Anu Rangarajan

1

Targeting Employment andJob Retention
Servicesto Welfare Recipients

• Anu Rangarajan
• Prepared for the Welfare to Work The Next
  Generation,A National Forum
• November 17, 2003
• Newfoundland, Canada

2
Overview of Talk

• Background Welfare reform in the U.S. context
• The challenge of job retention and advancement
• Targeting analysis and example
• Framework for designing new programs
• Retention and advancement strategies

3
Welfare Reform Has Led Many toEnter the Labor
Force in the US

• Increased emphasis on work since mid-1980s
• PRWORA legislation of 1996
• Work requirements
• Time limits (five years)
• Strong economic conditions during the mid- and
  late-1990s
• Over 1 million welfare recipients entered the
  labor market

4
Although Many Have Found Jobs, Sustained
Employment is Challenging

• Low-paying entry-level jobs
• Service/sales/administrative support
• Many jobs do not offer fringe benefits
• Seasonal/temporary is common, as is part-time
  work
• Nearly one-third are nonstandard hours

5
Employment Patterns Are Diverse

• Employment spells are short
• 60 to 70 percent end within a year
• First six months after job start is critical
• Reentry is common, but it takes time
• Employment experiences are diverse
• 30 to 50 percent stay employed for long periods
• Others cycle in and out of jobs or return to
  welfare

6
Diversity in Wage GrowthPatterns as Well

• On average, people progress to better jobs over
  time
• But, there is diversity in their experiences
• Some experience wage growth, while others
  experience no increase in wages or reductions in
  wages
• Some evidence of greater wage growth in more
  recent periods, but still most remain low-wage
  workers

7
Increased Focus on Job Retention andAdvancement
Programs in the US

• States just starting to experiment with job
  retention and advancement strategies
• Programs not sure whom to serve, what services to
  provide
• Often costly, and large numbers are entering the
  labor force
• Question How to best allocate limited
  resources?

8
Is Targeting a Feasible Strategy?

• Need a range of services Core services to more
  intensive services
• Will need oversubscription of services
• Is it even feasible? Can we identify those who
  are in greatest need? If so, how do we go about
  it?
• Used extensively in UI worker profiling

9
Step 1 for Targeting

• Define outcomes and goals that describe risk
  status
• NLSY example
• High risk Employed less than 70 percent of
  weeks during five-year period

10
Step 2 for Targeting

• Identify individual targeting characteristics
• NLSY example
• Age less than 20
• Employed less than half of the weeks during prior
  year
• Has no high school diploma or GED
• Has preschool child
• Received less than 8 per hour
• Has no fringe benefits
• Has no drivers license
• Has health limitations

11
Step 3 for Targeting

• Create decision rules using data on a sample of
  the caseload
• Target on a single characteristic
• Target on multiple characteristics

12
Step 4 for Targeting

• Assess performance of decision rules and select
  the best ones
• Two main performance criteria
• Large percentage of targeted cases should be
  high-risk cases
• Decision rule should perform better than randomly
  selecting cases for services

13
Steps 5 and 6 for Targeting

• Step 5 Select the numbers and types of clients
  to serve
• Step 6 Time the identification of clients for
  targeting

14
Targeting Analysis ResultsUsing the NLSY Data

• Single characteristics can be identified
• Has health problems has no high school diploma
  age younger than 20
• Programs can do better by using a combination of
  characteristics
• Over 90 percent are high-risk, if serve 20
  percent
• Over 80 percent are high-risk, if serve 50
  percent
• 66 percent are high-risk in sample

15
Checklist for Multivariate Targeting
Points if Has Barriers
Characteristic Age less than
20 2 Employed lt1/2 weeks in prior year 2 Has
no high school diploma 3 Has preschool
child 2 Received less than 8 per hour 2 Has
no fringe benefits 5 Has no drivers
license 2 Has health limitations 5 Total
Score
16
What Targeting Does andWhat It Does Not Do

• Only identifies those in need
• Allows us to rank the order of need programs can
  determine which segment to serve
• Does not always tell us what they need
• Does not tell us about the effectiveness of
  services

17
Guidelines for Program Design
1. Whom to serve? 2. What services to
provide? 3. How to deliver services? 4. How
much to promote services? 5. When to provide
services? 6. Who will run the programs?
18
Strategies to Promote Retentionand Advancement

• Case-management-based approaches
• Focus on initial job placement
• Broad systems-related changes
• Make work pay strategies
• Financial and nonfinancial incentives to affect
  behavior
• Employer involvement/sectoral-based strategies
  

19
Making Case Management Effective

• Contact clients early as problems begin during
  the first few weeks after job start
• Need to be flexible and creative, as engaging
  clients may be difficult
• Links with employers and other agencies
• Convey clear and consistent expectations
• Modest caseload sizes for intensive case
  management
• Integrate pre- and post-employment services

20
Focus on Initial Placement

• Good initial job match is important
• Provide soft skills training
• Basic skills and vocational training in areas of
  high demand
• Place clients where employer sponsored training
  available
• Career placement assistance

21
Broad (Systems-Related) ChangesMay Be Useful

• Seamless child care funding and availability
• Transportation subsidies in rural and suburban
  areas
• Housing subsidies
• Access to other public assistance programs (food
  stamps and health insurance)

22
Other Retention Strategies

• Make work more attractive by providing wage
  supplements
• Financial and non-financial incentives to affect
  behavior
• Encourage employer involvement and tailor
  programs to meet employers needs
• Create opportunities for effective mentoring/job
  coaching
• Opportunities for education and skills advancement

23
Designing Programs

• Assess needs for your own population
• Tailor services to meet clients needs
• Provide easy access to job search assistance and
  placement
• Provide opportunities for skills enhancement
• Consider separate strategies for the
  hard-to-employ

24
Conclusions

• Job retention and advancement is a challenge
• Efficient allocation of limited resources suggest
  targeting may be a desirable strategy
• Targeting is feasible job characteristics are
  good predictors
• States/provinces will have to experiment with
  various service strategies
• Good opportunity to learn from others and start
  building a knowledge base