Throughout the Middle East, unemployment rates of educated youth have been persistently high and female labor force participation, low. Researchers from the World Bank studied the impact of a randomized experiment in Jordan designed to assist female community college graduates find employment. One group of graduates was given wage subsidy vouchers that could be redeemed by their employers for up to six months for a value equivalent to the prevailing minimum wage; a second group was invited to attend 45 hours of soft skills training; a third group was offered both interventions; and the fourth group formed the control group. To conduct the study, researchers chose eight public community colleges with the largest female enrolment numbers. Four colleges were in Central Jordan (Amman University College, Princess Alia University College, Al-Salt College, Zarqa University College) and four in Northern and Southern Jordan (Al-Huson University College for Engineering, Irbid University College, Ajloun University College, and Al-Karak University College). Four individual level survey questionnaires were administered during the impact evaluation study. The baseline survey was conducted in July 2010, the midline in April 2011, the first endline was carried out in December 2011, and the second endline in January 2013.

Abstract

Throughout the Middle East, unemployment rates of educated youth have been persistently high and female labor force participation, low. Researchers from the World Bank studied the impact of a randomized experiment in Jordan designed to assist female community college graduates find employment. One group of graduates was given wage subsidy vouchers that could be redeemed by their employers for up to six months for a value equivalent to the prevailing minimum wage; a second group was invited to attend 45 hours of soft skills training; a third group was offered both interventions; and the fourth group formed the control group.

To conduct the study, researchers chose eight public community colleges with the largest female enrolment numbers. Four colleges were in Central Jordan (Amman University College, Princess Alia University College, Al-Salt College, Zarqa University College) and four in Northern and Southern Jordan (Al-Huson University College for Engineering, Irbid University College, Ajloun University College, and Al-Karak University College).

Four individual level survey questionnaires were administered during the impact evaluation study. The baseline survey was conducted in July 2010, the midline - in April 2011, the first endline was carried out in December 2011, and the second endline - in January 2013.

Geographic coverage

Amman, Salt, Zarqa, Irbid, Ajloun, and Karak.

Analysis unit

Female community colleges graduates from the class of 2010.

Kind of data

Sample survey data [ssd]

Sampling procedure

Researchers chose the eight public community colleges with the largest female enrollment numbers, which comprise over 85% of total female public community college enrollment. In July 2010 just before final graduation exams, data collectors conducted baseline surveys for most of the 404 male and all of the 1,776 female second-year students from these eight colleges. In August 2010, the researchers merged the baseline data with administrative data on examination results, which revealed that 324 men and 1,418 women passed their examinations. Of the 1,418 women who passed their examinations, the researchers randomly assigned by computer 1,349 of these graduates to be in the experimental sample. However, two of these graduates were male but incorrectly recorded as female. They were subsequently dropped from the sample.

The experimental sample of 1,347 was stratified into 16 strata and randomly assigned by computer into three treatment groups and a control group. The strata were created based on the following four characteristics: whether or not (1) the community college was in Amman (Amman, Salt, and Zarqa) or outside Amman, (2) an individual's Tawjihi examination score at the end of high school was above the sample median, (3) an individual indicated at baseline that she planned to work full-time and thought it at least somewhat likely that she would have a job within 6 months of graduating, and (4) she is usually permitted to travel to the market alone. Within each of the 16 strata, 22.2% of the students were allocated to receive the wage voucher only, 22.2% allocated to receive the soft skills training only, 22.2% allocated to receive both, and 33.3% allocated to the control group. This resulted in 299 or 300 in each treatment group, and 449 in the control group.

Sampling deviation

The only deviation from the sample design involved dropping two graduates from the sample because they were incorrectly recorded as female.

Mode of data collection

Other [oth]

Research instrument

All questionnaires were initially developed in English and subsequently translated into Modern Standard Arabic.

The questionnaire design process was based on standard labor force survey questions, academic literature on well being, mental health, and female empowerment, and inputs from Al Balqa Community Colleges, the Chamber of Commerce, the Ministry of Planning and International Cooperation, the Social Security Corportation, the Department of Statistics, Dajani Consulting, Business Development Center, and local firms.

The questionnaires were piloted and adjusted accordingly in each survey round.

Response rate

In the midline, first endline, and second endline surveys, researchers successfully followed up with 92%, 96%, and 92% of graduates in the sample, respectively. In the first and second endline surveys, the team collected a portion of the survey data (3% and 9%, respectively) by proxy through their relatives. This survey experienced very few problems with outright rejections to answer the survey questions although the proxy responses reflect graduates or their families refusing to allow the graduate respond for herself. The vast majority of attrition comes from disconnected cell phones and the inability to completely track individuals down.

The attrition rates are low and slightly vary by treatment status. The wage voucher group has the lowest attrition (3% midline, 1% 1st endline, 4% 2nd endline), which is likely due to the additional information gathered through monitoring the voucher usage. On the other hand, the control group experienced the highest attrition (11% midline, 7% 1st endline, 11% 2nd endline), which is likely because there was no additional contact with the control group outside of the surveys.