92255 from EVIDENCE to POLICY Learning what works for better programs and policies October, 2014 TURKEY: Do job training programs help people find jobs? Jobs are critical to ending the cycle of poverty because they and improving employment opportunities is part of this. give people the opportunity to support themselves and their In Turkey, researchers from the World Bank worked with families. But new entrants into the labor force or the un- the government to evaluate the impact of the Turkish employed may not have National Employment Agency’s (ISKUR) vocational the right skills or train- training program to reduce unemployment. The impact ing to fill the jobs that are evaluation found that vocational training didn’t improve SKILLS AND JOBS available. Policymakers people’s chances of being employed three years later, focused on helping people despite a small short-term improvement the first year. move into the workforce The evaluation, one of first randomized experiments often turn to vocational of a large-scale training program for the unemployed, training and retraining also found that trainings offered by private providers, programs. But do these particularly private providers facing more competition, had actually improve people’s a larger impact than those from the government. Based chances of finding work? on the results, the government increased courses by Do the jobs pay enough to justify the costs of the training private providers and took steps to ensure the quality programs? Who provides the best courses—private firms or of the offerings. The government also added job state organizations? Policymakers and researchers are look- search assistance, improved employment counseling, ing for answers to these questions and more in order to un- increased efforts to enroll beneficiaries of social derstand how to create programs that work best. assistance programs who are able to work, and offered The World Bank is committed to helping countries incentives to those beneficiaries of social assistance meet the United Nations Millennial Development Goals, who actively search for a job after the course. Context Nearly a million people enter the labor force every year in aged 15-24 rose to 25.3 percent in 2009 from 20.5 percent Turkey, a middle-income country with a population of 75 in 2008, while overall unemployment was 14 percent in million. Despite a strong economy, the country has strug- 2009, compared with 11 percent in 2008. gled with reducing unemployment and improving the over- In an attempt to improve employment prospects, the gov- all labor force participation rate, which is almost 20 per- ernment expanded its free vocational job training programs centage points lower than the European Union average of to enroll 250,000 people by 2011, up from 32,000 in 2008. 64 percent. The 2009 global financial crisis worsened the The training courses, offered in all of Turkey’s 81 provinces, situation in Turkey (as it did elsewhere), especially for job- are available in a wide range of fields, including services, me- seekers under the age of 25. Unemployment among youth chanics and computers. The courses are offered by both pri- vate and public providers and generally run for 6 hours a Average Applicant Profile: day, over a three-month period. Participants receive a small 27 years old daily stipend of 15 Turkish liras (about $10). Prerequisites 73 percent completed high school for enrolling include being at least 15 years old and having 61 percent female a primary school education, although specific courses may 63 percent have worked and have 3 years experience Only 12 percent worked even one hour in the previous month have specific requirements. Evaluation The trainings were so popular that in many cases, more number of applicants for whom they had slots, giving the people registered than there were slots. Researchers were researchers the opportunity to create control and treatment able to use the oversubscription as a means for carrying out groups. After being stratified by gender and age (under or SKILLS AND JOBS a randomized control trial to measure the program’s effect. over 25 years old) the applicants were randomly assigned Researchers started with 39 provinces that had at least two a slot, no slot, or they were waitlisted. Of the 5,902 ap- significantly oversubscribed courses. The 39 provinces plicants, 3,001 were assigned to the treatment and 2,901 to were divided into two groups—those above and below the control groups. The evaluation covered courses that ran the median unemployment rate at the time—and then 10 sometime between the last three months of 2010 and ended provinces were randomly picked from each group. Three no later than May 2011. A baseline survey was conducted other provinces, Antalya in the south and Gaziantep and either before the course started or within 11 days of it hav- Diyarbakir in the southeast, were included at the request of ing started. The follow up took place about a year after the the government agency to represent varying labor market specific course ended. Researchers did a further follow up conditions. almost three years after the end of each course by tracking In the 23 provinces, the evaluation team worked with participants through worker payments filed with the nation- regional employment officials to identify courses that were al social security system. likely to be oversubscribed and they included a mix of pub- None of the applicants were told about the evaluation. lic and private providers. In total, 130 courses were iden- Instead, they told they were being surveyed as part of an tified. Providers were asked to pick at least 2.2 times the effort to improve training services. Findings Participants in the training program were slightly The effects of the program were small but significant in more likely than people who didn’t get a slot to the first year. There was a three percentage point increase be formally employed one year after finishing in formal employment—compared with a 29 percent for- the course. But that didn’t last: After three years, mal employment rate for the control group—among those there was no difference between those who who took a training course. This disappeared by the third received the program and those who didn’t. year, when data showed that about 66 percent of people *This policy note is based on “The Impact of Vocational Training for the Unemployed: Experimental Evidence from Turkey,” Sarojini Hirshleifer, David McKenzie, Rita Almeida, Cristobal Ridao-Cano, Economic Journal (forthcoming). The impact evaluation received funding from SIEF, from the Turkish National Employment Agency, and others. Access the full paper at: https://drive.google.com/file/d/0B9C9RwWKZrUNd2JGZ2lqbHE4LVk/edit?usp=sharing from both groups had held a formal job at some point in 72 percent completed their courses. Although initial the post-training course period. analysis relying on administrative data from the social security office showed a longer-term positive impact on Initially, people who took courses provided by employment for courses with higher attendance, this didn’t private firms appeared to have better employment hold up on further analysis. rates than people who were in state-run programs, but this disappeared at the three year mark. When balanced against each other, the earnings of people who went through the training didn’t justify Training through private courses resulted in a four to the cost of the training. six percentage point increase in employment after one year, as compared with public courses. But this increase The cost of a training program was about 2,692 Turkish lira narrowed to 2.5 percentage points at the three year (about $1,600) for each person in a public course and about mark, which wasn’t statistically significant. Researchers 2,429 Turkish lira (about $1,800) for the private course, also noted that private course providers self-select who enrolls in their classes, making it hard to draw any interpretations from even this short-term gain. Training also didn’t give participants higher expectations for the future. One year after the courses ended, both the control and treatment groups were fairly optimistic about their futures, believing that their households would be better off economically five years later. They also shared the same outlook on their own chances of employment, with 54 percent of the control group and a similar percent among the treatment group thinking that they would have a job in two years’ time. The lack of impact didn’t have anything to do both of which includes the small daily stipend participants with course duration or staff quality. received. People who went through the training did show a slight increase in later earnings of an extra 26 Turkish lira The fact that the courses didn’t have significant impacts on (about $15) a month, which isn’t statistically significant. employment isn’t because courses were too short. In fact, At this rate, it would take 93 months for the income gain longer courses had less impact than short ones, possibly to offset the cost of the course. But as the impact of the because people were so busy getting trained that they had training, which includes the income gain, doesn’t last more little time to apply for jobs. The evaluation also found than one and a half years, the cost of the course wouldn’t that the impacts had nothing to do with the education or be made up by extra earnings. experience of the instructors. The private courses did a little better in the cost-benefit analysis. People enrolled in those courses earned on average Nor did this indicate that people weren’t attending an extra 66 Turkish lira (about $39) a month, and the classes. impacts from private courses lasted about two and a half years. The cost still outweighed the benefit, but based on Around three-quarters of those selected for training this, the employment agency has put more emphasis on programs continued to attend after the second day, and courses provided by private firms. Conclusion As policy makers continue to develop programs for the term effectiveness. It also underscores the importance of unemployed, the findings from this impact evaluation private training programs—which offer key incentives, can highlight some of the challenges—as well as as well as the ability to respond quickly and efficiently to opportunities—of vocational training programs. Indeed, market demands. The lessons from this impact evaluation the findings point to the program’s ability to provide will help policy makers in Turkey and elsewhere move short-term relief to people who need it, but as a whole, forward to create even more effective programs that give this evaluation serves as a cautionary tale about long- unemployed people the skills they need to succeed. SKILLS AND JOBS The Strategic Impact Evaluation Fund, part of the World Bank Group, supports and disseminates research evaluating the impact of development projects to help alleviate poverty. The goal is to collect and build empirical evidence that can help governments and development organizations design and implement the most appropriate and effective policies for better educational, health and job opportunities for people in developing countries. For more information about who we are and what we do, go to: http://www.worldbank.org/sief. The Evidence to Policy note series is produced by SIEF with generous support from the British government’s Department for International Development. THE WORLD BANK, STRATEGIC IMPACT EVALUATION FUND 1818 H STREET, NW WASHINGTON, DC 20433 Produced by the Strategic Impact Evaluation Fund Series Editor: Aliza Marcus; Writer: Daphna Berman The original had problem with text extraction. pdftotext Unable to extract text.