The National Shelter Study 2005-2009 — Highlights

The National Shelter Study is the first national analysis using consistent shelter data collected over an extended period of time to establish a baseline count and description of the characteristics of the homeless population in Canada. This study uses information gathered from emergency homeless shelters using the Homeless Individuals and Families Information System (HIFIS) and emergency homeless shelters in the City of Toronto.

Though not all homeless people regularly use them, emergency shelters are a common point of contact for those experiencing absolute homelessness. Emergency shelter use, measured over a period of years, currently serves as the best available indicator for understanding baseline trends in the size and composition of the homeless population. At the time of the study, there were nearly 400 emergency shelters with over 15,400 beds across Canada (see Table 1).

Table 1: Emergency shelters in Canada (2009)
Province/Territory Emergency Shelters Permanent Beds
Newfoundland and Labrador 5 64
Nova Scotia 6 190
New Brunswick 6 124
Prince Edward Island 2 12
Quebec 88 2,143
Ontario 132 6,806
Manitoba 13 627
Saskatchewan 18 328
Alberta 37 2,950
British Columbia 80 2,078
Yukon 3 16
Northwest Territories 5 109
Nunavut 1 20
Canada 396 15,467

About HIFIS

HIFIS 3 is a free, easy–to-use, electronic records management system built for emergency shelters and other homelessness service providers. It assists in daily operations such as registering clients and reporting on shelter use, and enables the collection of homeless statistics and data about the population using shelters. The system assists communities with their long-term planning and capacity building to address local challenges related to homelessness.

For more information about HIFIS, please e-mail NC-HIF-CP-PC-SIS-GD@hrsdc-rhdcc.gc.ca.

The National Shelter Study uses a stratified cluster sample to produce national estimates of the annual number of emergency shelter users between 2005 and 2009. Information from more than 655,000 client stays at over one quarter of Canada's emergency homeless shelters was used in the study (see Methods inset).

Methods

  • Study design: stratified cluster sample of emergency shelters
  • Eight sample strata based on shelter gender (serving males, females, or both) and target clientele (families, youth, general, or women with children)
  • Sampling frame contains information about all emergency shelters in Canada. Sample shelters selected with probability proportional to size
  • Only emergency shelters are used; the study does not include transitional housing or violence against women (VAW) shelters
  • Between 96 and 123 shelters, or 24 to 30 percent of all emergency shelters in Canada, contributed data for each year of the study
  • Estimates are based on over 655,000 shelter stays between 2005 and 2009
  • To avoid double counting, the sample design adjusts for clients who used more than one shelter

A National Portrait of Homelessness

In 2009, an estimated 147,000 people, or about 1 in 230 Canadians, stayed in an emergency homeless shelter. This number has not changed significantly since 2005, when about 156,000 Canadians used a shelter (because these estimates are based on a sample, there is a margin of error, see Figure 1). However, the number of bed nights used annually in Canada's shelters increased from just under 4.5 million in 2007 to nearly 5.3 million in 2009 (see Figure 2). On any given night in 2009, an average of 14,400 shelter beds were in use out of the approximately 15,400 permanent shelter beds available. Despite no statistically significant increase in the total annual number of unique individuals using shelters, there were over 2,000 more people sleeping in homeless shelters on an average night in 2009 compared to 2007.

Figure 1: Estimated annual number of unique individuals using emergency shelters in Canada*
2005 2006 2007 2008 2009
Unique Individuals 156030 150663 146884 151621 146726
95% Confidence Interval Lower bound 142804 138015 134177 137265 134224
Upper bound 169256 163312 159591 165977 159229

* margin of error (95% confidence interval)

Figure 2: Estimated annual number of bed nights used at emergency shelters in Canada
2005 2006 2007 2008 2009
Annual number of bed nights used 4759753 4668304 4463715 4783768 5263182

This can be explained by longer average stay lengths, especially at family shelters (see Table 2). The average length of stay at family shelters grew from around 30 nights at the beginning of the study to about 50 nights in 2008 and 2009. The increase was small yet statistically significant at other types of shelters—around two to three nights longer on average—but this translates into many more bed nights being used across Canada.

Table 2: Average length of a single shelter stay (nights) by shelter type
Shelter Type 2005 2006 2007 2008 2009
Family 33.4 32.2 31.4 49.8 50.2
General 12.2 11.7 12.7 11.1 13.9
Women with Children 11.2 13.9 15.0 14.9 17.4
Youth 18.9 18.2 16.5 15.1 18.8
Total sample 13.6 13.0 13.8 12.7 16.0

Looking more closely at the data, Figure 3 shows that the average stay length for clients with a single stay over the course of a year increased from about 20 nights in 2005 to nearly 30 nights in 2009 but held steady (around 10 nights) for clients with multiple shelter visits. Temporary, one-time shelter users (who accounted for 67% of the sample in 2009) are staying at shelters longer, suggesting they may be finding it increasingly difficult to resolve their homeless episodes.

Figure 3: Average length of stay (nights) for clients with a single stay versus clients with multiple stays
Client 2005 2006 2007 2008 2009
Clients with a single stay 19.8 21.3 23.3 24.1 29.0
Clients with multiple stays 10.1 9.7 10.2 8.7 10.6

Quick facts

  • The majority of shelter users (67%) had only one shelter stay in 2009
  • Only 7% of clients at family shelters had more than one stay in 2009
  • The average occupancy rate at family shelters was over 100% in 2009
  • Less than half of all stays by youth (age 16-24) were at youth shelters
  • Over one quarter of all shelter stays lasted just one night
  • The proportion of long stays (one month or more) increased from 12.6% of stays in 2005 to 16.7% in 2009

Demographic Trends

In 2009, 73.2% of shelter users age 16 and over were male, 26.7% were female and 0.1% specified another gender. Looking at gender proportions by age group (Figure 4), the gap between males and females widened with older age groups. Males comprised just over 60% of youth using shelters but nearly 80% of adults 55 and over using shelters.

Figure 4: Gender distribution by age group (2009)
Gender Under 16 16-24 25-54 55 and over
Female 50.1% 36.9% 24.2% 20.7%
Male 49.8% 63.0% 75.7% 79.3%

The average age of adult shelter users in 2009 was 37 years. Well over 80% of shelter users were between the ages of 16 and 54. As Figure 5 illustrates, the proportion of shelter users 55 and over drops sharply and just 1.7% of shelter users were over 65.

Figure 5: Distribution of shelter users by age group (2009)
Under 16 16-24 25-34 35-44 45-54 55-64 65 and over
Distribution (%) 6.5% 20.6% 19.9% 23.7% 21.0% 6.6% 1.7%

There were small but statistically significant changes in the age distribution of shelter users over the 2005 to 2009 study period. Children comprised 6.5% of shelter users in 2009, up from 4% in 2005. The proportion of adults 55 and over increased slightly as well, from 6.9% to 8.3% of all shelter users. The proportion of adults age 25-54 decreased from 67.7% in 2005 to 64.6% in 2009. Youth (ages 16 to 24) held steady, comprising about 20% of shelter users annually between 2005 and 2009.

Quick facts

  • The median length of a shelter stay for children increased from 16 days in 2005 to 24 days in 2009
  • The average age of children in shelters was 6.5 years
  • Approximately 80% of shelter users 55 and over were men
  • Older adults (55+) were much more likely to have a long shelter stay (one month or more) than adults ages 25 to 54
Figure 6: Estimated annual number of children (under 16) using emergency shelters
Users 2005 2006 2007 2008 2009
Under 16 6205 6533 7463 7290 9459

In absolute terms, the estimated number of children using emergency shelters grew from 6,205 in 2005 to 9,459 in 2009 (see Figure 6). Note that this study does not include children staying in Violence Against Women (VAW) shelters. In 2009, an estimated 29,964 youth and 2,567 seniors (65 or over) used emergency shelters (see Figures 7 and 8).

Figure 7: Estimated annual number of youth (age 16 to 24) using emergency shelters
Users 2005 2006 2007 2008 2009
Age 16 to 24 31891 30501 30210 29349 29964
Figure 8: Estimated annual number of emergency shelter users age 65 and over
Users 2005 2006 2007 2008 2009
Age 65 and over 2273 2214 2338 2467 2567

Conclusion

Despite a growing national population, the number of Canadians using emergency shelters remained stable at approximately 150,000 annually during the 2005 to 2009 study period. This should be considered a baseline estimate of homelessness, as there are undoubtedly many more people experiencing homelessness on the streets, couch surfing, or in unsafe, inadequate or temporary housing. Although the annual number of individuals using shelters did not change significantly over the 2005 to 2009 study period, the demographic composition of the shelter-using population has changed, especially with respect to children and families, and shelters are being used more intensely, with longer average stays.

For more information on the Homelessness Partnering Strategy, please visit www.hrsdc.gc.ca/homelessness.