Correlates of Homeless Episodes among Indigenous People

Abstract

This study reports the correlates of homeless episodes among 873 Indigenous adults who are part of an ongoing longitudinal study on four reservations in the Northern Midwest and four Canadian First Nation reserves. Descriptive analyses depict differences between those who have and have not experienced an episode of homelessness in their lifetimes. Multivariate analyses assess factors associated with a history of homeless episodes at the time of their first interview. Results show that individuals with a history of homeless episodes had significantly more individual and family health, mental health, and substance abuse problems. Periods of homelessness also were associated with financial problems. Among the female caretakers who experienced episodes of homelessness over the course of the study, the majority had been homeless at least once prior to the start of the study and approximately one–fifth met criteria for lifetime alcohol dependence, drug abuse, or major depression. Family adversity during childhood was also common for women experiencing homelessness during the study.

Correlates of Homeless Episodes among Indigenous Adults

American Indian and Alaska Native (AIAN) people make up only 1.0% of the U.S. population (U.S. Census Bureau, 2010), yet particularly in the 11 states that have significant AIAN populations, they are over-represented among urban homeless people. Although good estimates are extremely sparse and often the ethnic category for AIAN is omitted entirely from reports, there is evidence of over representation dating from the 1950s. In Chicago in 1957–1958, AIAN homeless people (1.9% of the population at that time) were approximately three times their proportion of the city’s population (.06%) (Bogue, 1963). This over-representation in certain cities has been documented over time in nation-wide estimates (Momeni, 1989).The National Survey of Homeless Assistance Providers and Clients (NSHAPC, 1996) estimates the proportion of AIAN urban homeless people at eight percent nationally (Burt, Aron, Douglas, Valente, Lee, & Iwen, 1999). However, the best estimates come from individual cities. For example, Koegel and Burnam (1988) projected that American Indian people made up five to six percent of Los Angeles County’s homeless population though they were only approximately 0.6% of the population. Gelberg and colleagues reported 5.5% American Indian homeless people among the Los Angeles beach population (Gelberg, Linn, & Mayer-Oakes, 1990). In the reporting year 2007–2008, AIAN people made up 6% of the sheltered homeless population in Seattle and 29% in Anchorage (HUD, 2008).

What we know about AIAN homeless people comes mostly from urban surveys. The most comprehensive report regarding reservation homelessness is the Wilder Research Foundation’s “Homeless and Near-Homeless on Northern Minnesota Indian Reservations” study (Owen, Heineman, & Decker, 2007). The authors provided estimates of the number of people who were currently homeless (9%), doubled-up near-homeless in substandard housing (14%), and doubled-up near-homeless in standard housing (77%) on six reservations in northern Minnesota. However, they did not ask the respondents about their lifetime history of homeless episodes. The research reported here expands our understanding of Indigenous (American Indian, Canadian First Nation) reservation/reserve homelessness by investigating characteristics of ever-homeless and recently homeless Indigenous adults who are part of an ongoing longitudinal study of Indigenous families on four U.S. reservations and four Canadian reserves in the northern Midwest U.S. and Canada. We also provide multivariate and bivariate correlates of having ever experienced homeless episodes.

Factors associated with homeless episodes among Indigenous people

Economic disadvantage and unemployment

According to a 2003 report by the U.S. Commission on Civil Rights (USCCR) Indigenous people, “rank at or near the bottom of nearly every social, health, and economic indicator” and that the “socioeconomic condition of the [Indigenous] population in the United States reveals a dire need for increased national attention” (8). Decades of research have documented high rates of unemployment (Sandefur & Liebler, 1997; USCCR, 2003) and poverty (The Henry J. Kaiser Family Foundation, 2003; 2007; Trosper, 1996) often coupled with inadequate job training and educational opportunities (USCCR, 2003). Nationally the 2009 unemployment rate for AIAN people was 13.6% vs. 8.2% for European Americans. But on some reservations/reserves the unemployment disparity is much greater. For example, estimates of unemployment on Pine Ridge Reservation in South Dakota range from between 83% – 85% (Schwartz, 2006) to 47.6% (Chapman, Conti, Frank, & Gansen, 2003). In 2007, the unemployment rate among the participants in this study was 18.4%. To cope with economic disadvantage Indigenous individuals and families are very mobile, moving on and off reservations and reserves in search of opportunities (Larriviere & Kroncke, 2004; Snipp & Sandefur, 1988). This combination of economic disadvantage and mobility can place them one lost job or one paycheck away from homelessness.

Overcrowding

According to the U.S. Census Bureau (2000) 14.7% of homes on tribal lands are overcrowded compared to 5.7% of homes in the general population. The Wilder Study indicates that among six northern Minnesota reservations, 62% of the near homeless had been living with someone for more than one year and 31% had been without their own home for three years or longer. Doubling up often means high mobility as people move away from overcrowded conditions or wear out their welcome. Among those in the Wilder study who had been without their own housing for one year or longer, only 19% had been in the same place for the last 12 months and 41% had lived in four or more places. Approximately two-thirds of the Wilder study respondents were living in overcrowded housing (Owen, Heineman, & Decker, 2007).

Access to affordable housing

AIAN people have difficulty accessing affordable housing either as potential buyers or as renters. Loan policies and availability of lenders have resulted in the lowest home ownership rate of any ethnic group (Youmans, 2002), less than half the rate for the general population (Enterprise Institute, 2004). A recent study by the National Health Care for the Homeless Council (NHCHC) found that on reservation land, individuals wait an average of 41 months for low income housing units (Zerger, 2004), twice the national average. In urban areas, renting can pose problems. There is evidence from Minnesota, Montana, and New Mexico of discriminatory practices favoring European Americans in the rental market, limiting housing choices, and increasing costs for AIAN people. The three-state average rate of discrimination against AIAN renters was 28.5% (Turner & Ross, 2003).

Substandard housing

Available housing on reservations is often substandard. According to the U.S. Commission on Civil Rights (2003), 40% of on-reservation housing is substandard, compared to about six percent nationwide. Census data indicates that on tribal lands, 11.7% of homes lack plumbing compared 1.2% in the general population (Census Bureau, 2000). In the Wilder Study 15% of the housing reported by those doubling up lacked electricity, running water, a flush toilet, or central heat (Owen, Heineman, & Decker, 2007).

Family and culture

Mobility is also affected by concerns pertaining to loss of culture, tradition, family connections, and homesickness (Joffer & Wagner, 1996). Living on a reservation/reserve represents a unique social context. For Indigenous nations lucky enough not to be completely removed from their home territories, reservations/reserves represent the remnant “homeland.” However, this “homeland” often occupies the least productive, least desirable area of what was once their vast territory and offers very few economic opportunities. As a social context, reservations/reserves are at once a symbol of what was and the representation of what has occurred. Simply living on reservations/reserves can be a reminder of ethnic cleansing, broken promises, continual encroachment on tribal lands, and economic disadvantage. At the same time, reservations/reserves may be a refuge from discrimination and a symbol of the living culture. They hold sacred places, extended kin, and remain the repository of cultural knowledge. In Joffer and Wagner’s (1996) study of return migration among Indigenous people in South Dakota, 41% said they left their reservation for economic reasons, yet only 16% gave “job” and 4.9% “cost of living” as a reason for returning. More than half (56.8%) gave social environment, family, and cultural reasons for returning. There is some evidence that episodes of homelessness away from reservations and reserves are more stressful than those on reservations and reserves. According to the Wilder report “the general level of distress in the American Indian homeless population not residing on reservations is somewhat higher than for adults in the general homeless population” (Owen, Heineman, & Decker, 2007, 2). When things go badly venturing off the reservation/reserve it is natural to want to return home even if it means doubling up.

Characteristics of Indigenous people who have experienced episodes of homelessness

The few studies of Indigenous people who are homeless indicate high levels of substance abuse typically associated with homeless people. Some studies have found that Indigenous homeless people manifest higher rates of substance misuse than other racial/ethnic groups (Kahn, Hannah, Kirkland, Lesnik, Clemens et al., 1992; Kasprow, 1998; Westerfelt & Yellow Bird, 1999). Others show Indigenous people who are homeless use alcohol and drugs at about the same rates as homeless people of other ethnic groups (Gamst, Herdina, Mondragon, Munquia, Pleitez, et al., 2006). In the Wilder study, respondents who had experienced actual homeless episodes rather than being near-homeless (i.e., doubling up) were more likely to report alcoholism or chemical dependency (52% vs. 32%) (Owen, Heineman, & Decker, 2007).

Other than some information on symptoms and measures of services utilization, there is very little information regarding the mental health status of Indigenous people who are homeless. In the Wilder study, the level of self-reported mental health problems of those who had experienced actual episodes of homelessness (56%) was very similar to that of the near-homeless (51%), but those who had been actually homeless reported higher levels of reported traumatic brain injury (40% vs. 29%) (Owen, Heineman, & Decker, 2007). Westerfelt and Yellow Bird (1999) reported a much lower rate of psychiatric hospitalization among Indigenous homeless people (6.7%) compared to their non-Indigenous counterparts (24.3%).

As is the case with many people who have experienced episodes of homelessness, a high percentage of homeless Indigenous people report dysfunctional family histories. In a recent study by Gamst and colleagues (2006), 25% of Native adults reported that prior abuse contributed to their homelessness, compared to 16.0% among the total sample. There were substantial differences in levels of physical abuse (46% vs. 37%), sexual abuse (33% vs. 25%), and neglect (34%, vs. 20%) between respondents in the Wilder study who were actually homeless and those who were near-homeless (Owen, Heineman, & Decker, 2007).

Theory and Research Questions

According to the life course perspective, there are developmental and contextual explanations for homeless episodes and the persistence of social marginality. Social location shapes life trajectories by influencing the occurrence and timing of major life transitions such as completing one’s education, marriage, entry into the work force, and career trajectories. Historical and social structural events may delay or eliminate the opportunities for “on-time” social transitions. For example, inadequate schools, poor academic performance, and early school leaving may result in lost job opportunities, as will living in an economically disadvantaged area with high unemployment. Negative life events occurring in disadvantaged social contexts may become self-perpetuating. Poor academic performance and school leaving coupled with high unemployment severely reduces chances of getting a job. What were originally contextual disadvantages may persist even if one moves to a new community where jobs may be more plentiful in that only low-paying, low-benefit jobs may be available. Mental health problems, substance abuse, or arrests exacerbate problems in finding and keeping well paying jobs. These combinations of life course factors place the individual a job or a financial emergency (e.g., a car break down, medical bills) away from not being able to pay the rent.

In this study we investigated the histories and characteristics of individuals who have experienced episodes of homelessness in their lifetimes. Our research questions focused on differences in life experiences and personal characteristics of those who have experienced episodes of actual homelessness (e.g., living on the streets, in cars, camping, squatting) and those who have doubled up. We were particularly interested in health, mental health, substance abuse, and family history correlates of lifetime homeless episodes. We use multivariate analyses to investigate correlates of ever having been homeless and to distinguish between correlates for doubling up and actual homeless episodes.

Method

Sample

The data used in this study are from Waves 1 through 5 of a longitudinal lagged sequential study currently underway on four Indigenous reservations in the Northern Midwest and four Canadian First Nation reserves. Several of the reserves were classified as remote in that they are considerable distances from even small towns and are accessed by non-paved roads. The reserves and reservations included in this sample share a common cultural tradition and language with minor regional variations in dialects. The sample is representative of one the largest Indigenous cultures in the U.S. and Canada. The long range purpose of the longitudinal study is to identify culturally specific resilience and risk factors that affect children’s well-being and to then use the information to guide the development of culturally-based interventions.

The project was designed in partnership with the participating reservations and reserves. Prior to the application’s funding, the research team was invited to work on these reservations/reserves and tribal resolutions were obtained and the researchers promised that participating reservations/reserves would be kept confidential in published reports. On each participating reservation/reserve, an advisory board was appointed by the tribal council. The advisory boards were responsible for advice on handling difficult personnel problems, advising on questionnaire development, reading reports for respectful writing, and assuring that published reports protected the identity of the respondents and the culture. Upon advisory board approval of the questionnaires, the study procedures and questionnaires were submitted for review by the university Institutional Review Board for approval.

All participating staff on the reservations were approved by the advisory board and are either tribal members or, in a few cases, non-tribal members who are spouses of tribal members. To ensure quality of data collection, all the interviewers underwent annual training for conducting computer-assisted personal interviewing and administering paper and pencil questionnaires. The training included practice interviews and feedback sessions regarding interview quality. In addition, all of the interviewers completed a required human subject’s protection training that emphasized the importance of confidentiality and taught procedures to maintain the confidentiality of data.

At the beginning of the project, each tribe provided us with a list of families of enrolled children aged 10–12 years who lived on or proximate to (within 50 miles) the reservation or reserve. We attempted to contact all families with a child within the specified age range. Families were recruited with a personal visit by an Indigenous interviewer at which time the project was explained to them. They were then presented with a traditional gift and invited to participate. If they agreed to be interviewed, each family member received $40 for their time when the interviews were completed. Study adolescents and at least one of their adult caretakers are interviewed once annually for the longitudinal project design. The recruitment procedure resulted in an overall response rate of 79.4%.

Interviews for Wave 1 began in the year 2002 and continued through 2003, with annual follow-up interviews completed thereafter. The interviews took place in the summer and fall of each year. The study had strong retention rates: 94.6% at Wave 2, 93% at Wave 3, 88.1% at Wave 4, and 90% at Wave 5. The sample for this analysis was made up of 873 parents/caretakers of adolescents aged 10 to 12 years at Wave 1. Non-Indigenous adults were excluded from the analyses. The 626 female caretakers ranged in age from 19 to 77 years with an average age of 38.4 years. The 247 male caretakers ranged in age from 21 to 68 years with an average age of 41.4 years. The majority (88.5%) of the caretakers lived on reservations or reserves. Approximately 9% of the caretakers lived on or near remote Canadian reserves.

Measures

A History of Homelessness was measured at Wave 1 with a single item, “has there been a time when you did not have a place to live.” Responses were coded 0,1 where a score of 1 indicated the respondent had ever been homeless. Follow-up questions were asked to determine, among those who reported a history of homelessness, where respondents stayed while they had no place to live. Near Homeless, as defined in the Wilder Study (Owen, Heineman, & Decker, 2007), includes those respondents who reported doubling up (i.e., staying with family or friends). Homeless Episode includes those who reported staying in hotels, motels, camps, cars, or shelters. Homelessness during the Study was measured by counting the number of times respondents stated that they had not had a place to live in the past 12 months. Only 2 respondents indicated that they had not had a place to live more than once, thus the resulting variable was dichotomized where a score of 1 indicates they had been homeless at some point between Waves 1 and 5.

Correlates of Homelessness

Health Problems were measured with yes/no questions where respondents indicated if they had had specific problems during the past year (see Table 2 for a full list of items). Financial Events were measured with sixteen yes/no questions where respondents indicated if they had experienced specific financial events in the past year (Table 3). Childhood History was measured with ten yes/no questions about respondents’ home environment while growing up (Table 4). Four questions asked respondents, while they were growing up, if anyone in their home had serious drinking problems, mental health problems, if individuals in their household were violent, and if their parents had marital problems. Six additional yes/no questions were included on alcohol, mental health, and emotional problems of their biological parents. Five Mental Health outcomes are measured with The University of Michigan Composite International Diagnostic Interview (UM-CIDI), including alcohol abuse, alcohol dependence, drug abuse, major depressive disorder, and generalized anxiety disorder (Table 5). The UM-CIDI is based o the Diagnostic and Statistical Manual-III-R (DSM-III-R) criteria and represents the University of Michigan revision of the CIDI used in the National Comorbidity Survey (NCS). The CIDI (WHO, 1990) from which the UM-CIDI is derived is a well-established diagnostic instrument that has shown excellent inter-rater reliability, test-retest reliability, and validity for the diagnoses that were used in this study. The UM-CIDI has been used extensively with non-clinically trained interviewers.

Table 2.

Percentage of People Experiencing Health Problems by Housing Situation

	Full Sample (n=873)		History of Homelessness(n=326)
	No History of Homelessness (n=530)	History of Homelessness (n=343)	Near Homelessness^ (n=214)	Homeless Episode^ (n=112)
Migraine	18.1%	26.8% **
Headaches			25.2%	28.6%
Asthma	6.6%	11.1% *	9.9%	14.3%
Underweight	1.9%	3.5%	3.3%	4.5%
Heart Disease	7.4%	2.9% **	3.3%	1.8%
Chest Pain	11.5%	14.6%	16.8%	12.5%
Overweight	44.3%	48.5%	46.7%	53.2%
Stomach Pain	15.1%	21.9% *	22.9%	22.5%
Back Pain	43.2%	53.6% **	50.5%	61.6% †
Eating Disorder	5.3%	7.6%	6.5%	9.8%
Weight Loss	11.2%	16.6% **	16.4%	17.9%
Weight Gain	34.5%	40.8% †	42.5%	39.3%

NOTE: Full Sample N varies slightly (964 – 969) based on item non-response; % are calculated by column n; The near and episode do not add up to 361 because of item non-response on the variables asking where people lived while they were homeless

^{^}Near Homeless is defined as doubling up with friends or relatives; Homeless Episode is defined as living in hotels, motels, camps, cars, or shelters

^†p<.10;

^*p<.05;

^**p<.01;

^***p<.001

Table 3.

Percentage of People Facing Financial Problems by Housing Situation

	Full Sample (n=873)		History of Homelessness(n=326)
	No History of Homelessness (n=530)	History of Homelessness (n=343)	Near Homeless^ (n=214)	Homeless Episode^ (n=112)
took a cut in wage or salary	15.7%	22.2% *	21.0%	25.2%
got laid off	12.1%	13.0%	11.8%	13.5%
suffered a financial loss in business, investments or property	7.8%	9.4%	8.5%	10.8%
lost some or all government benefits	8.9%	11.5%	10.0%	14.4%
got evicted from where you live	0.8%	3.5% **	0.9%	8.0% **
moved to a worse residence or neighborhood	2.3%	6.8% **	3.8%	10.8% *
had a car, furniture or other items repossessed	3.4%	3.8%	4.2%	2.7%
dipped heavily into family savings because of financial problems	10.0%	16.1% **	14.2%	17.9%
started receiving government assistance	27.7%	41.7% ***	39.3%	46.4%
took on financial responsibility for family member	20.4%	24.6%	22.5%	29.5%
had any other financial or employment problems	17.4%	27.0% ***	23.0%	34.2% *
changed residence to save money	3.0%	11.7% ***	10.7%	15.2%
reduced or eliminated insurance because of financial need	10.8%	19.9% ***	20.2%	22.5%
changed food shopping or eating habits to save money	17.7%	28.1% ***	23.0%	37.5% **
reduced driving the car to save money	14.6%	23.7% ***	20.7%	29.9% †
postponed medical or dental care to save money	8.9%	13.5% *	10.3%	20.9% **

NOTE: Full Sample N varies slightly (964 – 969) based on item non-response; % are calculated by column n; The near and episode do not add up to 361 because of item non-response on the variables asking where people lived while they were homeless

^{^}Near Homeless is defined as doubling up with friends or relatives; Homeless Episode is defined as living in hotels, motels, camps, cars, or shelters

^†p<.10;

^*p<.05;

^**p<.01;

^***p<.001

Table 4.

Percentage of People with Past Family Experiences by Housing Situation

	Full Sample (n=873)		History of Homelessness(n=326)
	No History of Homelessness (n=530)	History of Homelessness (n=343)	Near Homeless^ (n=214)	Homeless Episode^ (n=112)
While growing up…
did anyone in your home have serious drinking problem?	61.6%	72.9% ***	69.8%	82.7% *
did anyone in your home have mental health problem?	13.5%	21.0% **	21.7%	21.1%
was anyone in your family violent toward another family member?	49.0%	61.4% ***	57.7%	70.5% *
did your parents or people who raised you have serious marital problems?	36.0%	49.8% ***	47.1%	55.5%
Did your biological mother ever…
have drinking or alcohol problems	43.4%	62.9% ***	59.6%	70.6% †
get treatment for alcohol problems	29.9%	23.2%	21.1%	27.5%
get treated for mental or emotional problem	10.2%	12.3%	13.0%	11.9%
Did your biological father ever…
have drinking or alcohol problems	61.3%	74.0% ***	72.9%	78.9%
get treatment for alcohol problems	34.4%	34.8%	35.2%	35.4%
get treated for mental or emotional problem	6.2%	5.5%	5.0%	7.4%

NOTE: Full Sample N varies slightly (964 – 969) based on item non-response; % are calculated by column n; The near and episode do not add up to 361 because of item non-response on the variables asking where people lived while they were homeless

^{^}Near Homeless is defined as doubling up with friends or relatives; Homeless Episode is defined as living in hotels, motels, camps, cars, or shelters

^†p<.10;

^*p<.05;

^**p<.01;

^***p<.001

Table 5.

Percentage of People with Mental Health Diagnoses by Housing Situation

	Full Sample (n=873)		History of Homelessness(n=326)
	No History of Homelessness (n=530)	History of Homelessness (n=343)	Near Homeless^ (n=214)	Homeless Episode^ (n=112)
Lifetime
Alcohol Abuse	50.1%	49.1%	45.4%	59.5% *
Alcohol Dependence	17.5%	26.9% **	26.6%	27.9%
Drug Abuse	18.5%	29.0% ***	25.6%	35.1% †
Major Depression	13.2%	23.1% ***	20.3%	29.7% †
Generalized Anxiety Disorder	3.3%	6.3% *	5.3%	8.1%
12 Months
Alcohol Abuse	10.0%	12.3%	10.6%	14.4%
Alcohol Dependence	3.1%	5.1%	4.8%	5.4%
Drug Abuse	6.5%	10.5% *	9.7%	12.6%
Major Depression	5.3%	12.6% ***	12.1%	13.5%
Generalized Anxiety Disorder	2.2%	3.9%	3.4%	4.5%

NOTE: Full Sample N varies slightly (964 – 969) based on item non-response; % are calculated by column n; The near and episode do not add up to 361 because of item non-response on the variables asking where people lived while they were homeless

^{^}Near Homeless is defined as doubling up with friends or relatives; Homeless Episode is defined as living in hotels, motels, camps, cars, or shelters

^†p<.10;

^*p<.05;

^**p<.01;

^***p<.001

Predicting Homelessness

Female is a dichotomous item indicating if respondents are female (1) or male (0). Age is a continuous measure of respondent age in years. Remote is a dichotomous item indicating if respondents currently live in a remote location (1) or a rural location (0). Off Reservation is a dichotomous item indicating if respondents currently live off the reservation (1) or on the reservation (0). Employed Full-Time is a dichotomous item indicating if respondents are currently employed full time (1). Married is a dichotomous item indicating if respondents are currently married (1). Financial Events is a summed count of responses to the sixteen questions about financial difficulties described above. Alcohol Dependence, Drug Abuse, and Major Depression are taken from the CIDI measures described above.

Results

Histories of homeless episodes

Over one–third (39.3%) of our sample had experienced an episode of homelessness (Table 1). Women (41.5%) were significantly more likely to report a homeless episode than were men (33.6%). The majority (74%) had experienced homeless episodes while living on the reservation/reserve (73.2% males, 74.2% females) but 41.1% had been homeless when living off the reservation/reserve (41.8% males, 40.9% females). (Note that this is a duplicated count, respondents could have had episodes of homelessness both on and off reservation/reserve.) Almost twenty percent of respondents (17.7% males, 18.3% females, not shown on tables) stated that they had experienced episodes of homelessness when living both on and off the reservation/reserve.

Table 1.

Descriptives

	Total	Male	Female
Ever Homeless (n=873)	39.3	33.6	41.5 *
Homeless on Reservation (n=342)^	74.0	73.2	74.2
Live with Relatives or Friend (n=252)	94.8	93.3	95.3
Live in Hotel or Motel (n=232)	17.2	25.9	14.4 *
Live in Car or Camp Out (n=233	20.6	25.9	18.9
Live in Shelter (n=218	4.1	1.8	4.9
Homeless off Reservation (n=336)^	41.1	41.8	40.9
Live with Relatives or Friend (n=137)	88.3	90.6	87.6
Live in Hotel or Motel (n=125)	34.4	38.7	33.0
Live in Car or Camp Out (n=125)	19.2	35.5	13.8 **
Live in Shelter (n=120)	11.7	10.0	12.2

^{^}Homeless on Reservation and Homeless off Reservation are not mutually exclusive categories.

^†p<.10;

^*p<.05;

^**p<.01

Although 342 adults reported not having a place to live on the reservation, and 336 reported homelessness off the reservation, not all of them provided information on where they stayed while homeless. Consequently the numbers of adults who reported staying in various places do not come from the full sample of adults reporting homelessness. Almost all of the adults who were homeless on the reservation reported that they had lived with friends or relatives (93.3% males, 95.3% females). Men were significantly more likely to say that they stayed in a hotel or motel during a period of homelessness than were women (25.9% vs. 14.4%). Men (25.9%) also were more likely than women (18.9%) to have stayed in a car or camp out during episodes of homelessness on the reservation, although this difference was not statistically significant. Women were more likely to report staying in a shelter on the reservation (4.9% vs. 1.8%) but this difference was not statistically significant.

Similar to those who had experienced episodes of homelessness on the reservations/reserves, most of the adults who had been homeless while living off the reservation/reserves had stayed with friends or relatives (90.6% males, 87.6% females). About one–third (34.4%) had lived in hotels or motels (38.7% males, 33.0% females). Men (35.5%) were significantly more likely to have stayed in a car or camp while homeless off the reservation/reserve than were women (13.8%). As with those reporting homelessness on the reservation, a small proportion reported living in a shelter (10% of males, 12.2% of females).

Health Problems

Table 2 shows the proportion of respondents who answered yes to experiencing specific health problems in the past 12 months. Respondents with a history of homelessness were significantly more likely to report migraine headaches, asthma, stomach pain, and back pain than were those who had never been homeless. Those who had experienced homeless episodes also were more likely to report that they had lost (16.6% vs.11.2%) or gained weight (40.8% vs. 34.5%). Also compared to those who had never been homeless, individuals who had experienced homeless episodes were less likely to report heart disease (2.9% vs. 7.4%). Among those with a history of homelessness, the only statistically significant health difference between those who have doubled-up (50.5%) and those who had been actually homeless (61.6%) was back pain.

Financial Events

Table 3 shows the proportion of respondents who answered yes to experiencing specific financial problems in the past 12 months. Compared to the never homeless, respondents with a history of homelessness were statistically significantly more likely to have taken a cut in wage or salary, to have been evicted, moved to a worse residence or neighborhood, dipped heavily into family savings because of financial problems, started to receive government assistance, had other financial or employment problems, changed residence to save money, reduced or eliminated insurance, changed eating habits, reduced driving a car, or postponed medical or dental care to save money.

Adults who doubled-up reported experiencing fewer financial problems than those who reported an actual homeless episode. They were less likely to have been evicted, moved to a worse residence or neighborhood, to have had other financial or employment problems, changed eating habits, reduced driving the car to save money, or postponed medical or dental care than were those who had ever experienced a homeless episode.

Family History

Compared to the never homeless, those with a history of homelessness were significantly more likely to report negative family experiences while growing up (Table 4). Almost three–fourths (72.9% vs. 61.6%) reported someone in their home had a serious drinking problem and 21% (vs. 13.5%) reported someone in their home having mental health problems. Nearly two-thirds (61.4% vs. 49%) told us that when growing up a family member had witnessed family violence and one-half (49.8% vs. 36%) reported that their parents had marital problems.

Respondents who had been homeless were significantly more likely than the never homeless to report that their biological mother (62.9% vs. 43.4%) and biological father (74% vs. 61.3%) had alcohol problems. Biological mothers of the never homeless (29.9%) were more likely than those ever homeless (23.2%) to have gone through treatment for alcohol problems, but this difference was not statistically significant.

Compared to those who had doubled-up at some point, those who had experienced episodes of actual homelessness were more likely to have come from a home where someone had a serious drinking problem (82.7% vs. 69.8%), experienced family violence (70.5% vs. 57.7%) and to have had a biological mother with alcohol problems (70.6% vs. 59.6%).

Mental Health Problems

Respondents with a history of episodes of homelessness were significantly more likely to meet lifetime criteria for alcohol dependence (26.9% vs. 17.5%), drug abuse (29%, vs. 18.5%), major depressive episode (23.1% vs. 13.2%), and generalized anxiety disorder (6.3% vs. 3.3%) than those who had never been homeless (Table 5). The same pattern held for 12-month diagnoses. Those who had been homeless were more likely than the never homeless to meet criteria for past year drug abuse (10.5%, vs. 6.5%) and major depressive episode (12.6% vs. 5.3%). A larger proportion of adults with a history of homelessness met criteria for past year alcohol abuse, alcohol dependence, and generalized anxiety disorder, but the differences were not statistically significant.

Those who had an episode of homelessness were more likely than those who doubled up to meet criteria for lifetime alcohol abuse (59.5% vs. 45.4%), drug abuse (35.1% vs. 25.6%), and major depression (29.7% vs. 20.3%). There were no statistically significant differences in 12-month diagnoses between those who had been homeless and those who had doubled up.

Multivariate analyses

Correlates of a history of homeless episodes

We used multinomial logistic regression to investigate correlates of lifetime homeless episodes (Table 6). The first panel of results shows the predicted odds of experiencing near homelessness (i.e, doubling up) or an actual homeless episode relative to never experiencing homelessness. Being female, experiencing more financial events, and meeting criteria for alcohol dependence were each associated with increased odds of being “Near Homeless” compared to “Never Homeless”. Only marital status was associated with decreased odds of being “Near Homeless” compared to “Never Homeless.” Respondents who were married were 32% less likely than those who were unmarried to be near homeless, relative to never homeless. Being older and employed full-time were associated with decreased odds of belong in the “Homeless Episode” (actual homelessness) group than “Never Homeless”. Reporting more financial events, meeting criteria for drug abuse, and meeting criteria for major depression were each associated with increased odds of experiencing a homeless episode compared to never being homeless. Living off the reservation rather than on the reservation was associated with a 95% increase in the odds of being near homeless (O.R. = 1.95) and a 132% increase in the odds of experiencing an actual homeless episode (O.R. = 2.32), compared to never experiencing homelessness. This was the strongest predictor of both categories of homelessness.

Table 6.

Predicting History of Homelessness by Type of Homelessness (n=821)

	Never Homeless vs.						Near Homeless vs.
	Near Homeless			Homeless Episode			Homeless Episode
	B	O.R.	95% CI	B	O.R.	95% CI	B	O.R.	95% CI
Female	0.56 **	1.76	(1.16, 2.66)	−0.03	0.97	(0.59, 1.62)	−0.62 *	0.54	(0.30, 0.97)
Age	0.00	1.00	(0.98, 1.01)	−0.02 †	0.98	(0.95, 1.00)	−0.02	0.98	(0.96, 1.01)
Remote	0.37	1.45	(0.83, 2.55)	−0.90	0.41	(0.14, 1.21)	−1.21 *	0.30	(0.10, 0.93)
Off Reservation	0.67 *	1.95	(1.17, 3.24)	0.84 **	2.32	(1.24, 4.29)	0.07	1.07	(0.56, 2.06)
Employed Full Time	−0.02	0.98	(0.69, 1.40)	−0.68 **	0.51	(0.33, 0.80)	−0.65 *	0.52	(0.32, 0.85)
Married	−0.38 *	0.68	(0.47, 0.99)	−0.41	0.66	(0.41, 1.08)	0.02	1.02	(0.58, 1.78)
Financial Events	0.08 *	1.08	(1.00, 1.17)	0.14 **	1.15	(1.06, 1.26)	0.10 †	1.11	(0.99, 1.22)
Alcohol Dependence	0.44 *	1.55	(1.02, 2.36)	0.26	1.30	(0.76, 2.20)	−0.20	0.82	(0.46, 1.47)
Drug Abuse	0.35	1.42	(0.93, 2.16)	0.52 *	1.68	(1.02, 2.77)	0.16	1.17	(0.68, 2.01)
Major Depression	0.16	1.18	(0.74, 1.87)	0.63 *	1.88	(1.09, 3.23)	0.40	1.49	(0.81, 2.73)
Constant	−1.57 ***			−1.29 **			0.18

^{^}Near Homeless is defined as doubling up with friends or relatives; Homeless Episode is defined as living in hotels, motels, camps, cars, or shelters

^†p<.10;

^*p<.05;

^**p<.01;

p<.001

The second panel contains multinomial logistic regression results comparing experiencing an actual homeless episode to near homelessness (doubling up). Only financial events were positively associated with being in the “Homeless Episode” (i.e. actual homelessness) category. Each additional adverse financial event was associated with an 11% increase in the odds of experiencing a homeless episode compared to experiencing near homelessness. Being female, living in a remote location, and having full-time employment were each associated with decreased odds of being in the “Homeless Episode” category compared to “Near Homeless”. Marital status and mental health were unrelated to experiencing a homeless episode relative to near homelessness.

Episodes of homelessness between Waves 1 and 5

Only consistent caretakers (i.e. the same individuals interviewed during the course of the study) were included in the analyses. Forty-four women (7.8%) and eight men had experienced a homeless episode over the course of the study (i.e., in the past 12 months during Waves 1, 2, 3 or 5). Of the women who had been homeless after Wave 1 of the study, 56.8% had a history of at least one previous homeless episode (Table 7). Only 18.2% lived off the reservation/reserve and only 11.4% were from the remote Canadian reserves. About one-half (56.8%) of the women who became homeless were employed full time, and 15.9% were married. Almost one–fifth (18.2%) met lifetime criteria for alcohol dependence, 25% for drug abuse, and 18.2% for major depressive episode.

Table 7.

Descriptives of Homelessness over Time - Females Only (n=44)

	Mean
Homeless Ever	56.8%
Remote	11.4%
Off Reservation	18.2%
Employed Full Time	56.8%
Married	15.9%
Alcohol Dependence	18.2%
Drug Abuse	25.0%
Major Depression	18.2%
While growing up…
did anyone in your home have serious drinking problem?	84.1%
did anyone in your home have mental health problem?	29.3%
was anyone in your family violent toward another family member?	65.9%
did your parents or people who raised you have serious marital problems?	46.5%
Did your biological mother ever…
have drinking or alcohol problems	66.7%
get treatment for alcohol problems	32.0%
get treated for mental or emotional problem	21.6%
Did your biological father ever…
have drinking or alcohol problems	76.9%
get treatment for alcohol problems	30.0%
get treated for mental or emotional problem	5.6%

The family histories of those reporting homelessness over the course of the study were very similar to those homeless prior to the start of the study. Almost all (84.1%) came from homes where there were serious alcohol problems, and 65.9% had witnessed family violence. Two-thirds (66.7%) of their biological mothers and three-fourths (76.9%) of their biological fathers had alcohol problems and 21.6% of their mothers had been treated for a mental or emotional problem.

Discussion and Conclusions

This was not a random sample of reservation/reserve adults. Rather, the sample represents parents/caretakers of adolescents ages 10 to 12 years at Wave 1 of the study. This makes it all the more alarming that over one-third of the parents/caretakers had been homeless at some point in their lives. This number rose to four out of every ten among women. Moreover, eight percent (44 women) of the study mothers/female caretakers experienced an episode of homelessness between Waves 1 and 5 of the study. Most of the periods of homelessness were absorbed by family and friends, though about one in five of those who had been homeless had lived in a car or camped out. This number rose to almost one in four for off-reservation homeless episodes among men.

Individuals with a history of homelessness were worse off than the never homeless on nearly every indicator. They reported more health problems, more financial distress, and were more likely to meet lifetime criteria for alcohol dependence, drug abuse, major depressive episode, and generalized anxiety disorder. It is noteworthy that there were few health and financial differences between those who had doubled up and those who had been actually homeless. The major distinction was in mental health and substance use. Those who had been actually homeless were more likely than those who doubled up to meet lifetime criteria for alcohol abuse (without dependence), drug abuse, and major depressive disorder. Rates of alcohol dependence were almost identical between the two groups.

Individuals who had ever been homeless were more likely than the never homeless to have grown up in families where there were serious drinking problems, mental health problems and violence. They also were more likely than the never homeless to have a parent with serious mental health problems and/or alcohol problems. Those who had been actually homeless were more apt to have had childhood adversity than those who had doubled-up.

Multivariate analyses indicated that the parents/caretakers with a history of near homelessness (i.e., doubling-up with friends or family) were more likely those who had actually been without housing (i.e., “homeless episodes) to be female, live off the reservation or reserve, experience more financial difficulties, and to meet criteria for alcohol dependence, and less likely to be married. Those who had ever experienced a homeless episode (i.e., stayed in a camp, car, hotel, or shelter) were more likely to be younger, live off the reservation/reserve, experience more adverse financial events, and meet criteria for drug abuse or major depression. They were less likely to be employed full-time. Despite the higher prevalence rates of mental disorder and substance abuse disorders among the adults who had experienced episodes of actual compared to those experiencing near homelessness, neither the history of mental disorder or of a substance use disorder was statistically significant in distinguishing between the two groups.

The picture that emerges is one of cumulative and persistent disadvantage. Although we have no way to sequence events from these data, from a life course perspective what began as a disadvantaged childhood evolved into economic disadvantage, physical and mental health problems, and experiencing periods of homelessness. It may well be that the homeless episodes preceded the health and mental health problems, but it is very likely that homelessness occurred simultaneously or prior to economic problems. Financial difficulties may be both a cause and a consequence of homelessness. Regardless of the timing of events, the result is unrelenting disadvantage that manifests itself in current mental and physical health problems and current financial well-being.

Limitations

There are some important limitations to our current study. First, we are unable to determine exactly when the homeless episodes occurred. It may have been at any point in the respondents’ lifetime. Second, the respondents may be reporting on multiple homeless episodes. However, only 40 of the respondents of the 361 homelessness episodes did not have unique ID numbers. Even if the histories of homelessness included more than one episode of homelessness, this would be less than 12.5% of our sample, so this effect would be small. To correct for this potential clustering, we ran the model in MPLUS to adjust for standard errors and the results were unchanged. Third, there was no diagnostic data available on serious mental illness or schizophrenia. None of the parents/caretakers interviewed were judged to be psychotic at the time of their interviews (an exclusion criteria). This represents a sampling limitation in that those who experienced periods of homelessness in our study will not be similar to a sample made up of people who are currently homeless. However, one of the strengths of the study is the evidence of periods of homelessness among this sample of functioning parents/caretakers.

Along with these methodological issues there is the limitation that these data are from a single culture and cannot be generalized to other cultures. Also, the sample is made up of rural people who live on or near reservations and reserves and the results will not reflect the experiences or characteristics of urban Indigenous people. A sample made up of parents/caretakers of adolescents is both a limitation in terms of generalizing to all reservation/reserve adults and a strength in that it reflects the experiences of adults who are responsible for dependent children. Finally, there is the concern that diagnostic interviews may not be culturally appropriate. In this regard we used the same diagnostic interview schedules that have been used in major epidemiological studies of Indigenous people (e.g., Beals et al., 2005).

Conclusion

Homelessness is an important and often ignored risk factor in the lives of Indigenous people. Indeed, almost one-half of the mothers/female caretakers in our study had experienced an episode of homelessness in their lifetimes and about eight percent had been homeless during the first five years of the study. Importantly, our results suggest that reservation and reserve cultures quietly absorbs most homeless episodes, particularly those that occur on the reservation through making room for relatives and friends regardless of overcrowding (see also Owen et al., 2007). Our results are congruent with other studies that have shown that those who have ever been homeless manifest more physical health, mental health, and substance abuse problems suggesting that experiences of homelessness are related to more tenuous adjustment and that the experiences may be persistent (Kahn et al., 1992; Kasprov, 1998, Owens et al., 2007; Westerfelt & Yellow Bird, 1999). With the current data, it is impossible to discern the effects of parent/caretaker histories of homeless on the parenting of the study children. However, the associated mental health, substance use, and financial problems are likely to have a negative impact.

Although the culture is doing a remarkable job of informally taking care of their own, the consequences are overcrowding, possible substandard living conditions, and stress for all involved. Formal responses such as affordable housing, temporary shelters, and transitional living programs would be both more effective and exact less of a toll on everyone. Moreover, the correlates of homeless episodes suggest ongoing troubles that have resulted in persistent marginality. These characteristics may create tensions in overcrowding situations.

Because reservations/reserves appear to represent refuge in times of trouble, establishing formal services on reservation/reserves may be the most effective response. Prevention programs that provide comprehensive services for individuals and particularly families with children are an important step. Moreover, if those facing adversity tend to return “home,” support should be available to them. On many reservations/reserves there is simply not enough affordable housing to come home to. Programs that expand affordable housing and make transitional living programs and shelters available may be a highly effective first step. Interventions also should take the presence of children into account. Shelter programs that are welcoming to children, maintain educational opportunities, and provide day care for mothers who are working would ease the burden for those seeking education and employment. People who are in trouble seem to return “home” to the reservations/reserves. This should be taken into account in developing policy regarding homelessness among Indigenous people.

Biographies

Les B. Whitbeck, Ph.D. is the John G. Bruhn Professor of Sociology at the University of Nebraska-Lincoln. He graduated with his Ph.D. in sociology from Washington State University in 1986. He is principal investigator for the Midwest Longitudinal Study of Homeless Adolescents and is currently developing a program of research that focuses on homeless women and children. He is also principal investigator for a longitudinal diagnostic study of American Indian children aged 10–12 years. The project will follow the children and their caretakers as the children move through their high school years.

Devan M. Crawford, M.A. is the Director of Data Analysis for a longitudinal diagnostic study of American Indian children aged 10–12 years and the program director for a study of health, mental health, and HIV risk among homeless women at the University of Nebraska-Lincoln.

Kelley J. Sittner Hartshorn, M.A. is a doctoral student in the Department of Sociology at the University of Nebraska-Lincoln, and a research assistant on a longitudinal diagnostic study of American Indian children aged 10–12 years.