Results of the pilot study to estimate the size of the hidden homeless population in Metro Vancouver
Extracts from the Executive Summary
Margaret Eberle, Deborah Kraus, Luba Serge
Mustel Research Group, marketPOWER Research Inc.
Most of the elements of homelessness as defined by various studies are incorporated in existing data sources. For example, homeless counts in different cities in Canada demonstrate that communities are able to gather information about the number and characteristics of people living on the street, in shelters and, if known to outreach workers, places unfit for year-round habitation such as campgrounds or abandoned buildings. Statistics Canada and CMHC data provide several options for estimating the size and nature of the at risk population, including those that are overcrowded or living in inadequate housing. At present, there is no reliable source for data on the size of the hidden homeless population. This project tested an approach to estimate the size of the hidden homeless population in Metro Vancouver.
One of the key tasks was to clarify the definition of hidden homelessness to be used in this study, and specifically to develop a way to operationalize the definition in the context of the random household survey method. Two approaches were used – a literature review and a scan of communities for their definition of hidden homeless. The selected definition of hidden homelessness was intended to capture a range of situations:
Hidden homeless persons are people staying temporarily with another household and who do not have a regular address of their own where they have security of tenure.
The literature suggested that the following variables are important elements of a nuanced operationalization of the definition of hidden homelessness:
• Relationship to head of household (e.g. friend, relative etc.);
• Age group (e.g. under 25 yrs);
• Sleeping arrangement (e.g. couch, floor, basement, garage, etc.);
• Owner/tenant (or leaseholder) satisfaction/dissatisfaction with the arrangements; and
• Financial or in-kind contribution.
The number of these variables that could be operationalized in this short random survey was limited. As such, two of these variables were selected – family relationship and a version of owner/tenant satisfaction with the arrangement. While the relationship of the hidden homeless person to the host household was considered an important dimension for descriptive purposes, family relationship was not enough to exclude a person from consideration as hidden homeless. The rationale is that a homeless person may exhaust their welcome with family members, as well as others, with numerous or lengthy stays. Host dissatisfaction (whether family or not) with this type of arrangement ultimately results in a precariousness that represents lack of security of tenure. “Host household satisfaction with the arrangement” was felt to supersede family relationship. If the host household is dissatisfied with the relationship then family status, payment of rent, sleeping location or other variables is inconsequential. Host household satisfaction was determined by asking if the visitor could stay as long as they needed to establish a home of their own. Thus a person who was a member of the immediate family was considered hidden homeless if they could not stay in the host household as long as they needed to establish a residence of their own. Importantly, this approach eliminated from consideration as hidden homeless a situation where a youth was living in the family home while in school, for example, and could stay as long as he or she wishes.
Among the 1,027 completed household interviews, 35 host household representatives reported having 49 individuals living with them at the time of the survey. When the definition of hidden homelessness was applied using the screening question that the individual visitor “cannot stay with you until they are able to establish a residence of their own”, the number of positive responses was reduced. Applying this criterion reduced the number of households reporting a hidden homeless visitor from 49 to 8 households. They were accommodating 12 hidden homeless persons, 5 of whom were family members. Projecting to the total population of Metro Vancouver households, it is estimated that there were 9,196 hidden homeless persons at the time of the survey. Most of them would have been un-related to the host household. The number of hidden homeless individuals in Metro Vancouver in the past year was estimated to be 23,543 persons. Most (18,000 or 75%) of these individuals were non-family members.
Since the incidence of hidden homelessness is considered a statistically rare phenomenon, these estimates produce fairly broad interval estimates. The margin of error is 7,650 at the 95% confidence level. That means that 95 times out of 100 the interval from 1,545 persons to 16,846 persons includes the actual number of hidden homeless individuals in Metro Vancouver. A much larger (and impractical) sample size would have been necessary to provide narrow interval estimates. Despite the wide variance, and given that the survey has been conducted twice in L.A. with similar results, both the L.A. study statisticians and the statistician involved in the present study concluded that the method represents a reasonable approach for estimating the size of the hidden homeless population.
Qualitative research was designed to learn from hidden homeless individuals about their previous housing situation and barriers to obtaining stable housing. Two qualitative interviews with hidden homeless persons were completed. This limited number means it is not possible to draw any conclusions about the characteristics of hidden homelessness and pathways into and out of hidden homelessness.
The study provides an estimate of the size of the hidden homeless population in a Canadian community using empirical methods. It provides figures for the estimated number of hidden homeless individuals in Metro Vancouver over four weeks in January and February 2009 and an estimate of the number of hidden homeless individuals in Metro Vancouver over the course of a year. The figures may be an underestimate of the actual number hidden homeless due to the limitations of telephone survey research such as exclusion of some non- English speaking and cell only households. It represents findings for one community. Given variations in regional housing markets and other contributing factors, this estimate cannot be applied elsewhere.
Policy implications are related to the potential relationship between the number of hidden homeless persons and the number of absolute homeless persons. There were 2,660 absolute homeless people in Metro Vancouver counted on one day in March 2008. The estimate of 9,196 hidden homeless persons may be viewed as an indicator of housing instability or precariousness that may predict future levels of absolute homelessness. However, there is insufficient information to determine the existence or strength of this relationship. Data for both measures over several years would be required to determine the relationship.