Physical and Psychological Aggression towards a Child among Homeless, Doubled-up, and Other Low-income Families

Abstract

This study examines the extent of adverse parenting behaviors among low-income families with children and determines whether housing instability, measured by homelessness and doubling up with relatives or friends due to economic hardship, increases the likelihood of physical and psychological aggression towards a child, after considering the contribution of other relevant characteristics. Using data from 3 waves of the Fragile Families and Child Wellbeing Study, this study followed 2,332 low-income children in 20 large U.S. cities. Multivariate analyses involved logistic regression with generalized estimating equations. Adverse parenting behaviors were common among all low-income families regardless of their having experienced housing instability. Nonetheless, mothers with a homeless or doubled-up episode reported higher rates of physically and psychologically aggressive behaviors towards a child compared to the housed group. Having a homeless episode was significantly associated with a greater likelihood of reporting a high level of physical aggression towards a child. Child’s behavioral issues, maternal depression, and parental stress also contributed to adverse parenting behaviors. Findings suggest that housing instability can be a marker of adverse parenting behaviors and service professionals need to respond to parenting needs as well as housing needs for families in unstable housing. Areas of future research were discussed.

Introduction

Housing instability, particularly homelessness, has been reported to have a negative effect on parenting behaviors (Bassuk & Rubin, 1987; Boxil & Beaty, 1990; Hausman & Hammen, 1993; David, Gleber, & Suchman, 2012; Perlman, Cowan, Gerwitz, Haskett, & Stokes, 2012). However, there is ambiguity on whether and to what extent housing instability, independent of other factors common among families in poverty, contributes to adverse parenting behaviors. It is also unclear whether families who are homeless differ in their parenting behaviors from those unstably housed but are not literally homeless, including those doubled up with relatives or friends without formal rental agreements. Building on prior research, this study examined the extent of physical and psychological aggression towards a child among homeless, doubled-up, and other low-income families drawn from 20 large U.S. cities and investigated the role of homeless and doubled-up episodes on adverse parenting behaviors after controlling for other poverty-related factors. In doing so, this study sought to contribute to enhancing intervention strategies to improve parenting behaviors and parent-child interaction quality in these vulnerable populations.

Background

This study built on an ecological model of the determinants of parenting (Belsky, 1993; Bronfenbrenner, 1986) for understanding factors that contribute to a parent’s capacity to provide adequate care and protection for a child. As multiple stress-producing factors contribute to parenting outcomes, risk factors are organized by ecological levels. The first level includes individual factors related to the child (age, gender, whether child has a disability or behavioral problem) and parent (age, marital status, mental illness or substance use, immigrant status). Risk factors at the meso level include dynamics and interactions within the family, such as parenting stress, parent-child relationships and family conflict. Risk factors at the exo level include the larger support network around the family, socio-economic status, social isolation and community characteristics. Homelessness, doubling up, financial stress, and unemployment are all stress factors at this level.

Prior research suggests that families experiencing housing instability are at greater risk for adverse parenting behaviors. Children in homeless or doubled-up families tend to have a higher rate of emotional and behavioral problems (Park, Fertig, & Allison, 2011). Parents who experience housing instability are more likely than poor but stably housed adults to be depressed or to use substances, factors that are both associated with aggressive behaviors towards a child (Banyard & Graham-Bermann, 1998; Fertig & Reingold, 2008; LaVesser, Smith, & Bradford, 1997; Meadows-Oliver, Sadler, Swartz, & Ryan-Krause, 2007; Wenzel et al., 2004). The physical and psychological stressors that accompany unstable housing are likely to compromise a parent’s ability to address their children’s basic emotional and physical needs. Difficulties with fitting into a new residential environment (e.g., shelter, and other’s house), lack of privacy, and disconnection from neighbors and supportive networks may further undermine parenting capacity and strain relationships between parents and children (Hausman & Hammen, 1993; Schindler & Coley, 2007; David, Gelber & Suchman, 2012; Park, Metraux, Broadbar, & Culhane, 2004; Perlman et al., 2012). This is especially concerning as healthy child development requires attentive and responsive caregiving (David, Gelber, & Suchman, 2012).

Although studies have examined parenting outcomes in homeless families, little is known about families who are unstably housed but do not live in shelters. One such group includes individuals who are “doubled-up”, i.e. living with friends or family without resources to obtain their own housing. Doubled-up families are not counted as the homeless by the U.S. Department of Housing and Urban Development (HUD)’s housing assistance programs. Yet, doubled-up living situations are often related to overcrowding, incompatibility, or financial strain (Evans, 2005; Leventhal & Newman 2010; Solari & Mare 2012) to the point where such families can meet the criteria for homelessness (National Alliance to End Homelessness, 2012). Doubled-up families are counted as homeless by several other federal homeless assistance programs, including the Education for Homeless Children and Youth program. Given that doubling up is another important measure of precarious housing status and many doubled-up families strive to remain housed, it is important to understand whether homeless and doubled-up families are similar in their parenting behaviors.

Empirical studies that have compared parenting behaviors of low-income housed and homeless families have produced mixed results. Homeless mothers provided less learning stimulation and less warmth and affection for their children than did housed mothers (Koblinsky, Morgan, & Anderson, 1997). Other studies of low-income families have reported few significant main effects of homelessness on attachment security, caregiver-child interactions such as caregiver’s use of discipline, and developmental stimulation (Easterbrook & Graham, 1999; Howard et al., 2009).

While prior research has added to our knowledge about homelessness and its potential relationship with parenting outcomes, it has several limitations. First, it is unclear whether homelessness independent of other poverty-related risk factors is a predictor of physical and psychological aggression toward a child. Secondly, while studies have looked at parenting stress, caregiver’s use of discipline, attachment, and developmental stimulation among homeless families, limited information is available about aggressive parenting behaviors among those unstably housed but do not live in shelters. Lastly, previous studies were typically based on a sample from a single geographic area. In addition, variations in the age of the sample and settings from which the sample was extracted (e.g., shelters, schools) make it difficult to compare and generalize the findings.

Drawing on longitudinal data with rich information on a variety of factors at multiple ecological levels, this study focused on the contribution of housing instability to adverse parenting behaviors independent of other characteristics. The study is unique in that it includes both homeless and doubled-up families along with a comparison group of other low-income families. Additionally, the sample was recruited in 20 large U.S. cities, allowing for greater generalizability of findings.

This study addressed two main research questions:

1. What is the extent of adverse parenting behaviors, measured by physical and psychological aggression towards a child, among homeless, doubled-up, and other low-income families?

2. To what extent does housing instability, measured by homelessness and doubling up, contribute to physical and psychological aggression towards a child, after controlling for other relevant characteristics?

Method

Data and Sample

This study used data from the Fragile Families and Child Wellbeing Study (FFS). The FFS provides longitudinal information about a cohort of nearly 5,000 children born in 20 large US cities between 1998 and 2000 by surveying mothers at birth and again when the child was age 1, 3, 5, and 9 years.¹ Births to unmarried mothers were oversampled, and the extensive questionnaires administered to the mothers in the follow-up waves included questions on housing as well as a rich set of covariates. Of the original sample, 90 percent of the mothers responded to the 1-year follow-up, 88 percent to the 3-year follow-up, 87 percent to the 5-year follow-up, and 76 percent to the 9-year follow-up. An In-Home Longitudinal Study, which is a collaborative project following up the FFS sample from the 3-year interview, included a parental interview that asked detailed questions about parenting behaviors. Seventy-five percent of respondents of the 3-year core survey participated in the 3-year In-Home Study, 74 percent of respondents of the 5-year core survey participated in the 5-year In-Home Study and, 93 percent of respondents of the 9-year core survey participated in the 9-year In-Home Study.

Because the In-Home Study that collected information on parenting behaviors started when the 3-year interview was conducted, the analyses in this study focused on the 3, 5, and 9-year waves. The sample for this study was consisted of households that reported having income at or below the federal poverty threshold at one or more of the three waves. We excluded households with missing information on housing status or poverty status in all three waves (n=313). Our final analysis sample included 2,332 households.

Measures

Dependent Variables: Physical and psychological aggression towards a child

Three aspects of adverse parenting behaviors were examined based on maternal self-report: physical aggression towards a child and psychological aggression towards a child. These variables were based on questions from the Parent-Child Conflict Tactics Scales, a parenting measure with documented validity. Physical aggression included the following behaviors: shaking the child; hitting on the bottom with an object; spanking on the bottom with a bare hand; slapping on the hand, arm or leg; and pinching the child. Psychological aggression included shouting, swearing, threatening to send away, threatening to spank but not doing it, and calling the child a name, such as dumb, lazy. Each variable was based on five questions.

For each of the physically and psychologically aggressive behaviors, mothers responded using the following choices: a) this has never happened, b) yes, but not in the past year, c) once, d) twice, e) 3–5 times, f) 6–10 times, g) 11–20 times, or h) more than 20 times in the past year. If a mother had not been physically aggressive, then the number of physical aggression events was 0. If she had been physically aggressive, then we summed up the number of times a mother engaged in the behaviors using the middle value in the category (or 20 for the final category) across the five questions. The minimum value for each parenting behavior count measure was 0 and the maximum was 100. Mothers with a sum score more than one standard deviation above the mean on physical aggression and psychological aggression were coded as exhibiting a high level of adverse behavior on the dichotomous scales.

Independent Variable: Housing Instability

Homelessness was identified as to whether, at the time of the interview, the mother indicated that the household lived in temporary housing, in a group shelter, or on the street; or that, in the 12 months prior to the interview, they stayed in a shelter, an abandoned building, an automobile, or any other place not meant for regular housing, even for one night. Doubling-up was defined as living with relatives or friends or living in a house that is owned by relatives and for which the household does not pay rent. A household was also considered to be living in a doubled up arrangement if, in the past 12 months, the household moved in with other people because of financial problems. To ensure mutually exclusive groups, a household that reported being both homeless and doubled-up at the same interview was categorized as being homeless only.

Control Variables

Child factors included age, gender, physical disability, health status, and behavior problems. A dichotomous measure of child’s overall health status was created as to whether a child’s health was rated as fair/poor or good/excellent. At the 3-year interview, mothers were also asked if their child had a physical disability. At the 5-year and 9-year interviews, the survey asked a series of disability-related questions. We identified children as having a physical disability if the mother reported that their child had down syndrome, cerebral palsy, sickle cell anemia, a heart condition, blindness, deafness, or a problem with limbs. Finally, behavior problems were assessed using a series of questions from the Child Behavior Checklist (CBCL) for internalizing and externalizing problems. We created indicators for whether the child had CBCL scores in the clinical range (T≥64) on the Internalizing and Externalizing scales. Children with high internalizing scores exhibited signs of being anxious, depressed or withdrawn, whereas children with high externalizing scores exhibited aggressive or destructive behaviors.

Parent factors included socio-demographic characteristics, such as the mother’s race and ethnicity, immigrant status, educational level, number of children, and current marital and cohabitating status with the child’s father. We also added three measures of mother’s health and health behavior. Mother’s overall health status was measured by a dichotomous measure of whether her health is fair/poor or good/excellent. We also have a measure of whether the mother has a serious health condition that limits the amount or kind of work she can do. Depression was assessed using the Composite International Diagnostic Interview – Short Form (CIDI-SF). Mothers who reported feeling depressed with high frequency over a two-week period and who had two or more of the seven symptoms were considered to meet the diagnostic criteria for major depression (Walters et al., 2002). Finally, a mother was considered to have a drug or alcohol problem if she responded that drinking or drugs interfered with her work on a job or with her personal relationships in the last 12 months; if she sought treatment for drug or alcohol problems in the last 12 months, or if she used “hard” drugs (e.g. cocaine, crack, speed, LSD, heroin) recently.

Variables at the meso level included parenting stress, partner support, and exposure to domestic violence. Parenting stress was assessed during the in-home interview adapting 12 items from the Parenting Stress Index. The questions asked about the extent to which a parent felt trapped by parenthood, felt she sacrificed too much for her children, and to which extent she felt her children were the cause of problems. Mothers responded on a 5-point Likert scale ranging from 0 for strongly disagree to 4 for strongly agree. The average score for all twelve questions was used as the parental stress score. Partner support was created from 6 questions asking the mother about whether her current partner was willing to compromise, expressed affection, insulted, encouraged, listened, and understood her. We reversed the codes of the insults question and created a score by summing the responses of often or sometimes. The overall score for partner support ranged from 0 to 6, with a high scoring indicating good support and a low score indicating poor or negative support. Exposure to domestic violence was measured by questions of whether the mother was slapped or hit by the father of her child in the last 12 months and whether she was seriously hurt in a fight with the father in the last 12 months.

Key variables at the exo level included housing instability, household income, receipt of welfare/food stamps benefit, employment status, and neighborhood characteristics. Neighborhood characteristics, based on census tract-level socioeconomic and demographic characteristics, included the percentage of non-Hispanic Black, the percentage of foreign born residents, the percent of adults with a high school degree, and the poverty rate.

Data Analysis

We first conducted descriptive analysis to capture the sample characteristics We then used logistic regression with generalized estimating equations (GEE) on our pooled person-year sample to estimate the effect of homelessness, doubling-up and other explanatory variables on adverse parenting behaviors. GEE method accommodates correlated data that would otherwise violate assumptions of independence among variables, especially correlations across time as are present in this dataset (Allison, 1999).

Time-constant variables were obtained from the baseline and the community-level variables came from the 2000 Census. The variance inflation factor scores suggested no multicollinearity between variables in our specification.

Most of the data field had missing values of less than 5 percent of the observations. In order to handle missing data and maximize our precision, we used multiple imputation (with Stata’s MI commands). The results were estimated using 5 imputed data sets. All data analyses were performed using Stata statistical software, release 12.1 (StataCorp LP, 2011).

The institutional review boards of the University of Illinois and the University of Georgia approved this study before data analyses were conducted.

Results

Sample characteristics by housing instability

As Table 1 shows, 8% of the sample reported having experienced a homeless episode and additional 21% experienced a doubling-up episode during the 6-year observation period covered by this analysis. A significantly higher percentage of mothers with homeless episodes reported that their child has a physical disability and externalizing behavior issues.

Table 1.

Sample characteristics (N= 2,332)

	Ever Homeless (n=190)	Ever Doubled-Up (n=490)	Low-Income, Housed (n=1,652)	Total (n=2,332)
Individual Level: Child Factors
Child is female	53%	47%	48%	49%
Child has a physical disability	6%*	4%	3%	3%
Child’s health as fair/poor last wave	3%	3%	3%	3%
CBCL internalizing score in the clinical range	25%	23%	19%	20%
CBCL externalizing score in the clinical range	15%**	10%	8%	9%
Individual Level: Parent Factors
Black	66%**	55%	55%	56%
Hispanic	23%*	29%	31%	30%
Mother is immigrant	8%**	9%***	17%	15%
Mother w/HS+ education	17%	17%	19%	18%
Number of children in household	2.5	2.5	2.6	2.6
Mother’s age at birth	23.4	23.0***	24.3	23.9
Mother was married to child’s father	13%*	15%*	19%	18%
Mother was cohabiting with child father	26%***	31%***	43%	39%
Mother has a health condition that limits ability to work	17%**	12%	10%	11%
Mother has depression	33%***	33%***	21%	24%
Mother has drug or alcohol problem	6%***	3%*	1%	2%
Meso Level Variables
Parenting stress	1.4***	1.3***	1.2	1.2
Partner support scale (range 0–6)	3.6	3.6	3.9	3.8
Mother experienced domestic violence	22%***	15%**	10%	11.9%
Exo Level Variables
Homeless Episode last wave	50%***	0%	0%	4%
Doubled-up Episode last wave	10%***	61%***	0%	14%
Household income	$15,672	$14,358***	$17,633	$16,785
Mother received welfare in last 12 months	73%***	67%***	58%	61%
Mother was working	32%**	38%	42%	41%
% non-Hispanic Black in census tract	53%**	45%	45%	46%
% foreign-born residents in census tract	13%	12%**	14%	13%
% 25+ years w/HS+ education in census tract	66%	68%*	66%	66%
% families below poverty level in census tract	24%	20%**	22%	22%

Notes:

^*p<0.05;

^**p<0.01;

^***p<0.001. (Reference group: Low-income, housed mothers)

A comparison of the socioeconomic characteristics, health and behavioral issues, and the level of parenting stress, for the homeless, doubled-up, and low-income, housed groups indicates that homeless families were the most disadvantaged, followed by the doubled-up group.

Physical and psychological aggression towards a child among homeless, doubled-up, and other low-income families

As reported in Table 2, mothers in the homeless and doubled-up groups were more likely than other low-income, housed mothers to report a high level of physically aggressive behaviors toward a child. For instance, at 3 years, 29% and 18% of homeless and doubled-up mothers, respectively, reported a high level of physical aggression compared to 13% in the housed group. At 5 years, mothers who reported a high level of physical aggression were 25% in the homeless group and 14% in the housed group. The frequency of physically aggressive behaviors decreased in all groups as the children aged.

Table 2.

Physical and psychological aggression towards a child at respected years by homeless and doubled-up episodes (N=2,332)

	3 years			5 years			9 years
	H	D-U	L	H	D-U	L	H	D-U	L
Physical aggression (1 SD above mean)	29%***	18%*	13%	25%**	18%	14%	12%	17%*	11%
No. physical aggression in last year (0–100)	16.8*	14.6	12.7	17.3**	13.8*	11.3	7.0	8.0***	5.1
Psychological aggression (1 SD above mean)	27%	20%	19%	39%**	26%	22%	31%*	23%*	16%
No. psychological aggression in last year (0–100)	22.1	22.2	20.5	27.5*	26.0*	22.4	20.6**	18.3**	14.2

Note. H = Homeless group; D-U = Doubled-up group; L = Low-income, housed group.

^*significantly different from the corresponding low-income housed group at p<.05

^**significantly different from the corresponding low-income housed group at p<.01

^***significantly different from the corresponding low-income housed group at p<.001

Psychological aggression towards a child had a higher frequency than physical aggression among the sampled mothers. A greater proportion of the homeless and doubled-up groups than other low-income, housed mothers reported a high level of psychologically aggressive behaviors at the 5-year and the 9-year interviews (e.g., 39% of homeless mothers vs. 22% of housed mothers at 5 years).

Contribution of homelessness and doubling up to adverse parenting behaviors

Table 3 presents the results of logistic regression with GEE for the dichotomous outcome variables.

Table 3.

Results of logistic regression analyses with GEE (N=3,803)

	High level of Physical Aggression	High level of Psychological Aggression
	Odds Ratio	Odds Ratio
Housing Instability
Homeless Episode last wave	1.86**	1.17
Doubled-up Episode last wave	1.10	0.86
Individual Level: Child Factors
Child is female	0.81	0.83
Child has a physical disability	1.07	0.98
Child’s health as fair/poor last wave	1.16	0.58
CBCL internalizing score in the clinical range	1.15	1.25
CBCL externalizing score in the clinical range	2.11***	2.35***
Individual Level: Parent Factors
Black	1.43	1.21
Hispanic	0.76	0.76
Mother is immigrant	0.79	0.46***
Mother w/HS+ education	1.40	1.13
Number of children in household	0.98	0.99
Mother’s age at birth	0.94***	0.97**
Mother was married to child’s father	1.19	1.01
Mother was cohabiting with child father	0.95	0.97
Mother has a health condition that limits ability to work	0.89	0.95
Mother has depression	1.08	1.39**
Mother has drug or alcohol problem	1.32	1.36
Meso Level Variables
Parenting stress	1.37***	1.31***
Partner support scale (range 0–6)	0.98	0.93**
Mother experienced domestic violence	0.97	1.17
Exo Level Variables
Household income (log)	1.08	1.07
Mother received welfare in last 12 months	1.16	1.18
Mother was working	1.24	1.18
% non-Hispanic Black in census tract	1.00	1.00
% foreign-born residents in census tract	0.99*	1.00
% 25+ years w/HS+ education in census tract	0.99	0.99
% families below poverty level in census tract	0.98**	0.99
5 year interview wave	0.80*	0.88
9 year interview wave	0.54***	0.81*
Person-year observations	3,803	3,803
Number of mothers	1,859	1,859

Notes:

^*p<0.05;

^**p<0.01;

^***p<0.001.

The odds that a mother engaged in a high level of physical aggression were nearly 1.9 times greater for mothers with a homeless episode than other low-income, housed mothers. Doubling up was not significantly associated with adverse parenting behaviors.

Of child factors, having a CBCL externalizing score in the clinical range was a big contributor to both physical and psychological aggression, increasing the odds more than two times.

Of parent factors, maternal depression increased the odds of psychologically aggressive behaviors towards a child by 39%. Mothers who are an immigrant had a significantly lower likelihood to report a high level of psychological aggression. Giving birth at an older age was also associated with the odds of a high level of physical and psychological aggression.

Parenting stress was associated with a significantly greater likelihood for adverse parenting behaviors, increasing the odds of physical aggression by 37% and those of psychological aggression by 31%. On the other hand, more partner support contributed to lower odds of psychologically aggressive behaviors.

Neighborhood characteristics, such as percentages of foreign-born residents and households below the poverty threshold, were significantly associated with the odds of parent’s aggressive behaviors toward a child. Yet, the magnitude of those coefficients was quite small.

The 5 year and 9 year interview wave dummies were associated with decreased odds of aggressive behaviors, indicating that as children get older, the likelihood of physical and psychological aggression declined.

Discussion

Overall, mothers with a homeless or doubled-up episode reported a higher frequency of physical and psychological aggression towards a child compared to those in the low-income, housed group. While the frequency of physical aggression decreased in all three groups as the children aged, that of psychological aggression increased or remained at similar levels over time. Homeless families were the most disadvantaged in terms of their socioeconomic characteristics, health, behavioral issues, and parenting stress, followed by the doubled-up group, and then by the low-income housed group. The pattern was similar in the extent of physical and psychological aggression.

Having had a homeless episode was significantly associated with a greater likelihood of physically aggressive behaviors towards a child (OR=1.86, p<.01). It suggests that difficulties related to homelessness, including the individual and interpersonal strain as well as housing instability, can be detrimental to relationships between parents and children and may magnify adverse parenting behaviors. This finding is in line with the family stress model, which posits that economic strain increases parental emotional distress and results in harsher parenting practices (McLoyd, 1990).

Families experiencing homelessness seem to be at greater risk for adverse parenting behaviors than those in other unstable housing status. Doubled-up living situations can be tenuous and stressful, but their independent effect on adverse parenting was not demonstrated in this study.

The study also identified other factors that contributed to adverse parenting behaviors. Similar to findings in the literature, child behavior problems and maternal depression contributed to adverse parenting behaviors (Kim et al., 2010; Kohl, Kagotho, & Dixon, 2011). These findings suggest that mothers and children with mental health problems consist of a high risk group for less optimal parenting and they may benefit from parenting-related support in addition to mental health services.

Immigrants were less likely than non-immigrant families to report psychologically aggressive behaviors toward a child. It is unclear whether immigrant families were less likely to report this form of adverse behavior or they engage less in such forms of behavior. More research is needed on this topic.

Parenting stress was higher among homeless and doubled-up mothers and it was a consistent and strong predictor of physical and psychological aggression towards a child. This is consistent with prior literature showing that economic hardships are associated with increased parenting stress that compromises both parenting practices and the quality of parent-child interactions (Belsky, Woodworth, & Crnic, 1996; Calkins, Hungerford, & Dedmon, 2004; Crnic, Gaze, & Hoffman, 2005). These findings suggest that the role of homelessness on adverse parenting behaviors can be better understood in its relation to the extent of parenting stress a mother perceives.

Partner support functioned as a protective factor for adverse parenting. A level of psychologically aggressive behaviors was lower if a partner had expressed more affection, encouraged, listened and understood the mothers.

This study has several limitations. First, the study relied on mothers’ reports of adverse parenting behaviors. In future research, it would be important to use multiple sources of data on parenting, including direct observations of mother-child interactions. Nonetheless, the extent of adverse parenting behaviors among low-income families is alarming and it underscores the high degree of strain low-income mothers were living under and the multitude of difficulties they were dealing with in parenting their children. Second, the data for this study did not have information on the severity and frequency of housing instability episodes, and we were unable to assess how these factors contributed to adverse parenting behaviors. It is possible that more frequent, prolonged experiences of housing instability have a different impact on adverse parenting. Lastly, this study did not determine whether homeless families were different in adverse parenting from other low-income families prior to homeless episodes and, if so, when such differences have emerged. Further research addressing these elements is needed to enhance our understanding of the relationship between housing instability and parenting behaviors.

Despite these limitations, this longitudinal study disentangled the role of housing instability and other poverty-related characteristics on adverse parenting behaviors, and it points out the importance of addressing parenting needs of low-income mothers and to target their abilities to cope and deal with stress. Higher rates of adverse parenting behaviors among mothers with a homeless or doubled-up episode suggest that housing instability is an important marker of adverse parenting behaviors and that social service professionals need to identify and respond to parenting needs as well as housing needs for families in unstable housing. In addition, concrete support (e.g., housing assistance) is important in helping homeless families stay afloat, but providing support that bolsters a mother’s own ability to cope with stress and that meets her own unique emotional and behavioral needs may play an important protective role against adverse parenting behaviors. Intervening to identify and address child behavior problems and bolstering a partner’s ability to provide support may also be important.