Program

Aug 04, 2016

OVERVIEW

The Building Strong Families (BSF) Program – which provides a combination of group education and individual support – was developed to provide support and linkages to services for young couples who were unmarried at the time they conceived a child.  It was designed to impact relationship duration, relationship quality, and child well-being.  At both the 15- and 36-month evaluations, few positive impacts were found.  The exception to this was that in Oklahoma City positive, but diminishing, impacts were found in outcomes in all three areas.  Elsewhere the program had few positive impacts on any of the three desired outcomes.

DESCRIPTION OF PROGRAM

Target population: Couples who were not married when they conceived and are newly parenting

The BSF program aims to develop relationship skills for couples who were unmarried when they conceived a child and who are parenting an infant.  The program aims to equip couples with skills in communication, conflict management, and parenting.  The program also aims to increase the percentage of couples who are married and co-parenting their children, but also looks at relationship satisfaction, ability to manage conflict or avoid bad conflict management, father involvement, faithfulness, and perceived partner support.  It does this with a core curriculum, taught through group sessions, that addresses these skills, as well as individual support from family coordinators that consists of encouragement, reinforcement of skills, emotional support and referral to services, as needed, for education, employment, mental health, child care, housing, and legal services.  BSF can be implemented in a number of different settings, including as part of existing federal programs and community agencies, as well as in community centers developing this program as a stand-alone program.  The core group sessions consist of 30 to 42 hours of programming, and can take anywhere between six weeks to five months to complete, depending on the length and frequency of the sessions.

EVALUATION OF PROGRAM

Wood, R. G., McConnell, S., Moore, Q., Clarkwest, A., and Hsueh, J. (2010). The Building Strong Families Project: Strengthening unmarried parents’ relationships: The early impacts of Building Strong Families. Mathematica Policy Research Report, 1-31.

Wood, R. G., Moore, Q., Clarkwest, A., and Killewald, A. (2014). The long-term effects of Building Strong Families: A program for unmarried parents. Journal of Marriage and Family, 76, 446-463.

Evaluated population: There were just over 5,000 couples (10,000 individuals) in the evaluation, distributed across eight implementation sites. There were 930 couples in Atlanta, 602 couples in Baltimore, 652 in Baton Rouge, 695 in Florida counties, 405 in Houston, 466 in Indiana counties, 1,010 in Oklahoma City, and 342 in San Angelo, California.  Couples were eligible if both members wanted to participate, they were romantically involved with each other, they were either expecting a baby together, or had an infant who was less than three months old, were unmarried at the time the baby was conceived, and were both 18 years of age or older.  They were deemed ineligible if they reported experiences of intimate partner violence that the study leaders thought would be exacerbated by the program.

Across the sites, the demographic characteristics of the sample differed slightly.  However, 52 percent of the couples in the overall sample were both African-American, 12 percent were both white, 16 percent were both Latino, and 16 percent were either mixed-race couples or races or ethnicities that were not African-American, white, or Latino.  Two of the sites – Florida and San Angelo – recruited parents who already had babies, while the others targeted couples who did not yet have a child (although in Indiana just over half did).  In none of the sites did more than 50 percent of the couples report that both members had a high school diploma, suggesting very low levels of formal education, and earnings hovered around $20,000 a year.  Just under half of couples reported that at least one of them had children from a previous relationship, except in Atlanta and Baltimore, where just over half did so.  Finally, most of the couples were both over 21.  The majority of couples in all sites, except for Baltimore, reported that they expected either a “pretty good chance” or “almost certainty” that the relationship would end in marriage. The Baltimore and Baton Rouge sites had the lowest rate of marriage (four percent), and the Houston site had the highest (11 percent).  The majority of the couples across all sites were either married or cohabiting at baseline.

Approach: The evaluation compared the control and intervention groups, both pre- and post- intervention. Control-group participants were not eligible for Building Strong Families services, although they could seek services elsewhere. Couples were randomized using a computer-generated procedure after they were screened for eligibility.  The study examined a number of outcomes: outcomes related to the relationship status and conflict management skills (measured with two scales for the ability to manage conflict effectively, and the ability to avoid destructive conflict).  These and scales on relationship quality were averaged across the male and female partners, because these were conceptualized as couple-level, rather than individual outcomes.  Additional outcomes measured included attitudes towards marriage, experiences of intimate partner violence (IPV), co-parenting (including father involvement), depression, and economic well-being (including employment in the past month, earnings in the past year, being below the poverty level, having had trouble meeting housing costs in the last year, or receiving TANF or food stamps in the last year).  Finally, in the long-term impacts report, measures of socio-emotional development of the child were evaluated.  These included whether the child had behavior problems, and whether the child report feeling emotional insecurity.

The report outlining long-term impacts shows the baseline equivalence of the two groups.  There were differences in income (though both groups were low-income), with the BSF couples making about $1,000 more in annual income than couples in the control group.  There were also differences in cohabitation at baseline, with the BSF couples slightly more likely to be cohabiting.  Otherwise, the groups were similar at baseline.

Data were collected at baseline when couples applied to be in the BSF program, and at 15 months and 36 months through a telephone survey.  At least one parent responded from 4,425 of the couples (87 percent).  This included 83 percent of the mothers and 72 percent of the fathers.  Analyses were conducted with the entire sample, by site, and with sub-groups by race and ethnicity.  At follow-up, relationship quality measures were only available for those still in a relationship, and conflict management only for those who were still in touch. Pooled sample results were calculated by generating estimates for each site and then averaging those estimates across the eight sites.

Wood, R. G., McConnell, S., Moore, Q., Clarkwest, A., and Hsueh, J. (2010). The Building Strong Families Project: Strengthening unmarried parents’ relationships: The early impacts of Building Strong Families. Mathematica Policy Research Report, 1-31.

Results I (15 months): Because control-group participants could have received services elsewhere, and intervention group participants could have turned down services, an essential component of the evaluation was whether or not access to training actually differed in the two groups.  Results indicate that the BSF program was successful at increasing the amount of training that people received.  This varied by site, but everywhere those enrolled in the program received more hours of education.  In Baltimore and Baton Rouge, those in the intervention group received only five to six more hours of training, while in San Angelo those enrolled in the program received 21 more hours (16 hours of group sessions and five hours of individual support).  There was a three-percentage-point difference in the number of fathers in the BSF group who reported education, a two-percentage-point different in the fathers who reported mental health counseling, and a five-percentage-point different in the fathers who reported any kind of support.  All three differences were significant.  For mothers, there was a one- percentage-point difference (insignificant) in mothers who reported receiving education, a two-percentage-point difference in mothers who reported receiving mental health counseling, and a two-percentage-point difference in the mothers who reported receiving any support service.

There were no statistically significant differences between the intervention and control groups in terms of whether the couple was still romantically involved, living together, or married at the 15 month mark.  There was a small positive impact on mothers’ attitudes toward marriage, but otherwise no significant impacts on attitudes towards marriage, relationship happiness, perception of support and affection from one’s partner, conflict management, or fidelity.  There were no differences in experiences of intimate partner violence (IPV) by intervention sub-group.   There was no impact on co-parenting quality or father’s involvement, though mothers in the BSF group reported significantly less frequently spanking their child and less frequently feeling stress or aggravation due to parenting.  Mothers in the BSF group reported significantly lower depression scores than those in the control group, though this was not true for fathers.  There was no impact on economic well-being.

There were slight differences in results by site. In Oklahoma, BSF couples were more likely than those in the control group to be romantically involved; to report relationship happiness, support and affection, fidelity, and effective conflict management; to be effectively co-parenting; and for the father to report providing substantial financial support.  Meanwhile, in Baltimore, BSF couples were less likely than those in the control group to still be romantically involved or report feelings of support and affection from their partner.  They also were more likely to report IPV, less likely to be co-parenting, and less likely to report any of the three types of father involvement.

Finally, couples where both parents were African American fared better under the BSF program.  Specifically, they saw an increase, relative to couples of other races, in feelings of support and affection for each other, better conflict management skills, more fidelity, lower IPV experienced by the men, and improved co-parenting.  There were no impacts on relationship status.

Wood, R. G., Moore, Q., Clarkwest, A., and Killewald, A. (2014). The long-term effects of Building Strong Families: A program for unmarried parents. Journal of Marriage and Family, 76, 446-463.

Results II (36 months): At the three-year follow-up, BSF couples were less likely to be romantically involved, and less likely to be living together, than those in the control group.  There were no differences in marriage rates.  There were no impacts on relationship quality, the quality of co-parenting, the ability to manage conflict, or the economic well-being of the family.  Fathers in the BSF group were slightly less likely to spend time regularly with their child or provide financially for their child.  There was no difference in the likelihood that children would live with both parents at three years.  The only outcome for which there was a marginally positive impact was in in regard to behavior problems.  Children of BSF couples were slightly less likely to report behavior problems, though there was no impact on emotional security.

There were also differences between 15 and 36 months, by site and by sub-group.  By the three-year follow-up, most of the negative impacts in Baltimore were gone way (that is, there were no impacts); similarly, most of the positive impacts in Oklahoma City had also disappeared by this point.  In Oklahoma, the one positive impact that remained was in terms of family stability.  More children of BSF parents in Oklahoma City reported living with both of their biological parents.  In Florida, fewer couples in the BSF group were romantically involved at the three-year mark than in the control group.  In the other sites, the null findings discussed above held.  There were no varied impacts by sub-group (including by race) at the three-year follow-up.

SOURCES FOR MORE INFORMATION

References

Wood, R. G., McConnell, S., Moore, Q., Clarkwest, A., and Hsueh, J. (2010). The Building Strong Families Project: Strengthening unmarried parents’ relationships: The early impacts of Building Strong Families. Mathematica Policy Research Report, 1-31.

Wood, R. G., Moore, Q., Clarkwest, A., and Killewald, A. (2014). The long-term effects of Building Strong Families: A program for unmarried parents. Journal of Marriage and Family, 76, 446-463.

Websites:

http://www.acf.hhs.gov/programs/opre/research/project/building-strong-families

http://www.buildingstrongfamilies.info/

http://www.mathematica-mpr.com/~/media/publications/PDFs/family_support/BSF_36month_impact_ES.pdf

http://www.mdrc.org/project/building-strong-families#overview

Contact Information

Mathematica Policy Research

Office of Planning, Research, and Evaluation – Administration for Children and Families (OPRE/ACF)

KEYWORDS: infants, toddlers, young adults, males and females, urban, community-based, parent or family component, parent training/education, family structure/marriage

Program information last updated on 8/4/16.