Absence of conflict of interest.
Citation
Jacobson, L., & Petta, I. (2000). Measuring the effect of public labor exchange (PLX) referrals and placements in Washington and Oregon (Workforce Security Research Publications 2000–06). Washington, DC: U.S. Department of Labor.
Highlights
- The study examined the impact of receiving public labor exchange (PLX) job referrals and placements on the duration of Unemployment Insurance (UI) benefit receipt for UI recipients in Oregon and Washington.
- The authors used statistical models to compare two distinct treatment groups to a comparison group within each state. The treatment groups both received PLX job search assistance, a job referral, and a job offer; one group did not actually work at the position offered while the other did. The comparison group received job search assistance but not PLX job referrals or placements. The authors used data from the Oregon Employment Department and the Washington State Employment Security Department.
- The study found that individuals in the treatment group tended to receive UI benefits for fewer weeks than individuals in the comparison group; this was a statistically significant difference.
- The quality of causal evidence presented in this report is low because the authors did not ensure that the groups compared were similar before the intervention. This means we are not confident that the estimated relationships are attributable to PLX job referrals or placements; other factors are likely to have contributed.
Intervention Examined
Other
Features of the Intervention
PLX services were offered as part of the Wagner-Peyser Act of 1933. State employees through federally funded state employment security agencies implemented the PLX services.
PLX services included in-person or over-the-phone job search assistance, workshops, access to resource rooms that provide computers to prepare resumes, phones to communicate with employers, newspaper classifieds, online job boards, and a job and career planning library. Individuals could have identified promising job listings and been referred to a job. Individuals could have then pursued interviews for those jobs, completed the interview process, and received and accepted job offers. An individual was considered to have received a placement if he or she actually worked at the position offered. PLX was available to anyone in the state, and individuals can access these job openings in person or online.
In addition to these services, PLX placed requirements on UI claimants to ensure they were seeking employment and offered services to employers, such as listing job openings and hosting job fairs.
Features of the Study
This study is a nonexperimental analysis that included 138,280 unemployment spells among the 58,980 individuals who received UI benefits in Oregon in 1995, and 328,815 unemployment spells among individuals who received UI benefits in Washington between 1987 through 1995. Using data from the Oregon Employment Department and the Washington State Employment Security Department, the authors compared the duration of UI benefit receipt among the unemployment spells of two treatment groups with the duration among the unemployment spells of the comparison group. The treatment group members received PLX job search assistance, a job referral, and a job offer; those in one treatment group did not actually work at the position offered, while those in the other treatment group did. The comparison group members could have received PLX job search services but did not access job listings or receive a referral to a job. In their statistical models, the authors controlled for age, race, gender, and earnings or occupation before the intervention. Individuals could have multiple unemployment spells during the study period, so the analysis included multiple data observations for some individuals. Analysis took place separately by state.
Findings
Benefit receipt
- Individuals who received PLX job referrals but did not actually work at the position offered tended to receive UI benefits for about one-fifth of a week less in Washington but about two-thirds of a week more in Oregon relative to the comparison group. Individuals who received PLX job placements tended to receive UI benefits for about two weeks less in Oregon and about four weeks less in Washington relative to the comparison group. These were statistically significant differences.
Considerations for Interpreting the Findings
The authors did not account for existing differences between the groups before receiving PLX job referrals and placements. Individuals in each treatment group, all of whom received a job referral and offer, may differ from individuals in the comparison group in unobservable ways. These existing differences between the groups—and not the PLX job referrals or placements—could explain the observed differences in outcomes.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because the authors did not ensure that the groups compared were similar before the intervention. This means we are not confident that the estimated relationships are attributable to PLX job referrals or placements; other factors are likely to have contributed.