• -0.35,1.00
  • -0.10,4.00
  • 0.24,1.50
  • 0.19,1.00
  • 0.12,3.00
  • 0.15,4.00
  • 0.04,4.00

Jobs First was a statewide welfare reform initiative in Connecticut that promoted work and encouraged people to seek, obtain, and retain employment.

Jobs First was a statewide welfare reform initiative in Connecticut that promoted work and encouraged people to seek, obtain, and retain employment.

Jobs First was one of the demonstration projects made possible by Section 1115 waivers to the rules in effect at the time for the Aid to Families with Dependent Children (AFDC) program. These Section 1115 waivers allowed states to test new approaches to advance the objectives of the AFDC program.

Jobs First had three main features. First, participants could receive no more than 21 months of cash assistance. They could receive 6-month extensions if they attempted to find employment but their family income was still below the threshold. Second, Connecticut disregarded participants’ earned income below the poverty level when calculating cash grants and food stamps benefits to encourage and reward work. This meant that people could receive earnings from work and continue to receive cash assistance income. In addition, Jobs First participants could have $3,000 of assets, compared with just $1,000 under AFDC in 1996. Finally, Jobs First participants were required to look for a job on their own or through training courses that taught job search and job maintenance skills.

If participants still could not find a job, Jobs First provided additional education and training services. People who did not search for a job, secure a job, or participate in education and training services were subject to sanctions. The first time they failed to meet requirements, participants’ payments were reduced by 20 percent for three months. Payments were reduced by 35 percent the second time participants did not meet requirements and were cancelled after the third time.

Jobs First participants could receive a maximum of 21 months of services, unless they qualified for the extension. All applicants to and recipients of AFDC in Connecticut were eligible for Jobs First.

Year evaluation began: 1996
Populations and employment barriers: Cash assistance recipients
Intervention services: Academic instruction, Financial incentives, Sanctions, Training, Soft skills training, Job search assistance
Setting(s): Tested in multiple settings

Effectiveness rating and effect by outcome domain

Back to top
View table help Need more context or definitions for the Outcome Domain table below? View the "Table help" to get more insight into terms, measures, and definitions.

Scroll to the right to view the rest of the table columns

Outcome domain Term Effectiveness rating Effect in 2018 dollars and percentages Effect in standard deviations Sample size
Increase earnings Short-term Supported favorable $2,677 per year 0.128 4803
Long-term Supported favorable $837 per year 0.040 4803
Very long-term No evidence to assess support
Increase employment Short-term Supported favorable 8% (in percentage points) 0.186 4803
Long-term Supported favorable 5% (in percentage points) 0.133 4803
Very long-term No evidence to assess support
Decrease benefit receipt Short-term Not supported unfavorable $955 per year 0.347 4803
Long-term Not supported unfavorable $286 per year 0.104 4803
Very long-term No evidence to assess support
Increase education and training All measurement periods Little evidence to assess support favorable 2% (in percentage points) 0.032 772

Participant race and ethnicity
Black or African American
White, not Hispanic
Hispanic or Latino of any race
Another race

Implementation details

Back to top

Dates covered by study

Jobs First started operating in January 1996. The evaluation enrolled program participants from January 1996 to February 1997, and the final report assesses outcomes four years after participants enrolled. Connecticut still operates a Jobs First program under the Temporary Assistance for Needy Families program but introduced changes to the time limit and benefit extensions in October 2001, after the evaluation.

Organizations implementing intervention

Connecticut’s Department of Social Services (DSS) operated Jobs First.

Populations served

Jobs First served applicants and recipients of AFDC. Individuals who did not have a child younger than age 1 were randomly assigned to either Jobs First or traditional AFDC when they came to the DSS office to apply for AFDC benefits or recertify their benefits eligibility.

The study focused on two public benefits offices (New Haven and Manchester). Across the two sites, 39 percent of participants were Black, 38 percent were White, and 22 percent were Hispanic. The majority (88 percent) had work experience, but more than 40 percent had never worked full time for a single employer for 6 months. Further, less than 20 percent had earned $5,000 or more in the 12 months before enrolling in the study. Two-thirds of participants had a high school diploma or GED or higher level of education. The average age of participants was about 30.

Notable differences were present in demographic characteristics and work experience between the participants at the two benefits offices, which the study describes further.

Description of services implemented

Jobs First was created under the federal Section 1115 waivers to test new approaches to achieving AFDC’s goals. Jobs First replaced Connecticut’s AFDC program with Temporary Family Assistance (TFA) and introduced two main policy changes, in addition to employment services for participants:

  • Benefits time limit. TFA recipients could receive benefits for 21 months, after which their benefits were discontinued. There were two exceptions to the time limit policy: (1) families in which all adults were exempt from participating in employment-related activities were exempt from the time limits, and (2) recipients who reached the time limit despite efforts to find employment and whose family income was below the maximum monthly grant for their family size could receive 6-month extensions of their benefits. There was no limit on the number of extensions families could receive.
  • Financial work incentives. Participants’ earned income was not included when calculating a family’s monthly benefit amounts. Families would receive their full TFA benefit amount until they earned more than the federal poverty level for their family size, after which they would stop receiving benefits. This earned income disregard meant benefits were not immediately reduced or replaced by earned income. Jobs First also increased participants’ asset limit for eligibility from $1,000 under AFDC to $3,000 under TFA.

Jobs First’s employment services included the following:

  • Job search. The initial program model intended for participants to first engage in self-directed job search. If participants did not report employment after three or six months, they would be called to the DSS office to verify with eligibility staff that they had been searching for a job. In practice, staff did not often assign participants to self-directed job search because many participants were not searching for jobs on their own and staff did not feel comfortable leaving participants unmonitored for so long.

  • Job search assistance. Case maintenance workers referred participants to an employment services orientation, where staff from the Connecticut Department of Labor (DOL) or Regional Workforce Development Boards (RWDBs) met with recipients individually to determine their first activity. Participants considered job ready worked with DOL staff to find a job. Participants with barriers to employment worked with the RWDB or one of its contracted providers to engage in job-readiness activities, which might include training or education. DSS staff could refer participants who were unsuccessful in their job search to Job Search Skills Training, a program consisting first of two weeks of classroom instruction in job-searching and job-holding skills and, second, a monitored job search.

  • Supportive services. Jobs First participants received child care assistance if they were engaging in work activities while receiving cash assistance. They could continue receiving child care assistance after leaving welfare for employment if their income was below 75 percent of the state median income. Jobs First also helped pay for transportation for participants engaging in employment activities.

  • Sanctions. Jobs First participants who did not comply with employment activity requirements and did not have an acceptable reason for noncompliance could face sanctions, which included a temporary reduction or suspension of benefits. DSS staff referred such participants to a caseworker with the Connecticut Council of Family Service Agencies. This caseworker could develop an individual performance contract with participants that outlined agreements for complying with an employability plan.

Service intensity

Participants were expected to engage in job search activities until they found a job. Although the program was largely self-directed, participants could access more intensive employment supports as needed. Slightly more than half of participants (53 percent) reached the 21-month benefits time limit during the study’s four-year follow-up period. About 29 percent of study participants continuously received benefits throughout their time in the program, reaching their time limit 21 months after random assignment. The remaining 24 percent reached the benefits time limit within 22 to 48 months after random assignment, meaning they received temporary exemptions from Jobs First or temporarily stopped receiving benefits during that time.

Comparison conditions

The study used a randomized controlled trial design to compare employment and earnings outcomes of Jobs First participants to people who participated in the traditional AFDC program, which had no time limit on cash assistance receipt. The comparison group also received more individualized activity assignments. However, AFDC recipients had a smaller earned income disregard and received just one year of Medicaid coverage after leaving the AFDC program for work, whereas Jobs First participants were eligible for coverage for two years. AFDC recipients also received child care and transportation support but with different limitations.


Various organizations subcontracted by DSS provided employment services to program participants. In 1998, DOL and the RWDBs and their contractors started administering employment services.

MAXIMUS, a private for-profit company, verified eligibility for child care assistance and issued subsidy payments for eligible participants starting in September 1997.
Child Care Infoline, a toll-free telephone information service, helped parents find child care providers.

United Way of Connecticut, a community foundation, and the Connecticut Council of Family Service Agencies, a network of family service providers, offered services for finding employment and overcoming barriers.


Specialized intake staff processed applications for public benefits and described the Jobs First program to applicants.

Case maintenance workers conducted initial assessments of participants, referred them to DOL for employment activities, oversaw the sanctioning process if participants did not comply with employment-related mandates, and administered food stamps and Medicaid to families. Case maintenance workers served both Jobs First and AFDC participants. In the Manchester office, 11 case maintenance workers each carried a caseload of about 200 cases. In New Haven, 40 case maintenance workers each carried a caseload of 120 to 150 cases.
For the first year and a half, a group of DSS staff members assigned recipients to employment activities and monitored compliance. In mid-1998 when DOL took over responsibility for employment-related services, DOL and RWDB staff assumed these responsibilities.

Social workers in each DSS office assisted participants facing homelessness or other challenges.

The study authors did not include information on staff’s training, degrees, or certifications.

Local context

Jobs First was implemented throughout Connecticut, but the evaluation focused on two public benefits offices, New Haven and Manchester. In 1990, New Haven was one of the poorest cities in the country, with about one fifth of residents living below the federal poverty line. The New Haven office served about 22 percent of the state’s TFA caseload. The Manchester office served a more suburban clientele. The poverty rate for Manchester’s coverage area in 1990 was 3.9 percent.

Fidelity measures

The study did not discuss any tools to measure fidelity to the intervention model.

Funding source

Jobs First was funded by Connecticut’s federal AFDC grant dollars.

Cost information

The total cost per program participant over a five-year follow-up period was $8,040 (in 1999 dollars). This figure accounts for program operating costs and supportive services paid by Jobs First agencies and external agencies while participants were in the program, including case management, employment services, and transitional child care assistance. It also includes costs for employment and supportive services after participants exited the program and stopped receiving benefits, including ongoing child care assistance. The cost of supporting a Jobs First participant was $2,252 (1999 dollars) more than for supporting an AFDC participant.

A benefit-cost analysis found that Jobs First participants received an average of $5,651 more in income and services (in 1999 dollars) than AFDC participants over the five-year follow-up period. Costs to the government from Jobs First in terms of program operations, payments, and services were $4,222 higher per Jobs First participant (in 1999 dollars) than for AFDC participants over the follow-up period. For every dollar the government invested in Jobs First, participants gained about $1.30 in benefits. Jobs First resulted in a net benefit of $1,250 (1999 dollars) to society as a whole, which assesses the combined financial perspectives of program participants, payroll taxes paid by employers, and the government budget.

Studies of this intervention

Back to top
Study quality rating Study counts per rating
High High 1