Mixed-method and experimental data on working poor families and children with troubles participating in the New Hope anti-poverty experimental initiative in Milwaukee are described. Sixty percent of these families had at least one child who had significant problems (learning, school achievement and/or behavior, home behavior, retardation, other disabilities). Control group families with children who had troubles had more difficulties in sustaining their family routine than did New Hope experimental families. In the context of the many other challenges these parents face, adaptation to children with troubles does not stand out as sharply compared to middle-class European American families. There is less family adaptation specifically due to, or in response to, the troubled child, and more adaptation to the struggles of making ends meet.
There is ample evidence in the research literature that children with developmental and behavioral problems can make the daily routine of family life difficult. Such pressures may be translated into increased parental workloads, decisions to forego a career or to stay in a job for reasons of insurance or convenience, curtailed family activities, intensive information-seeking about educational or health services, increased responsibilities placed on siblings, reliance on extended family networks, problems locating appropriate childcare, serious financial burdens, and others. All of these effects have been documented in studies of families with children who have varying disabilities (Barnett & Boyce, 1995; Beresford, 1994; Brotherson & Goldstein, 1992; Freedman, Litchfield, & Warfield, 1995; Gallimore, Bernheimer, & Weisner, 1999; Gallimore, Weisner, Kaufman, & Bernheimer, 1989; Harris & McHale, 1989; Wishart, Bidder, & Gray, 1981). However, much of the existing research has tended to focus on middle socioeconomic status (SES) families, and, until very recently, European Americans. There is much less literature about the impact of children with problems on the daily lives of low-income families.
Just making ends meet is a significant achievement for many working poor families quite apart from having a child with troubles or disabilities. Often, despite full-time work, these parents live near or below poverty and have jobs that do not offer security, benefits, or flexibility in hours (Burton, Allison, & Obeidallah, 1995; Burton, Obeidallah, & Allison, 1996; Edin & Lein, 1997; Friedlander & Burtless, 1995; Handler, 1995; Holloway, Fuller, Rambaud, & Eggers-Pierola, 1997; Jencks & Peterson, 1991; Newman, 1999; Rosier & Corsaro, 1993). Many low-income families lack adequate transportation to commute to work or to provide school or day care for their children (Olson & Pavetti, 1996; Sherman, Amey, Duffield, Ebb, & Weinstein, 1998). A high proportion are likely not to have stable partners in their lives to assist in child care or in bringing in income (Garfinkel & McLanahan, 1995; Quadagno, 1994; Sherman et al., 1998; Wertheimer, 1999). Working poor parents face problems with drugs or alcohol or mental health at a greater rate than do other families (Olson & Pavetti, 1996; Schmidt, Weisner, & Wiley, 1998). Reliance on extended family and other economically poor parents can be of great value, but also a house of cards that can and does unexpectedly collapse at times (Danziger et al., 2000; Newman, 1999).
Although the overall prevalence of chronic conditions in children and the variation in different sociodemographic groups are influenced by the definitions of those conditions chosen (Stein & Silver, 1999), the existing data indicate a high incidence of child disabilities and health problems in low-income families (Birenbaum, 2002; Duncan, Brooks-Gunn, & Klebanov, 1994; Fujiura & Yamaki, 2000; Moore & Driscoll, 1997; Newacheck & Halfon, 1998; Olson & Pavetti, 1996; Vandivere, Moore, & Brown, 2000). For example, poor children are 1.3 times as likely to evidence developmental delay as children who are not poor (Brooks-Gunn & Duncan, 1997). Luprest and Acs (1996) reviewed the literature on families receiving Aid for Dependent Children (AFDC) and reported that 11% to 16% had children who were limited in the kind or amount of typical childhood activity. Meyers, Brady, and Seto (2000) estimated that between 20% to 25% of all families receiving welfare support have either a child or parent (11% to 16% of children, 16% to 19% of mothers) with significant disabilities and impairments.
The associations between poverty and disability appear to extend to other indicators as well. Zill, Moore, Smith, Stief, and Coiro (1995) reported that poor children were significantly less healthy, more than twice as likely to fail in school, and more likely to present serious conduct and discipline problems than children who are not poor, regardless of whether or not their families received AFDC support. These findings are supported by those of Takeuchi, Williams, and Adair (1991): Children from families under persistent financial stress have significantly higher mean scores on antisocial behavior and depressive symptoms than do those whose families are under no financial stress or who perceived financial difficulties only at one time.
Whether or not there are children at home with problems, poor mothers are now facing new work requirements. The legislation resulting from passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) in 1996 put time limits on the receipt of cash assistance and mandated states to move recipients off of state supports to work. Most states are still in the process of setting definitive policies related to disability issues in welfare reform, and there is a dearth of information on the true incidence, demographics, or familial concerns in this subset of the welfare population (LeRoy, Harrison, & Johnson, 1999). Based on the evidence from previous research, however, low-income families with children who have problems are likely to have additional significant barriers to obtaining and maintaining employment (Acs & Loprest, 1994; Wolfe & Hill, 1995). Brandon and Hogan (2002) showed, for example, using the nationally representative 1996 Survey of Income and Program Participation, that the likelihood of exit of a family from AFDC supports is reduced as much by having a child with a disability in the household as it is for a mother's disability.
The families in our study were all part of New Hope, an antipoverty experiment aimed at moving welfare applicants to work and greater self-sufficiency. New Hope was based in Milwaukee, Wisconsin, and active between 1994 and 1998 (Bos et al., 1999). Families targeted by New Hope had to meet four eligibility criteria: (a) have lived in one of the two targeted neighborhoods in Milwaukee, (b) been older than 18, (c) had an income at or below 150% of the poverty line, and (d) been willing to work 30 or more hours a week. Those who volunteered for the program were randomly assigned either to New Hope or to a control group. The New Hope program offered a package of benefits to eligible participants: (a) a wage supplement (to ensure that income remained above the poverty threshold for their family), (b) subsidies for affordable health insurance, (c) child care vouchers, and (d) a full-time community service job opportunity for those unable to find work on their own. Members of control and experimental groups were free to seek out any federal or state public assistance programs, but individuals in the experimental program also had access to New Hope benefits.
To understand the effects of a child with troubles on the family, we conducted an ethnographic study with a stratified random sample of the New Hope participants. The focus was on the family's daily routine and activities, how families were able to sustain their daily routine, whether the New Hope intervention impacted sustainability, and how it did so. This perspective relies on ecocultural theory (Weisner et al., 1999, Weisner, 2002). In this report, we focus on family sustainability when there are children with troubles in the family. Family routines are considered sustainable when they (a) fit current resources, (b) reflect balance among competing and often conflicting family interests, (c) are stable over time, and (d) reflect the pursuit and achievement of meaningful family goals and values (see Method section for more detail about these dimensions). Recent welfare law changes may reduce the numbers receiving state assistance, but are those parents who are now working but not receiving AFDC assistance, and their children, able to sustain a meaningful, stable daily routine, given the low wages and episodic work situations most of them face? In our longitudinal study of European American, primarily middle-class, families of children with developmental delays, we demonstrated that children's difficulties are related to sustainability of the daily routine and family adaptations more generally (Gallimore, Coots, Weisner, Garnier, & Guthrie, 1996; Gallimore et al., 1999). Children in the European American sample with more severe behavioral problems led to more difficulties in sustaining a daily routine as did low incomes and problems with stable work and insurance. Given these findings for middle-class families, we expected that sustainability and work issues would be even greater among the working poor families of children with troubles. We felt that supportive programs such as New Hope, although not specifically targeted to children with disabilities and their families, might assist some of these families in sustaining their family routines.
The New Hope Child and Family Survey sample
After the first 2 years of New Hope, we identified a child and family subsample of 745 families (366 experimental, 371 control) who had at least one child between the ages of 1 and 10 at baseline. Two children (balanced for gender if there were 2 children in the household ages 1 to 10) were randomly selected as focal children. Survey researchers interviewed parents about the use of programs and services other than New Hope, economic outcomes (hours of work, hourly wages, and the type of jobs held), and noneconomic outcomes regarding family functioning, parent well-being, and children's development. Administrative records provided data on a variety of other economic outcomes, including welfare payments, food stamps, Medicaid benefits, and receipt of earned income tax credits.
Investigators began the 3-year ethnographic study in spring 1998, during the final year of the New Hope experiment. The ethnographic study provided a richer, more detailed understanding of the impact of New Hope on participating families than could be gained from the child and family subsample survey alone. The New Hope Ethnographic Study investigators drew a stratified random sample of 45 families from the full child and family subsample, with equal representation of both the experimental and control groups. One family dropped out very early in the study, and 2 families did not begin until spring 1999, leaving a final sample of 42 Ethnographic Study families with 67 focal children for whom full ethnographic data were available for this report. In return for their participation, each family was given financial compensation amounting to $50.00 for every 3 months of their participation. Descriptive statistics for both the full child and family subsample and the 42 families in the full ethnographic sample are presented in Table 1. More information on the Ethnographic Study is available in Duncan and Gibson (1999), Gibson and Weisner (2002) ,Lowe and Weisner (in press), Weisner, Bernheimer, Lowe, and Gibson (2000), and Weisner, Bernheimer, Lieber et al. (2000)
Fieldworkers used ethnographic methods to engage Ethnographic Study parents in conversations about their lives, their concerns and hopes, and their everyday routines, using techniques and methods from the standard set of ethnographic and qualitative research methods (Bernard, 1995; 1998; Miles & Huberman, 1994; Pelto & Pelto, 1981). Researchers jointly developed a lengthy set of domains and topics to organize these conversations. Topics covered included work experiences, child care and child monitoring, take up and use of the New Hope offer, roles of fathers and partners in mothers' lives, family finances, and beliefs about the welfare system, to name a few. If parents did not bring up topics, fieldworkers raised them and probed for information. Thus, fieldwork data are as complete as possible across all the ethnographic cases.
Fieldworkers had a median of nine contacts with each of their families during the first year of the ethnography (ranging from 2 to 16), with fieldnotes following each contact. Across all families, a total of 402 contacts occurred. Open-ended conversations began our fieldwork relationships; each participant was encouraged to “tell their story,” and these stories then led to more focused qualitative interviews on topics important to them and to the project. Fieldworkers did participant observation in homes, took families out for lunch and dinner, went with them to church, shopping, and family visits, and visited the children's schools. The methods we used are further described in Weisner et al. (1999). Although we were not able to undertake a comprehensive community and institutional ethnographic study, the methods are ethnographic because we attempted to understand the overall ways of life of the families in the study in a holistic way, using both interviews and field observations over an extended period of time. Qualitative fieldnotes, comprising hundreds of pages of notes for each family, were gathered and analyzed using the EthnoNotes system (Lieber, Weisner, & Presley, in press). Quantitative scores also were derived from the fieldnotes (see quantitative methods below), and survey and administrative data from the overall New Hope study were also available. Our mixed-methods approach links holistic, contextual, and quantitative description and assessment to better describe and assess (in this case) families and children (Fielding & Fielding, 1986; Greene & Caracelli, 1997; Weisner, 1997, 2002)
In order to understand families with children who have troubles, we analyzed the data that had been collected through the first year's worth of fieldwork by the Ethnographic Study team. We assembled all comments by parents and fieldworkers regarding their children's problems, examined data from the survey on their progress, and talked with each fieldworker about what they had observed directly from fieldwork and participant observation. We used the fieldnotes as well as scores for sustainability of families' daily routines, and the four variables that affect sustainability—resources, balance and congruence, meaning and goals, and stability. In the following section we describe the quantitative measures created from the ethnographic data.
Child Troubles Score
At the end of the first year of the Ethnographic Study, fieldnotes were used to create a “troubles” score for each of the 67 focal children in the sample. Child troubles were defined by the presence of any of the following items: (a) special educational placement in school; (b) behavior problems at home beyond the typical, such as chronic aggression; (c) behavior problems at school sufficient to lead to school intervention, such as phone calls to parents, being suspended or expelled from school; (d) difficulties with schoolwork that persisted; and (e) chronic health problems, for example, asthma or allergies. All scores were verified with the fieldworkers who knew each family intimately. These scores were used to describe the type and prevalence of child problems in the New Hope families.
We based our data on parents' reports about what others (educational and health professionals) have said their children needed, what services they were currently getting, parents' perceptions of their child's status, and on the fieldworker's own observations of the child. Our approach is compatible with that of Stein and Silver (1999), who defined chronic conditions by examining the consequences of diverse medical, behavioral, or cognitive disorders rather than by using diagnostic labels. We use the term children with troubles deliberately: Parents and others have said to us, and in many cases fieldworkers have directly observed, that these children are not doing well by various criteria and are of significant concern to parents. We usually also have explicit, independent professional diagnoses or signs from schools and clinics that these children have educational or health or behavioral problems, based on the criteria used by those professionals or schools and clinics (i.e., children were in special education; had been held back in school due to insufficient academic progress; were going to a clinic for a significant health problem; or, in fact, had been in the child services system at one or more points). Our fieldwork judgments and ratings are thus typically confirmed by at least one other independent piece of evidence.
The New Hope survey data provided yet an additional perspective on our ethnographic assessments of child troubles. Parent and teacher survey data were collected 6 to 18 months prior to the family's entry into the Ethnographic Study. We analyzed the scaled scores from the parent and teacher survey data on children in the study to search for children in special education or with academic problems, behavior problems at home, behavior problems at school, and those with two or more troubles. Because of missing survey data for a few of the ethnographic sample families, the N was 38 for the parent problem behavior and ranged from 25 to 38 for the teacher data (SSRS Academic subscale, Teacher Positive Behavior, Disciplinary Action, Classroom Skills, and Teacher Problem Behavior). There were no significant differences on the six measures between children in special education and the rest of the sample, although all differences were in the expected direction. However, there were statistically significant differences for the children with two or more troubles identified in the ethnography. They differed significantly from children with no troubles or one kind of trouble on three scores: They were lower on the Academic subscale and Teacher Positive Behavior and higher on Teacher Problem Behavior. For the rest of the groups, the significant differences were in the expected direction. The independently obtained psychological assessments of the children from the survey data generally support the ethnographic results.
Assessment of Family Sustainability
Fieldworkers rated their families on sustainability and its four dimensions: resources, balance, stability, and meaning. An 11-point scale was used (0 to 10). The four sustainability dimensions are defined as follows:
Resource fit is the match between resources available, and the family's activities. Lack of resource fit can lead to interpersonal tensions, poorer health, and lower levels of well-being for parents and their children. Although more income and greater education certainly make sustainability easier for many families, more is not necessarily better. Resource fit assesses making ends meet, given the family's routines and activities. Our definition would give a relatively high score, for example, for working poor families, who, even with limited income, organized a sustainable routine containing activities that provided for their needs and goals given the resources available.
Balance and congruence: The varied goals, values, and interests of family members motivate and direct the activities that make up family routines. When these goals, values, and interests mesh together well, or are congruent, the overall family routine is more sustainable over time, and there is less conflict. However, when interests and goals do not match, as often happens, sustainability is threatened by high levels of conflict and a lack of balance. In a family with high congruence, parents assess the inevitably competing interests of family members and take action to try and maintain some balance among them. This involves the complicated task of assessing whose concerns should take precedence when they are in conflict. For working poor families with children who have problems, matters of balance and congruence can be highly salient because these children have the potential to create conflict and threaten the sustainability of family routines more than other children do.
Meaning and goals: Routines are more sustainable when they are meaningful to participants and fit with their goals. When the family achieved goals in their daily routines that parents personally and culturally valued, the routine was scored higher. An example of a lower rating on fit of the daily routine and activities with parental goals is a working single mother who placed her toddler in day care, despite her belief that young children should be cared for exclusively by family members.
Stability/predictability: Children and parents benefit from a manageable routine that can be reasonably well-predicted, with expectable patterns of activity. Of course, change and adjustment are required if routines are to be sustained in the long term. Such change is often a positive sign for adaptation and healthy development. However, constant, unpredictable change in the everyday activities of children and the significant others around them is a sign of low sustainability. Families who moved several times a year, whose children had to change school each time, received a lower score on stability than did families who remained in the same home. Families whose work hours unpredictably fluctuated from week to week also received lower scores.
To test whether our fieldworkers' sustainability ratings could be made reliably by other observers, we had three raters who were not part of the New Hope fieldwork team take the cues written by the fieldworkers as qualitative evidence for their rating and rate seven randomly selected cases again. The raters did this without any knowledge of the original fieldworkers' ratings or child troubles scores. Intercorrelations for each of the four ratings across all seven cases were then calculated and the average intercorrelation was used to compute Cronbach's alpha (.95) for all four raters. We conclude that the translation of fieldnote information to rating scales can be reliably replicated.
Family Level Troubles Scores
Because the focal child troubles score is based on a different unit of analysis than our family-level sustainability measure, we also computed a family-level child troubles score. The family measures are based on the assumption that the child in the family with the most difficulties will have the greatest impact on family adaptation. First, we rated the family based on the focal child with the greatest number of troubles (out of the five possible). If none of the focal children had any troubles, the family received a score of zero. If a focal child had four troubles, the family scored a four. Thus, each family was given an initial troubles score from zero to five, based on the most difficult focal child in the family.
Second, we analyzed these family-level troubles scores and found a distinct break point, between families with scores of zero or one, and those with scores of two or higher. These families had children with chronic, as opposed to episodic, problems and, therefore, a greater potential impact on the daily routine. Thus, we also coded each family for the presence of a focal child with two or more troubles versus those families with focal children who had only one or no troubles.
Our analyses addressed the following three questions: (a) What is the type and prevalence of child troubles in the Ethnographic Study? (b) What was the experimental impact of New Hope on families with and without troubled children? (c) What is the relationship between child troubles and sustainability of the daily routine?
Type and Prevalence of Troubles
Table 2 shows the characteristics of the children identified as having troubles. Of the 67 focal children in the ethnographic sample, 33 (49%) had at least one significant problem. Sixty-seven percent of the children with troubles were boys and 42% were in special education. Labels reported by parents from schools or clinics for these children included learning disabilities (n = 6); emotional disorder (n = 3); mental retardation (n = 2); speech/language problems (n = 2); and hearing impairment (n = 1). Eighteen children (55%) had other school problems (grade retention, below grade level performance in academic subjects) but were not receiving any special education services. Fifteen children (45%) had behavior problems in school, and 16 (48%) had behavior problems at home. Six children (18%) had chronic health problems (e.g., asthma, allergies). Of the children with problems, 45% were African American (compared to 46% in the full Ethnographic Study), 30% were Latino (compared to 37% in the Ethnographic Study), and 24% were European American (compared to 16% in the Ethnographic Study.)
It is also useful to frame our findings in the context of what might be expected across children in Wisconsin. The Twentieth Annual Report to Congress on the Implementation of the Individuals With Disabilities Act contained data on children in Wisconsin who received special education services in 1998. The 21% of Ethnographic Study children in special education was almost twice the rate of approximately 11% of children between 6 and 17 years of age in Wisconsin who received special education services. The percentage of children in each category of disability was approximately twice as high in the Ethnographic Study as in the state, with the exception of emotional disabilities, where there are almost three times as many children in the Ethnographic Study with this problem. Obviously the Ethnographic Study families were not representative of all families in Wisconsin. However, the substantially higher rates of child problems in the Ethnographic Study gives some frame of reference regarding the magnitude of the numbers of children with troubles we found in the Ethnographic Study.
Twenty-five (60%) of the 42 families reported at least one child with one or more troubles; 38%, at least one child with academic problems; 31%, at least one child with behavior problems at school or at home; or a child who had been placed in special education. Fourteen percent of the families had a child with a chronic health condition. The incidence of families with no father/father figure in the home was identical in families of children with troubles and in the entire Ethnographic Study sample (36%). There were no associations between father involvement (consistent, sporadic, or none) and child troubles (two or more vs. less than two).
Additional information about the mothers used in the sections on ethnographic findings may be found in Table 3.
In terms of learning difficulties, many of the 18 children who were not in special education, but who were identified by their parents as having significant problems in school, appeared in danger of falling through the cracks at school—neither receiving special services nor learning in the regular classrooms. School learning problems included the fallout from behavior problems, difficulties with English, and academic problems in the absence of behavior problems. Maria, who spoke only Spanish when she was enrolled in English immersion at the age of 5, was still well behind her classmates at the age of 9. Her mother, Dulce, was convinced that she would have done fine if she had been placed in a bilingual class. She made sure her two younger children went into bilingual programs and reported that neither of them had any academic difficulties. Erik's mother, Faye, was frustrated by his poor grades; his last report card contained only Ds and Us (unsatisfactories). He did not pay attention in the classroom and spent too much time “ clowning around.”
Some of it could be my fault because I should be demanding to see his homework,” she explained. “But when I do that he says he doesn't have any. Sometimes I am so tired, I am so tired … Other things on my mind, I am so tired. I'm cooking that day or doing what I'm doing. I'm like, ‘good night.”
Behavior problems at school included inattentiveness in class, defiance in response to teacher instructions, chronic lateness, and truancy. For example, Carla talked about her son's inability to get up in time to get to school once she started working. She even took him to the doctor because he slept so much; the doctor just told her that her son was sleep-deprived. In several cases there were reports of teachers reaching out to parents in an attempt to solve the problems. Erik's teacher dropped by Erik's house to explain to his mother that he was so busy being the class clown that he was not concentrating on any of his work and was in danger of failing. She suggested that Faye get him involved in Big Brothers or sign him up for an art class because he was really good in art. Faye had done that when he was younger, but he was not able to stay consistently involved as she frequently needed him to stay home and care for his younger brother.
Behavior problems at home ranged from withdrawal, defiance, and constant fighting with siblings to involvement with drugs and gangs and running away. Mothers had different theories about their children's behavior problems. One single mother attributed her daughters' defiance to the fact that they had no father figure. Two mothers blamed the father or father figure for the children's negative behaviors. In one case, the mother suspected sexual abuse was the cause of her daughter's emotional withdrawal. In another, the mother accused her partner of treating the child with the behavior problem less well than he did his own biological child. Several mothers talked about the impact of their working on their children's behavior. Samantha talked about the effect of her work-related stress on her two daughters.
They sit here and argue and have fits nonstop. To the point that I lock them in their room until they come downstairs and can get along. I can't do anything else with them. They are stressing me out worse than I stress myself out. They argue. I have to go to work and deal with all this crap. Then I come home and they're arguing.
In some cases, children's behavior problems took a heavy toll on the mothers. Belinda admitted that she did not like being a parent. She did it because God meant for her to have her children, and they were her responsibility; but they were getting on her nerves. She felt that she was ready to snap. She reported that last week she yelled at them and told them she hated them. She said if they were not careful they might come home and find her gone.
Our ethnographic work suggested that working poor parents probably have more troubled children to deal with than are being picked up through formal surveys or screening programs. Of the 22 families with a child who had troubles for whom New Hope survey assessment data were available, only 16 (73%) were identified by the survey as having a child with troubles. Problems found by fieldworkers but not picked up by the survey included placement in special education, academic problems, behavior problems in the home, and chronic health problems. On the other hand, the survey identified four Ethnographic Study families (18%) as having one or more children with problems who were not identified as such by fieldworkers. Those problems included two grade retentions, two suspensions from school, and one placement in special education. Because the survey data were collected between 6 and 18 months prior to the ethnographic data collection, we are cautious in making comparisons between the two data sets. However, we are quite confident of the pattern among those children who were not captured in the survey but were troubled according to parents and fieldworkers: they clearly were troubled, and those troubles existed during the survey period but had not been identified. We are less certain in the other direction: Some of the children who were not considered as troubled by fieldworkers could have had problems when the survey was done that caused parents to report troubles in the survey that were not present 6 to 18 months later.
Experimental Impact of New Hope on Families With and Without Troubled Children
Because New Hope did have impacts on the family and children, it was plausible to consider whether there was a differential New Hope program impact mediated by child troubles in the household. New Hope had effects through its package of benefits (income supports, child care, health insurance, and full-time work if needed) in exchange for 30 hours or more of work each week (Bos et al., 1999). For example, New Hope increased incomes and work hours among participants in the experimental group who at baseline were not employed (Bos et al., 1999). New Hope also augmented the use of child care by its offer of a subsidy and the use of HMO medical care through a subsidy if parents were working 30 hours a week (Bos et al., 1999). Teachers reported that boys from the New Hope intervention group were doing better in school and had fewer behavior problems than did the boys from the control group, although there were no such findings for girls (Huston et al., 2001). All these types of experimental impacts might have affected family sustainability differentially for families living with troubled children.
To test this, we compared the differences in sustainability ratings for the families with high (2 or more) and low (0 to 1) child troubles for the experimental (n = 22) and control groups (n = 20) separately (see Figure 1). Recall that children with two or more troubles were generally those with chronic, as opposed to more episodic, problems. Because chronic troubles are often symptomatic of underlying disability (in fact, we knew directly that many of these children had a disability), we assumed that these children's more severe troubles were present in some form prior to random assignment. Therefore, the family and child differences we found were almost certainly prior and exogenous to the experimental intervention itself.
As Figure 1 indicates, control group families with a child who had two or more troubles did have significantly lower sustainability ratings than did families with a child who had only one or no troubles. This was the case for overall sustainability as well as for each of the four dimensions of sustainability. However, no such difference appears for New Hope families; New Hope closed the gap. Indeed, sustainability ratings for control group families with low or no child troubles are similar to the sustainability ratings for New Hope families who had a child with significant problems. Because families were randomly assigned to New Hope or control groups, and in the Ethnographic Study we randomly sampled families from both experimental and control groups, these data suggest that the kinds of programs offered by New Hope did have some positive impacts on family adaptation.
Two cases from the Ethnographic Study illustrate the positive spillover effects of the New Hope treatment. Wendy's son Jordan was in a special education program for children with emotional disturbances, and he had behavior problems at home as well. New Hope helped Wendy obtain a child-care license and open up her own child care program. “I'm hoping and praying to be stable with this occupation because of the kids,” she commented. She looked forward to being able to go down to the school more often when there was a problem with Jordan. Edith's son, Max, had a hearing problem and behavior problems at home and school. Thanks to the medical insurance offered by New Hope, she was able to have her son's hearing problem treated surgically and access some psychological counseling for him as well. Thus, a well-run antipoverty program can make a difference in the daily lives of families with children who have significant problems. In the absence of access to a program like New Hope, however, child troubles can and do lower family sustainability for the working poor.
Child Troubles and Family Sustainability
For the entire sample of 42 families, the number of child troubles was negatively correlated with our ratings of the sustainability of the family daily routine. The correlation between overall sustainability and number of troubles approached significance, r =−28, p = .07. The mean sustainability scores of those families with a child who had two or more troubles were lower than those among families with children who had only one or no troubles, although the differences were nonsignificant.
Selected items on the parent survey were examined to determine whether there were differences between families of children with two or more troubles compared to the other Ethnographic Study families. We examined New Hope survey data on a number of work and personal measures. There were no statistically significant differences, but the trends reveal that families who had at least one child with two or more troubles reported more financial worries, greater financial strain, and less regularity in their daily routine. These families were more likely to have used AFDC or food stamps in the 24 months prior to the survey. Families of children with troubles also reported being employed more weeks in the 24 months prior to the Survey (M = 85.1 weeks, standard deviation [SD] = 26.0) than did families of children without troubles (M = 68.3 weeks, SD = 41.0), even though they had more financial strain.
We next examined the fieldworkers' notes on each case, searching for evidence of patterns in the ways child troubles affected parents' abilities to create a sustainable daily routine. Instead of simple patterns, we found a tangled web of multiple pressures and conflicts occurring along with child troubles. These included stressful home situations, boring jobs, insensitive employers and teachers, dangerous neighborhoods, lack of transportation, and abusive men. These kinds of problems, equally as or much more salient than child troubles, at times push child troubles into the background. Child troubles were certainly one, but only one, of many pressures on a sustainable daily routine. Child troubles in particular, by themselves, often were not the critical factor in lower sustainability in the context of the many other difficulties parents faced.
Child troubles and low sustainability
In Katrina's case, when her car broke down, she was unable to get it repaired for 2 weeks because she had to wait for her next paycheck. During those 2 weeks, she stopped going anywhere other than work, and the family ate only frozen pizzas. In this same period she had to move her youngest son out of day care because he was biting people. Fortunately, the day care was within walking distance of her house. However, she also had to find transportation for her other three children to get to day care in a different part of the city. She talked with the fieldworker about her frustration with 8-year-old Neil, who had significant behavior problems at home and at school: ”I've got the other [three] kids, the world cannot just be him…. We gotta eat. I gotta work. I am really angry with him because he takes away from everything. … He's making it hell for everybody.” Katrina said that nothing was easy with him; even eating dinner turned into an event. She wondered about increasing his Ritalin but was not sure it was a good idea. “I don't want the boy to veg out. I don't want him to be a vegetable. He's got to have some life.”
The teacher told Neil she did not like him, and, not surprisingly, his behavior problems at school escalated. Neil's father, who was recently released from jail, stayed home with the children while Katrina worked. She was not sure how long this would last. The moment she walked in the door, he left because he was so eager to get away from them. Katrina said, “He is getting tired of them too.”
The complex relationship between overall difficulties with family adaptation and sustainability and child troubles is well-illustrated by the case of Trisha and her four children. All four had problems, including the 13- and 14-year-old girls in the ethnographic sample. In the 18 months that Trisha had been in the study, she was not able to hold a job and continued to live month to month, using a combination of child support checks and Supplementary Social Security Income (SSI) payments. Although she reported that she was not using drugs, her partner still used cocaine and possibly crack. She continued to have high levels of involvement with the police and the child protective services in Milwaukee. She was planning to move to a small town in Wisconsin and believed that this would be a fresh start, but it is hard to see how she would be able to leave her significant problems behind her. In the meantime, both daughters had become sexually active, started using drugs, ran away from home with some older gang members, been arrested for shoplifting, and ticketed for violating the city's curfews. Trisha had left (or been asked to leave) at least two jobs because of her daughters. In the first, a cash register position, “they had to pull me off the register [to leave work and go get my daughter], and I had 15 people [waiting]. You can't do that at work.” Her situation was complicated by the fact that her boyfriend resented her working, in part because he did not want to be stuck at home with the kids. “If you weren't fighting with my kids all the time, maybe I could keep my job,” she told him. She implied that the girls were also the reason she was fired from a job in a travel agency. “It's not like I can't hold a job. It's just that with everything that keeps going on with the girls… every time I get a job that's when they screw up.”
Our qualitative data suggest that among parents who reported that they were struggling with work schedules and child care, parents with children with behavior problems were neither much worse nor a lot better at managing their daily routine than those who were not dealing with behavior problems. At the same time, a careful review of our cases showed that in every family in which fieldworkers perceived that the family, overall, had low sustainability in their daily routines, there always was at least one child in the household who was reported to have significant behavior problems.
Impact of child troubles on employment
Surprisingly, in 20 of the 25 families with children who had troubles, there was no evidence that child troubles in particular constituted a barrier to employment or influenced the number of hours worked. We have described and quoted some of the mothers who did have such difficulties, but there were more cases in which this was not an explicit or major issue. The majority of mothers were either employed or not working for reasons unrelated to their children (e.g., some parents were not working due to their own health problems, other family conflicts, or lack of skills and experience). Most of the employed mothers gave no indication that their children with troubles specifically and adversely affected their job situation. For example, Janet, whose son Nick had language and speech delays and was receiving special education help, opened a child care center in her home. She chose to do child care because it allowed her to stay home with her children and be her own boss. When she wanted to speak with Nick's teacher or observe in the classroom, her aide took over the child care. Marisa worked as an aide in the public schools and was home by the time her daughter Luz, a teenager with learning disabilities and behavior problems, was out of school. Although Luz did not have a direct impact on Marisa's work, Marisa commented that her patience was worn thin by working with children all day and then coming home to face her difficult daughter.
In 5 cases (20% of the 25 families with children who had problems), however, the children did directly influence the mother's ability to find or keep a job or the number of hours worked. Three of these cases were in the control group. One such case was Trisha's, a single mother of four children with problems. Another was Christy, a single mother without a father/father figure available to help out. Christy felt unable to work because of the needs of 5-year-old Andy, who had Down syndrome. He functioned like a 3-year-old. He was not toilet-trained, which created a major problem for Christy. She felt she could not leave him with anyone because he could not speak and make himself understood. She had to keep the apartment very clean so that he did not pick up anything from the carpet and eat it. She had to make her apartment childproof—one time he opened the freezer door and left it open and then turned on the oven when Christy fell asleep. She could not get Andy to understand—or at least to remember—that he should not do these things.
Caroline reported that she and her husband had not found any discipline strategy that worked with Damien. Damien had been diagnosed as emotionally disturbed and by the age of 13 had been in 14 different programs. Caroline claimed that she lost four jobs because of calls from the school about his behavior or because he ran away. The fourth case, Kristina, was actually working but only because the father of one of her children had recently been released from jail and was available to watch the children while she worked. Carla, the fifth case, held a full time job in a factory and frequently had to leave work to go to Anthony's school and speak to his teacher about his learning and behavior problems. She felt fortunate that she was able to work overtime and make up the lost income.
Impact of employment on child troubles
There is also evidence that employment was among the many factors that exacerbated child troubles over the course of our fieldwork. One mother talked about her fear that their children were “turning bad,” in part because of the hours she spent working. Two of Carla's seven children were “acting out” at home and at school. She was not sure what was wrong.
I don't know if it's their friends, the neighborhood, what's going on. Now that I'm working… you know, before I wasn't working and they didn't act this way. Now that I'm not home, not taking care of them full-time, they are getting bad.
Carla went on to talk about the difficulties in keeping up with school visits, doctor appointments, or Women, Infants, and Children Program appointments because she worked during the day. She reported that she had thought about taking on a second job to bring in more money, but she worried about not seeing her children at all. She felt that she had too little time with them. She used to be able to go to the kids' schools, get involved in activities and field trips, and do things with them in the evening. Now she had neither the time nor the energy.
Alicia talked about the problems she had with her two sons, who exhibited behavior problems at home. They were constantly bickering, and she was always after them. One of them had started to lie a lot. She said she loved her kids but was always tired. It became worse since she began working the third shift because she really did not have the time or energy to give them the individual attention they needed. When she came home she wanted to relax and not be bothered. Much of the time she closed herself in her room and told them not to bother her. She resorted to whipping when they “really push her,” but she did not like doing it. She described herself as the type of mother “that will yell and carry on one minute, then give big hugs and kisses the next.”
Taken together, the qualitative and quantitative data support the findings of others who have reported a higher incidence of child disabilities and problems in low-income families than in higher income families. Further, the ethnographic researchers spotted significant problems that had not been recognized by schools or other service professionals: children (and families) who had “fallen through the cracks” in one way or another. Many of the children with significant academic problems were not receiving any special help, and those with serious behavior or health problems were not getting regular treatment. Roughly an additional third were found than had been identified by survey assessments or outside professionals.
Our qualitative data do not support the seemingly evident notion that child troubles alone, considered apart from the overall family context, constitute a significant barrier to obtaining and keeping employment. In 84% of the families in the troubles group, there was no evidence that the children with troubles, as a single cause, interfered with the parent's ability to work. What is more likely is that there is no single, discrete cause for varying work patterns among the working poor and that among the multiple factors, having a child with significant troubles adds to the cumulative burden and can sometimes tip the balance and lower family sustainability of daily routines. The ethnographic evidence suggests that there is an association between child problems and reduced abilities to sustain a daily routine but that the effects are cumulative and interact with the many other perturbations and strains in these parents' lives.
We note that much of the research on the impact of child problems on low-income parents' employment is focused on children with an identified condition. However, we would argue, as do Stein and Silver (1999), that the disease-specific approach used in much of the earlier work does not seem to match the experiences our parents reported and fieldworkers observed. The actual experience of attempting to care for, find treatment for, and adapt to children with troubles among the working poor depends less on disease-specific diagnoses and more on the broader problems of sustaining a daily routine for the family and dealing with low-wage, no-benefit, episodic jobs.
Our data confirm work by Newman (1999), Edin and Lein (1997), and others who found that working poor families continually face a “house of cards” concerning support and burdens that they overcome in making ends meet and sustaining their family activities. Public assistance, wage work in the informal economy, kin, and targeted services all appear to be crucial factors in keeping the family routine going at various points. Sustaining a daily routine of activities constitutes a constant balancing of multiple constraints and opportunities taxing the carrying capacity of these parents' cultural ecologies. For the working poor, there is relatively little in reserve. It is not often possible for parents to concentrate on just the troubled child, largely exclusive of everything else. Theirs is not a single-problem world, with reserves of cash, equity, insurance, and a variety of options. With few or no such reserves, it is not easy to change one aspect of the daily routine without affecting many of the other elements.
The impact of a troubled child, so commingled as it is with the many other pressures on these working poor families, contrasts with our data from European American, mostly middle-class families, where a child with developmental delays has a more clearly defined and independently recognized impact on family adaptation (Gallimore et al., 1996; Gallimore et al., 1993; Weisner, Matheson, & Bernheimer, 1996). In middle-class families, the impact on family adaptation is usually a result of the many accommodations parents make to having a child with problems: for example, seeking services, arranging child care, redistributing the workload among family members to allow for more attention to be given to the child with problems, dealing with siblings, or cutting back on social activities. In contrast, the Ethnographic Study families incorporated their child much more matter-of-factly into their everyday life. Middle-class families with more resources may struggle to include the child with problems in outings and family events even when it is very difficult and leads to family conflict and then eventually leave the child at home because of the hassle of including him or her. The families in our study had fewer resources for outings and were less likely to have caregivers solely for the child with problems. The Ethnographic Study families had so many concerns when trying to make ends meet that many of the child-focused interventions found in middle-class families just did not exist in our study sample.
Many of the adaptive problems found in middle-class families were not apparent in the Ethnographic Study families either. For example, in some middle-class families, a large proportion of the material and human resources over and above basic subsistence are allocated to the child with a problem, sometimes to the detriment of the couple's relationship or siblings. To date, we have not seen this in Ethnographic Study families. Middle-class families report struggling to decide between jobs, careers, or staying home. Some mothers leave or defer professional careers to take less prestigious jobs or to stay home. The working poor rarely have the luxury of such options. Their conflicts involve just trying to keep often poorly paid jobs or dealing with chronic job changes and the transaction costs of trying to get public assistance. In working poor families, the accommodations made by parents are more often for the purpose of making ends meet.
The specific pathways between child troubles and sustainability are less direct and more difficult to tease out in the Ethnographic Study sample. However, this demonstrates the magnitude and number of other problems parents face, not that troubled children are not of great concern. As we show in the summaries of ethnographic reports, parents were very well-aware of the problem behaviors and health issues of their children and struggled to address them.
In conclusion, low-income families with children who have significant disabilities and other learning and behavior troubles comprise a population that deserves special consideration from a policy perspective. We note that due to reductions in federal funds for SSI, children with serious mental, emotional, and behavioral disorders as well as respiratory problems have been disproportionately represented among those losing benefits (Luprest, 1996; Ohlson, 1998; Rosman & Knitzer, 2001). Our study shows, as the survey-based literature also suggests, that such children make it harder to sustain a daily routine, and they and their parents deserve additional support. The package of benefits offered by New Hope appears to be one way to offer selective help for some families. Additional benefits would assist this population, including waivers from work where indicated, along with improved home care options for parents. These parents would benefit from more flexible work, including time to take telephone calls and attend school conferences. They need health care, school services, and family services. In many cases, the “problem” is as much a service problem (lack of available services) and job site problem (absence of work flexibility, no child care, and/or no health benefits), as it is a child, parent, or family adaptive problem. An adequate menu of benefits can mitigate the impact of family difficulties around work and increase work hours and income, as the New Hope experimental data demonstrated was possible for all families, including those with children who had problems. The New Hope experimental impacts and the experimental findings from our ethnographic studies suggest that making such benefits (e.g., child health care, child care subsidies, increased Earned Income Tax Credit benefits for working women with children) a standard part of work, rather than a specialized add-on benefit, would improve the ability of poor families with children who have disabilities and significant troubles to sustain their daily routines and enhance the well-being of the whole family.
The authors thank the New Hope fieldworkers (Conerly Casey, Nelle Chmielewski, Victor Espinosa, Christina Gibson, Eboni Howard, Katherine Magnuson, Andrea Robles, Jennifer Romich, Devarati Syam), and the families who shared their lives with us. Without these families and fieldworkers, the study would not have been possible. We also thank Eli Lieber and Sonya Geis of the Qualitative Fieldwork Laboratory at UCLA. Research and data analysis were supported by the Joint Center for Poverty Research (Chicago/Northwestern), the MacArthur Network on Successful Pathways Through Middle Childhood, National Institute of Child Health and Human Development Grant R01HD36038-01A1 (Aletha Huston, Greg Duncan, Robert Granger, and Thomas Weisner, principal investigators), and the UCLA Fieldwork and Qualitative Data Laboratory in the Center for Culture and Health. This study is part of an evaluation of New Hope conducted by the Manpower Demonstration Research Corporation.
Authors: Lucinda P. Bernheimer, PhD ( email@example.com), Research Scientist, Department of Psychiatry, Center for Culture and Health; Thomas S. Weisner, PhD, Professor, Departments of Psychiatry and Anthropology; and Edward D. Lowe, PhD, Assistant Research Anthropologist, Department of Psychiatry and Biobehavioral Science, Center for Culture and Health, UCLA/NPI, 760 Westwood Plaza, Box 62, Los Angeles, CA 90024-1759. Requests for reprints should be sent to the first author