Evaluating the Impact of Restore Hope’s 100 Families Initiative on Families Referred through Differential Response in Arkansas

The intervention evaluated in this study is a referral to Restore Hope Arkansas’s 100 Families initiative, which provides intensive case management and coordinated connections to community-based services for families who have been screened into the Arkansas Differential Response system. Families in the treatment group will receive a formal referral to 100 Families. Restore Hope case managers will then contact the family, complete an intake assessment, and begin monthly check-ins focused on goal setting, service navigation, and ongoing support. Services may include referrals for housing assistance, employment support, education and training, mental and behavioral health care, substance use treatment, transportation assistance, and other wraparound resources. Engagement and service delivery will be tracked through Restore Hope’s HopeHub platform, which facilitates communication and coordination across partner agencies.

Families assigned to the control group will continue to receive standard DR services through the Arkansas Division of Children and Family Services. Standard services include a safety assessment conducted by DR staff, limited case management, and referrals to community providers based on assessed needs. These services reflect the existing child welfare prevention pathway available to all DR families. Control group families will not receive a systematic referral to 100 Families, although they are not prevented from accessing 100 Families on their own if they choose to do so.

The goal of the intervention is to determine whether providing structured, proactive, and coordinated case management at the point of DR intake improves family stability and reduces future child welfare involvement. The study examines whether a referral to 100 Families increases program take-up, enhances access to community services, and improves outcomes related to child safety, housing, employment, education, health, and economic well-being. The intervention seeks to provide evidence on whether community-based case management can strengthen early prevention efforts within the child welfare system.

2025-11-17

2034-09-30

The primary outcome is whether the family enrolls in Restore Hope’s 100 Families program following referral. Enrollment is defined as having successful contact with Restore Hope after referral and completing an intake assessment.

Program take-up will be defined as a binary indicator equal to 1 if the family completes enrollment in 100 Families or 0 if the family does not complete enrollment in 100 Families after being randomly assigned to receive a referral.

Secondary outcomes will focus on child welfare involvement, family stability, and economic and well-being outcomes. Child welfare outcomes will include any repeat child welfare hotline call, any substantiated maltreatment finding, any new DCFS case opening, and any child removal within specified follow-up windows. Family stability and economic outcomes will include housing stability (such as entries into shelters or HMIS records of homelessness), employment status and quarterly earnings, school attendance and chronic absenteeism for children, and receipt of public benefits including SNAP, TANF, and SSI. Additional secondary outcomes will include health care utilization such as emergency department visits and inpatient hospitalizations, as well as criminal justice involvement such as arrests, court cases, and incarceration. If funded, survey-based secondary outcomes will focus on domains not captured in administrative data, including self-reported family well-being, stress, access to services, material hardship, and other indicators of household stability.

These outcomes will be constructed primarily from Arkansas administrative data sources, including DCFS records, state employment and wage records, education data, HMIS housing data, Medicaid or all-payer claims data, benefits records, and criminal justice data. For child welfare outcomes, repeat hotline calls, substantiations, case openings, and removals will be coded as binary indicators measured over pre-specified follow-up periods (for example 6, 12, and 24 months after randomization). Housing instability, health care utilization, benefit receipt, and criminal justice involvement will similarly be coded as binary indicators of any event within a given follow-up window, with additional continuous or count measures where appropriate such as number of ER visits or total days incarcerated. Employment and earnings outcomes will use quarterly employment indicators and continuous earnings measures. Survey-based outcomes, if implemented, will be constructed from parent or guardian responses, with individual items analyzed separately and, where appropriate, combined into indices that average standardized items within domains such as material hardship or family well-being.

This study uses a randomized controlled trial to evaluate the impact of Restore Hope’s 100 Families initiative on families referred through Arkansas’s DR system. The main research questions are: (1) Does a referral to 100 Families increase take up and persistence in 100 Families services among DR families, and (2) Does a referral to 100 Families reduce subsequent child welfare involvement and improve family stability outcomes such as housing and employment? Study enrollment will begin with a pilot period that runs from November 17, 2025, through February 27, 2026, with the intention to extend to a larger full study period.

Families will be eligible for the study if they are accepted into Arkansas DCFS’s DR pathway, are a parent or legal guardian of at least one child under age 18, have household income at or below 200 percent of the federal poverty level, and are a United States citizen or lawful permanent resident. Among eligible DR cases in counties served by 100 Families, the DR manager will randomize families at the case level to either a treatment group or a control group. In the pilot, families will be assigned to treatment or control with equal probability within each study county that has 100 Families capacity. Treatment families will have a standardized note placed in the DR case directing staff to “Refer to 100 Families,” while control families will have a note instructing staff to “Follow regular protocol but do not refer to 100 Families.” All families, regardless of assignment, will continue to receive the existing standard DR services from DCFS.

Randomization will be implemented using a pre-specified rule based on DR case numbers to ensure that assignment is systematic and not influenced by staff discretion. After randomization, DCFS will transmit to the research team a file of eligible DR cases that includes a study identifier, treatment assignment, and limited baseline information required for analysis and for linkage to other administrative data sources. Outcome measurement will rely entirely on administrative records from DCFS, Restore Hope, and state agencies, including child welfare outcomes, 100 Families engagement, and longer term outcomes related to housing, employment, education, benefits receipt, health care use, and criminal justice involvement. If funding is secured, a brief follow up survey of parents or guardians may be added to measure additional family well being outcomes. This randomized design will provide causal evidence on whether referrals to 100 Families improve stability and reduce deeper child welfare involvement for families at a critical point of contact with the system.

Not available

Families will be randomized using a predetermined rule based on the last digit of the DR case number. Odd case numbers will be assigned to one study arm and even case numbers to the other, ensuring a simple and unbiased assignment process that does not rely on staff discretion. Given that DR case numbers are equally likely to be even or odd, this method should lead to a 50-50 split between treatment and control.

Unit of randomization: Family (DR case).

No

750 families (DR cases referred through Arkansas’s Differential Response system)

750 families (DR cases referred through Arkansas’s Differential Response system). Some analyses will be at the child level and thus will have greater than 750 observations.

375 families assigned to treatment (referral to 100 Families), 375 families assigned to control (standard DR services without referral to 100 Families).

The pilot RCT will enroll approximately 750 families, with 375 assigned to treatment and 375 assigned to control. The primary outcome is a binary indicator for 100 Families service enrollment. Using a two-sided test with a 5 percent significance level, 80 percent power, a baseline enrollment rate of approximately 10 percent in the control group, the study is powered to detect a minimum detectable effect size of 6.98 percentage points in enrollment. This corresponds to an increase from about 10.0 percent of enrollment in the control group to 16.98 percent enrollment in the treatment group. For this binary outcome, the implied standard deviation in the control group is approximately 0.30.