anything that would generalize about that. But I think if you look at the difference between Hispanic populations, which are not over-represented among the homeless, although they are equally poor, proportionately, and the African Americans, I think you have to be looking at issues of family structure and culture to try to understand why, however poor and however fractured, one of those cultures at least at the moment seems to be more willing to keep everybody inside, and the other one is less able to."

    Ms. Marshall: "I would suggest that as you look at African Americans, you also have to look economically at unemployment rates, look at marginalization from the economic system, I think, and recognize that what you're talking about here is a problem of persistent and chronic poverty."

Rochelle FriedmanRochelle Friedman, McCauley Institute

"What effect do you suppose the 1996 welfare legislation has had on what we're talking about today in terms of the needs and so forth of homeless services?"

    Ms. Marshall: "As time runs out on people's welfare entitlement, I predict that we're going to see lots and lots more women and children in the system. It presents, again, one of those policy challenges, and it is a question that is beginning to emerge here in Washington. Again, I think different programs all serving the same people need to better coordinate, to make sure the people are receiving services."

    Ms. Roman: "What we have seen is a really enormous increase in demand for food. And it seems like what's happening is it's not that people get sanctioned or removed from welfare and immediately the next day go to the homeless shelter to check in, it's probably going to take a lot longer. What they are doing is trying to accommodate, we think, the loss of income with an income substitute like food."

James HobenJames Hoben, HUD Office of Policy and Research

"One of the things that this study started to suggest to many of us, and I just wanted to put a light on it, is that homelessness may well be as much a product of social poverty as it is economic poverty.

"This is what I mean by social poverty, if you talk to the homeless providers and others. There are two frameworks: One is of persons who are totally broken and are lacking in self-esteem and self-worth, the other—with the case of males—is an entire lack of a social support network, somebody who cares enough to help you, let you live in their house, give you a couch, give you some money. There are incredible data here in this study on the vastly larger portions of homeless who have had no marriage, who are separated, the men, from their families. There is incredible data about the foster care experiences. If you look at the data on the source of incomes on these households, and we've got a fairly complete picture of it, not only do half have only half of what the American poverty level is for their household size, but very negligible amounts of money from any friends or any families.

"And I challenge the group here, and others, to look at these data from the standpoint of the issue of the coincidence of social and economic poverty resulting in homelessness.

Andy Mollison, Cox Newspapers

"Is the proportion of children the same in both the '87 and '96 samples?"

    Ms. Burt: "No, it's higher in '96. In '87, 10 percent of the households were households with children, and 23 percent of the people were in those households, of which—don't hold me to this, if you really want it I'll look it up—but I think it was 9 percent were the adults and 14 percent were the children. And now you're looking at 34 percent who are in—with the October estimate, 34 percent, who are either the adults or the children in the households, and the kids are about 21 percent of that."

Mr. Mollison: "There was another group that had a news conference this morning, at which they objected to the custom in some 40 cities of sending children from shelters to separate schools rather than to what I would call the normal public schools. Is there any indication at all that there might or might not be any difference in the results if you sent children who were being served by agencies to school or if you didn't send them to school, or if you educated them right in the shelter, do you have any indication about which is preferable?"

    Ms. Marshall: "My gut feeling tells me that you educate children as close to home as you can. That you present as few disruptions to their lives as possible. And I don't think that it serves any child well to be set apart in one more way from society. So we very strongly advocate neighborhood-based services that would allow children to be educated in the schools that they're used to."

Marsha MartinMarsha Martin, HHS

"I'm wondering if there's a way, with all of these efforts where we've seen government assist, is there a way to really tie this conversation to the challenges of the economy? How do we link these data, how do we link this information, how do you take your numbers and your needs and link it to services and take it back to where we want to go?"

    Ms. Marshall: "We are in the process of going through a strategic planning exercise that includes mainstream systems, the substance abuse agencies. I think the whole coordination between the service needs that we have and mainstream programs needs to be paramount. I think people who are in positions of leadership, both at local and national levels, should talk about coordination, talk about interdisciplinary approaches to solving problems.

    Ms. Roman: "I would say there are three big things we could do to make a difference. One is, we need this kind of data at every local level, because there's no way we can end homelessness if we don't know who is becoming homeless and why they're becoming homeless, how long they're staying in the system, and what the outcomes of that intervention are. So this is important national information, but every local jurisdiction needs to have this so that they can say, look, here's where we are now, and here is where we want to be with zero and what do we need to do to get there.

    "Second, these people are interacting already with the range of mainstream services that are failing them, essentially. So we've got to push this problem back up into those mainstream services. We've got to give them incentives to take care of these very poor, very troubled, complicated people. And if incentives don't work, then we've got to provide disincentives to allowing them to have the bad outcomes they're having now.

    "And then the third thing is, we need to break the population down into manageable pieces that we can do something about. Instead of taking people who are chronically homeless and chronically ill and housing them in the shelter system, which is inappropriate for them, very expensive to us because they not only live in the shelter system, they also interact with the criminal justice [system], emergency rooms, and so forth, at tremendous cost. We should be helping them with permanent supportive housing. It's cost-effective, it's 85 percent successful with the most disabled and chronically homeless people. It's the opposite of what we are doing."

    Ms. Burt: "There are a couple of comments that I would like to add to that. One is that we rely—everybody relies—on the big cities. The big cities are not, however, the instigators of all the homelessness that ends up being found there. If you are in a rural area, you move to the fringe of an urban area. If you're on the fringe of an urban area, you move into the center city. If you're in a small center city, you move to a bigger center city. People are not necessarily leaving because they're drawn to services, but they're leaving in part because there are no services and in part because they are in trouble where they are, and no one is helping them.

    "So some of the smaller communities also need to get on board here and not just ship their homeless people off to the big city. So you could stop some of this migration and then disconnection from the people who would help you if we had more assistance at the local level.

    "But I would also say that for 10 years now we haven't dared to say that housing subsidies would help. But if we fundamentally deal with this as a problem of housing, there's definitely people with vulnerabilities. It's a combination of poverty, housing, and other vulnerabilities. And so the people that fall off are the people that have lots of strikes against them. They don't have supportive families, they don't have support from the very frayed social safety net. They don't have skills and abilities. And they do have disabilities. When you have four or five of those things going against you, any crisis can put you on the street, and it's very hard to get back out.

    "So, if we start with the idea of offering or making available housing, if we just did the worst-case housing needs, we would be solving a fair amount of this problem, and we could then help people get skills, then help people get jobs, then help people beat their substance abuse problems, or perhaps they wouldn't need to be drinking quite so much because they would have some other things going for them."

Guy Raz, National Public Radio

"Would you say that the District of Columbia has the worst homeless problem in the country?"

    Ms Burt: "I doubt it. We don't have any local data from the National Survey of Homeless Assistance Providers and Clients. NSHAPC had a sample of 4,200 people all together, that covered the whole U.S. [United States]. There were roughly 76 areas, about 50 people came from each of them. And I am not willing to make any generalizations about New York City homelessness and D.C. homelessness based on 50 people from those jurisdictions. I don't know anybody who has actually looked at differences across localities in a systematic way."

    Ms. Marshall: "I would say that I don't think the District has the worst problem. I think we get into semantics about what does worst mean. I think the fact that the District was fortunate to have a cash investment from HUD in order to transform the system means that we have been well situated to begin to deal with the problem."

Full Transcript of the Session

Topics/Tags: | Housing

Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact [email protected].

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.

Email this Page