This sample Definitions and Estimates of Homelessness Essay is published for informational purposes only. Free essays and research papers, are not written by our writers, they are contributed by users, so we are not responsible for the content of this free sample paper. If you want to buy a high quality essay at affordable price please use our custom essay writing service.
Definitions and counts or estimates of a phenomenon are inextricably intertwined. One cannot count something—widgets, for instance—if one cannot tell whether an object is or is not a widget. In other words, the first problem that arises when trying to count “homeless people” is that defining the term is extremely difficult, both in the abstract and in the practical methods of social research. Further, definitions of a phenomenon such as homelessness usually embody one or more social purposes. They are not neutral, but rather are constructed to influence public concern and action. Thus they are “programmatic.” They may be trying to make a phenomenon visible, to get it defined as a problem and therefore worthy of intervention and amelioration, or to do the opposite. They may be trying to limit, or to expand, the scope of action expected or demanded. Or they may be trying to influence the value placed on one way of life in comparison to others.
Yet definitions are essential. From the perspective of immediate action, definitions identify who is eligible to receive whatever assistance is available specifically for homeless people. From a research perspective, definitions are necessary to identify who should be counted and described. And from a policy perspective, definitions are necessary to identify who should be planned for and what types of assistance they will need. The problem is, each of these purposes may require that homelessness be defined and measured in a different way. For this reason, the most useful research methods do not rely on a single definition of homelessness, but collect enough information about housing situations to allow analysts to construct samples based on different definitions for different purposes.
This entry looks first at the rather simple definition of homelessness used to allocate federal funding and at national estimates of the size of the homeless population based on that definition. It then discusses several issues surrounding definitions of homelessness and examines the issues involved in selecting various methods for obtaining estimates of population size.
The Federal Definition Of Homelessness
At present in the United States, government policy and access to particular kinds of government-supported assistance are driven by a clear but narrow definition of “literal” homelessness, which is based upon a person’s sleeping arrangements. Literal homelessness is defined on a day-by-day basis and refers to sleeping either in places not meant for human habitation, in an emergency shelter or transitional housing program serving homeless people, or in emergency accommodations paid for by a voucher from a program serving homeless people. This federal definition narrows the group of people identified as homeless to a fairly small proportion of the precariously housed or unhoused population. It is meant to help providers determine whom to serve, and to help planners calculate the levels of service to provide.
1996 Estimates Of The Number Of Homeless Persons
The most recent reliable source of national estimates of homelessness is the 1996 National Survey of Homeless Assistance Providers and Clients (NSHAPC). NSHAPC did not use a single definition of homelessness, but collected information that allows researchers to construct samples based on a variety of definitions.
Burt and her colleagues used NSHAPC data to develop point-in-time estimates for February and October of 1996 using the federal definition of homelessness. Their results indicated that between 444,000 (October) and 842,000 (February) people were homeless at those particular points in time and using homeless assistance services. These estimates include the children in homeless families as well as all service-using homeless adults.
Point-in-time estimates like these were the only type produced for more than a decade after homelessness was acknowledged to be a national problem in the early 1980s. However, during that period, awareness increased of the complexities of homelessness. One aspect of that complexity is the amount of movement into and out of homelessness over time, including the exit and reentry of the same people and the first entry of new households into homelessness. Making reliable estimates of period prevalence—the number of people who have experienced homelessness during a specified period of time such as one year—is much harder than taking a point-in-time count, because one needs a way to be sure that each person is counted only once.
The first good estimates of annual prevalence were not national. Dennis Culhane and his colleagues (1994), analyzing data from the Philadelphia and New York homeless management information systems (HMISs), found that four to six times the number of people homeless on a given day passed through the shelter systems of these cities in the course of a year. Calculations for Philadelphia and New York showed that during a single year’s time, these people included about 1 percent of the entire city population, and came close to or exceeded 10 percent of its poor people.
No HMIS exists on the national level that could produce annual prevalence estimates for the country as a whole. Nevertheless, as appreciation grew for the very great differences between point-in-time and annual numbers, the issue of annual estimates was important enough for Burt and her colleagues to develop a way to make projections using the February and October 1996 NSHAPC point-in-time estimates based on the federal definition of homelessness. The results: From 2,325,000 (based on October estimates) to 3,494,000 (based on February estimates) people experienced homelessness in a year’s time. As large as these projections seem, they are completely in line with the documented shelter use rates in New York, Philadelphia, and a few other cities—0.9 to 1.3 percent of the entire population, and 6.3 to 9.6 percent of poor people in the United States.
Issues In Defining Homelessness
Definitions of a phenomenon such as homelessness frequently require balancing between the two horns of a dilemma. If definitions are too inclusive, they become useless because too many people are ultimately covered and the phenomenon becomes too diffuse. With homelessness, this tendency is manifested by definitions that threaten to include the entire population of people in poverty, or everyone who is poorly housed. But if definitions are too specific, they focus too exclusively on the homelessness of the moment. They lead to policies and practices that are ameliorative but not preventive, because they do not address the larger question of desperate poverty and the pool of people at high risk for periodic bouts of literal homelessness.
During the past two decades, we have become increasingly sophisticated in our policy approaches to homelessness. We no longer can sustain the belief of the early 1980s that homelessness is an emergency situation that will go away as a national issue, and we increasingly look for ways to prevent homelessness as well as to end it for those to whom it occurs. As policy needs change, so too must we reconsider the value of definitions that were created to serve a particular policy purpose. For instance, the federal definition of literal homelessness that served throughout the 1990s, with its emphasis on being without a bed tonight or last night, is conveniently precise from some perspectives (it is relatively easy to ascertain where a person slept on the previous night and therefore easy to use in determining whom to help). However, it is overly precise and therefore misleading from other perspectives. If people can afford to pay for a motel room three nights a week but sleep in the park for the other four, week after week, because they cannot afford a room, it seems relatively meaningless to say that they are homeless only on the nights they sleep in the park. Truly helping them would mean helping them achieve stable housing. Similarly, if someone has no stable place to stay but a relative is willing to provide housing for two or three days until space is available in a homeless shelter, it seems relatively meaningless to say that the person is housed (because he or she is sleeping in conventional housing) and therefore not eligible for the homeless service. The ultimate failure of this narrow definition is that it does not help us address the larger issues of how to end homelessness for the long term, not just for a night. It only helps people decide who should receive services at any particular time.
Issues In Counting And Estimating Homeless Populations
Every number claiming to represent the size of “the homeless population” is an estimate, regardless of the method used to obtain it. Even when the basis of the number is a street count and enumerators know they have actually seen every person in the count, the final number is still an estimate. Enumerators will always have missed some unknown number of people and will try to compensate for those not counted in various ways. It is a rare count that, when published, does not include a statement that “there are at least this many homeless people” or “we doubled the number we actually interviewed to account for those we missed” (note that the decision to increase the reported count by any percentage is pure guesswork). As most counts are done at night, the odds of missing people who do not want to be found are quite high. The only numbers that are not estimates are those that claim merely to be reporting the number of people contacted—either through a survey, a street count, or a shelter-tracking database—and not “the size of the homeless population.”
Focusing on estimates of population size, we can try to understand the various factors that make an estimate more or less likely to reflect the population adequately. These include (1) where one looks, (2) how much information one gathers to help distinguish among different housing situations, (3) when and how long one looks, and (4) whether one uses a sample or tries to make contact with everyone.
Where One Looks
If one does not go to certain places where one might expect to find homeless people, the people staying in those places will not be included in one’s estimate.
This is the definitional bottom line—whatever the theory, the reality of who is included in an estimate depends on where the enumerators go. Therefore the first thing that should be checked when reading reports of homeless counts is what locations the enumerators did and did not include. For example, studies that go only to shelters miss homeless people who do not use shelters. Many local efforts to count homeless people do only shelter counts. Studies that only search downtown streets miss anyone who stays outdoors in residential areas.
Very few street counts cover an entire jurisdiction, yet in some communities homeless people with specific disabilities—mental illness in particular—avoid downtown areas because they feel too vulnerable there to harassment and victimization. Along with gaining agreement on which housing circumstances should be called homelessness, the biggest challenge to obtaining good estimates is finding people who do not use shelter arrangements.
It is relatively easy to count and interview people who sleep in a residential program. The challenge is choosing which programs to include in one’s study. Emergency shelters are pretty obvious. But lines are increasingly blurred around programs that offer various degrees of permanence for people coming directly from the streets. For instance, some no/low demand residences such as safe havens allow residents to stay “as long as needed” although the housing is not intended to be permanent—are the people staying in them homeless or are they in permanent housing? Some transitional housing programs put households directly into housing they can retain even after supportive services end—are these tenants homeless or in permanent housing? Every study needs to make these decisions, but different studies may decide differently, making their estimates difficult or impossible to compare. Finding people who do not use the homeless assistance network’s shelter and housing resources is even more of a challenge. Various techniques have been used, including street searches, following outreach teams, and going to feeding programs (for example, soup kitchens and mobile food vans), health care programs, and drop-in or warming centers. All these approaches require a method for assuring that a particular person is not counted more than once, since the same person may use more than one service, either on a given day or during the days the enumeration is taking place. In general, no matter how thorough, street counts will miss many homeless people because they do not want to be found or because finding them might be dangerous to the searchers. For most communities, techniques involving sampling from relevant services (including outreach) will capture higher proportions of non-shelter-using homeless people than straight street counts, especially if the counts are done at night.
Finding people who are not presently homeless but are at imminent risk is the hardest task of all, and one that cannot be done within the same study framework as most studies of homeless populations. The places one would look for the imminent risk population include institutions (psychiatric hospitals or wards, jails, and prisons) and conventional housing units. As prevention becomes an increasingly important part of systematic efforts to end homelessness, we will need better techniques for estimating the population at imminent risk, which will indicate the demand for interventions to prevent homelessness.
The more places one includes as locations to search for homeless people, the greater the odds that some of the people in those places will not be homeless now and may never have been homeless. For instance, feeding programs such as soup kitchens are excellent locations for finding non-shelter-using homeless people, but on average about half of soup kitchen users are not currently homeless. If one includes soup kitchens in one’s enumeration to increase the coverage of street homeless people, one will need to collect enough information to separate the homeless users from the non-homeless users.
Therefore one cannot merely count heads; one needs to obtain specific information about housing situations, at present and in the past. The more details one has about current, recent, and even long-ago housing situations, the more flexible the data will be for accommodating a variety of definitions of current or former homelessness and housing instability. Another consequence of searching in many types of places for homeless people is that the same people are likely to have used more than one of the places. Early homeless counts attempted to deal with the problem of duplicate counting by limiting their time frame to a few hours at night, and some still do so today. The problems and inaccuracies of night counting quickly became apparent, however, and led to the search for alternative approaches for assuring that people are not double and triple counted in the final estimate. Two approaches emerged—developing unique identifiers, and getting service use information from each respondent. Unique identifiers, which are most likely to be used in studies that try to capture all homeless people, allow analysts to compare identifying information from interviews or contacts with homeless people done at many locations to detect and eliminate duplication in population counts, while still retaining the information about how many people use each different service location. Service use information from respondents allows analysts to use statistical techniques to unduplicate, and it is most common in studies that use sampling rather than trying to make contact with every homeless person.
When and How Long One Looks
“Point prevalence” refers to the number of people who are homeless at a single point in time—usually one day or one night. “Period prevalence” refers to the number of people who have been homeless during some longer time period such as a month or a year. Most efforts to enumerate homeless people cover a very short time period and produce estimates of point prevalence. They give a count and provide descriptions based on the people who are homeless on a single day or single night. There have also been some attempts to use interview data from point-in time samples to estimate the number of people who might be homeless during the course of a year. To come close to the results of shelter-tracking databases when doing this type of projection, one needs a large, statistically valid sample that is very inclusive of sampling sites and information from respondents about the length and pattern of their homelessness. The annual estimates derived from NSHAPC, presented above, used this technique.
Two alternatives to the one-night approach provide important techniques for getting accurate estimates of population size. The first is the computerized shelter-tracking database that covers all (or most) “homeless beds” in a jurisdiction. Because each person in these databases has a unique identifier, all service episodes (for instance, nights in shelter) used by one person can be linked to that person. These databases, now known most frequently as homeless management information systems, give a jurisdiction knowledge of how many distinct individuals have used the system during any time period of interest, from one day to one year to however long the system has been operating. The Philadelphia and New York data cited earlier come from systems of this sort. It is important to remember, when using population counts from an HMIS, that the counts cover only the types of people the HMIS covers—usually emergency shelter users. In most communities, this is not the entire homeless population. The second alternative to a one-night count is an enumeration over an extended period of time—say six to eight weeks—conducted in mainstream agencies such as welfare offices, food pantries, food stamp offices, and community action agencies as well as in programs targeted specifically to homeless people. The Kentucky Housing Corporation has used this technique twice in statewide studies to estimate population size. The mainstream agencies use two brief screening questions to identify people among their clients who may be homeless, and a brief follow-up questionnaire gathers information to construct unique identifiers as well as answer some basic questions about the homeless condition. This technique is particularly appropriate for jurisdictions with relatively sparse populations and/or relatively few homeless specific services. Kentucky developed the approach because all but 7 of its 120 counties are rural.
Sampling or Counting
Counting seems like a straightforward process. But in fact, producing an accurate estimate of the size of the homeless population at a particular point in time from simple counting techniques is extremely difficult. Understanding the flows of people into and out of homelessness cannot be accomplished with a simple count. HMISs can provide both literal counts and descriptions of flows (assuming they contain the right data fields), but an HMIS is a very difficult thing for a community to develop, and it will rarely cover the entire homeless population.
Studies based on sampling techniques require more methodological knowledge but also give more flexibility. One need interview fewer people for the same results and can go into greater depth with them about their circumstances and conditions. One can use screening questions to make it feasible to check for homeless people in places where they may be a small minority of service users, which is useful for finding homeless people in jurisdictions without many homeless-specific programs. The Kentucky Housing Corporation survey provides an example of a simple two-question screener: (1) In what type of place are you now staying? (2) Is that your permanent place to stay? If the answer to the second question is “no” or “unsure,” the interview continues on the assumption that the person is homeless. Finally, if one wants to find people at imminent risk of becoming homeless (to estimate, for instance, the likely level of need for services), sampling techniques are essential for examining overcrowded households, particular neighborhoods, and institutionalized populations that are major senders of people into literal homelessness. However, when it comes to estimating the annual prevalence of homelessness, sample-based studies are limited to projections based on what currently homeless people say about their homeless experiences.
No one has done a perfect enumeration of homeless people, and no one is likely to do so. Resource constraints, the slipperiness of homeless definitions, and multiple policy purposes for homeless studies make this prediction almost certain. Also contributing to the difficulty are the different sources of data for estimating the size of the homeless population and the different legitimate uses to which such numbers can be put. There is no one right number. Different types of estimates serve different purposes, and all are useful. Someone interested in service planning needs to know about the expected level of service contacts on a given day and will not care so much whether the contacts are made repeatedly by a relatively small group of people or only once each day by a very large number of different people. Someone who wants to create permanent housing for the longest-term homeless people with disabilities will need to know the actual number of such persons, not just at one point in time. And someone who wants to stop family homelessness through prevention efforts needs to know how many families might be at risk of homelessness during a particular time period and whether they need temporary crisis assistance or long-term supports to remain in their housing.
The 1980s began with only the crudest ways to estimate the level of homelessness. Throughout the decade and into the 1990s, the assessment of homelessness has become more sophisticated, both in the use of numbers for different purposes and in the ability to determine these numbers. It is critically important for those who wish to make policy or to influence it to have a very clear understanding of what they are trying to do and for whom they want to do it. It will then be much easier to identify the right number for that specific purpose. It is also important to realize that bad policy will result from using the wrong numbers or from using numbers that confuse rather than clarify the nature of the policy task.
- Burt, M. R. (1994). Comment. Housing Policy Debate, 5(2), 141-152. Retrieved April 14, 2003, from http://www.fannie maefoundation.org/programs/hpd/pdf/hpd_0502_burt.pdf
- Burt, M. R. (1996). Homelessness: Definitions and counts. In J. Baumohl (Ed.), Homelessness in America. Phoenix, AZ: Oryx Press.
- Burt, M. R. (1996). Practical methods for counting the homeless: A manual for state and local jurisdictions. Washington, DC: Urban Institute Press.
- Burt, M. R. (1999). Demographics and geography: Estimating needs. In L. Fosburg & D. Dennis (Eds.), Practical lessons: The 1998 National Symposium on Homelessness Research. Washington, DC: U.S. Departments of Housing and Urban Development and Health and Human Services.
- Burt, M. R., Aron, L. Y., & Lee, E. (2001). Helping America s homeless: Emergency shelter or affordable housing? Washington, DC: Urban Institute Press.
- Bylund, R. A., Rudy, D. R., & Parkansky, S. (2001). 2001 Kentucky homeless survey. Frankfurt, KY: Institute for Regional Analysis and Public Policy, Morehead State University, and Kentucky Housing Corporation. Retrieved April 14, 2003, from http://www.kyhousing.org/Publications/resources/2001HomelessReport.pdf
- Culhane, D. P., Dejowski, E., Ibafiez, J., Needham, E., & Macchia, I. (1994). Public shelter admission rates in Philadelphia and New York City: Implications for sheltered population counts. Housing Policy Debate, 5(2), 107-140. Retrieved April 14, 2003, from http://www.fanniemaefoundation. org/programs/hpd/pdf/hpd_0502_culhane.pdf
- Culhane, D. P., & Hornburg, S. P. (1997). Section I: Defining, counting and tracking the homeless. Understanding homelessness: New policy and research perspectives. Washington, DC: Fannie Mae Foundation.
- Kentucky Housing Corporation. (1993). Kentucky homeless survey preliminary findings. Frankfurt, KY: Author. Retrieved April 14, 2003, from http://www.kyhousing. org/publications/resources/1993HomelessSurvey.pdf
Free essays are not written to satisfy your specific instructions. You can use our professional writing services to order a custom essay, research paper, or term paper on any topic and get your high quality paper at affordable price. UniversalEssays is the best choice for those who seek help in essay writing or research paper writing in any field of study.