81683 R e s u lt s S e r i e s   |   O C T 2 0 12 Can skills training programs increase employment for young women? The Case of Liberia Y oung people age 15 to 29 make up about a quarter of the world’s population, yet they con- stitute nearly half of the world’s unemployed. The The first AGI pilot—the Empowerment of Adolescent Girls and Young Economic Women (EPAG) project—was launched in Liberia World Bank is helping to increase viable employ- in late 2009. Preliminary results from the midline ment opportunities for youth. In many countries, survey show that EPAG has been very successful restrictive gender norms make it harder for girls to in achieving its primary objectives—increasing access training and employment opportunities. To employment and earnings among young women. ensure that girls and young women are included The magnitude of the results is impressive when in this agenda, the Bank launched the Adolescent compared to findings from other youth training Girls Initiative (AGI) in 2008. The program is being programs in developing countries. It is expected piloted in eight low-income countries—including that successful economic empowerment programs some of the toughest environments for girls. Each like EPAG can also indirectly bring about positive intervention is tailored to the country context, behavioral changes and provide spillover benefits and includes an impact evaluation to build the for the families and communities of trainees. evidence base to help adolescent girls and young women succeed in the labor market. THE EPAG PROJECT Youth unemployment is one of the main obstacles unsafe for young women; they can be expected to to development in Liberia. The country has a very exchange sex for grades or depend on older boy- young population—more than 60% of the popula- friends to finance their education. Gender-based vul- tion is under the age of 24—many of whom grew nerabilities are severe in Liberia; approximately 30% up during the Liberian conflict (1989 to 2003).1 of 15 to 24 year-old females have experienced physi- There are few opportunities to enter in wage cal violence, and in 2007 rape was the most com- employment; most of Liberia’s youth earn income monly reported crime.2 These issues are increasingly on a day-to-day basis by trading on local markets well-recognized by the Government and gender and or as day laborers in precarious conditions. youth are positioned prominently in the country’s The situation of young women is particularly poverty reduction agenda. challenging. Boys’ education tends to be prioritized EPAG implementation is led by the Liberian by Liberian families and domestic duties (includ- Ministry of Gender and Development (MoGD) ing childcare) limit young women’s available time with a team of service providers. The program for school and work. Moreover, schools are often consists of six-months of classroom training 1 United Nations Population Division: http://esa.un.org/ 2 The Population Council and UNFPA. 2007 Demographic unpd/wpp/index.htm and Health Survey: Adolescent Dataset. 2 A G I   |   R e s u l t s Serie s EPAG’s innovative design elements: i. EPAG is designed around girls’ needs: service providers hold morning and afternoon sessions to accommodate participants’ busy schedules; trainings are held in the communities where the girls reside; and every site offers free childcare. ii. Participation is incentivized creatively through attendance prizes, contests, business plan competitions, etc. iii. Girls receive support throughout the followed by six-months of placement and support program from volunteer EPAG coaches (including micro-enterprise advisory services and who are based in the classrooms, as well internship and job placement assistance). The aim as through peer groups of EPAG girls. is to smooth the transition from the classroom to iv. The M&E framework includes frequent wage or self employment. 70% of girls are trained and unannounced visits to ensure that in business development skills (BDS)3 and 30% service providers maintain a high- in job skills (JS)4 targeted to sectors with high quality learning environment. demand for workers.5 All JS trainees also receive v. Performance bonuses are awarded training in entrepreneurship skills. All participants to training providers that successfully receive life skills training specifically designed for place their graduates in jobs or “[My favorite part Liberian girls. In addition, trainees receive small micro-enterprises. of the training] is stipends contingent upon classroom attendance painting, because and are assisted in opening savings accounts at Recruitment—particularly for younger girls who it is strange to see local banks. Girls who complete the training are were both not in school and literate—proved chal- young women awarded a small bonus ($20 USD). The EPAG proj- lenging and the age limit was raised to 27. EPAG ect also includes a capacity-building component originally recruited 2106 trainees.6 painting in Liberia.” for the Ministry of Gender and Development that The EPAG impact evaluation employs a ran- EPAG Graduate has resulted in a newly created Adolescent Girls domized pipeline research design in which train- Unit to help mainstream gender-sensitive pro- ees were randomly assigned to receive training in gramming and policies in Liberia. either the first round (March 2010 to February 2011) EPAG was originally targeted to young women or second round (July 2011 to June 2012). Impact who were i) between age 16 and 24; ii) possessed is defined as the change in outcomes between the basic literacy and numeracy skills; iii) were not in time the program starts and six months after the school (although night school was okay); and iv) classroom training ends, as compared to a statisti- resided in one of nine target communities in and cally similar control group (the second round train- around Monrovia and Kakata City. ees). 1273 trainees were were randomly assigned Community mobilization campaigns to the treatment group and 769 to the control. The began in September 2009 and trainee recruit- results presented here are based on two rounds ment took place from December 2009 through of quantitative surveys—a baseline conducted in January 2010 in all nine target communities. 2010 and a midline survey conducted six months after the classroom-based phase of the training 3 BDS includes training in entrepreneurship principles, program ended in 2011. market analysis, business management, customer service, money management, and record-keeping. 4 When possible, the participant’s track preference was honored; however, there was excess demand for the JS track so many participants were placed into BDS. 6 Additional recruitment was done in 2011 because the 5 1) hospitality, 2) professional cleaning / waste project was designed to train 2500 girls. Participants management, 3) office / computer skills, 4) professional recruited in the second batch, some of whom also received house and office painting, 5) security guard services, and additional literacy training, are not included in the impact 6) professional driving. evaluation. worldbank.org/gender/agi 3 THE FINDINGS were self-employed, Figure 1. Likelihood of Working and only 12% were 0.7 wage-employed. 0.67 Percent with at least 1 IGA 0.65 The EPAG pro- “With the skills I got I 0.6 gram increased aver- am sure that if I apply 0.55 age weekly income Treatment for work in my area 0.5 by 115%.8 Between 0.45 0.45 Control baseline and midline, of training I will get 0.4 0.39 the treatment group’s the job, but if I do not 0.35 0.38 average weekly earnings find a job I can try to 0.3 increased by 948 LD start my own business Baseline Midline (approximately $13.54 USD) more than the with the business control group’s earn- training I got.” EPAG achieved remarkable program participa- ings. Given that the commonly accepted minimum EPAG Graduate tion. The classroom training was held from March wage in Liberia is 25 cents per hour, an additional through August, during which the project achieved $13.54 USD per week is a large sum of money. a 95% retention rate (far higher than similar pro- Considering only those who are working, the aver- grams in Liberia and elsewhere) and a high average age earnings of the treatment group do not sig- attendance rate of 90%. Anecdotally the high par- nificantly differ from those of the control group. ticipation is attributed to EPAG’s innovative design In other words, the impact on earnings is mostly elements (see box above). driven by the large increase in employment rather The program led to a 50% increase in employ- than the earnings per se. ment among trainees, compared to those in the EPAG significantly increased girls’ savings. control group. Before the first round of training, the At midline, the treatment group on average had a treatment and control groups were approximately total of $44 USD more in savings than the control equal: about 38% of young women in both groups group. Given that the program helped girls open reported being engaged in at least one income-gen- an account at a local bank, it is not surprising that “I used to braid erating activity.7 Both groups improved by the mid- savings in banks increased. In addition to saving hair and just eat line, but the change was significantly larger among in banks, the treatment group also reported hav- the treatment group. The control group improved by ing increased savings in informal savings clubs. the money without nearly 20% (from 38% to 45%), and the treatment The increase in savings amounted to more than the thinking about the group improved by more than 70% (from 39% to trainees earned through the program stipend and future, but now 67%). This means that the program led to about a the completion bonus of $20 USD. I consider saving 50% increase in employment. The midline survey showed no significant The positive employment outcome was changes to borrowing or lending. Only 4% of as being important much stronger among the business skills train- midline survey respondents reported having any for the fact that ees. The employment rate at midline was 76% loans, with no difference between treatment and saving helps me to among BDS trainees and 52% among JS trainees. control groups. Although these results are pre- have some money Both JS and BDS graduates were more likely to liminary, they indicate that the economic improve- be involved in self-employment at the time of the ments in terms of earnings and savings have thus when things midline survey than in wage employment. 50% of far not led to an increased rate of borrowing or become tough.” JS graduates who reported working were involved lending. This finding will be explored further in EPAG Graduate in self-employment (the other 50% were engaged the end-line analysis.9 in wage employment). For BDS graduates, 87% 7 An income-generating activity encompasses the full range 8 Respondents reported their cumulative earnings across of activities through which people earn money, e.g. wage all income-generating activities over the past week, in employment (jobs), self-employment (either in a small addition to the value of an in-kind compensation. business or through petty trade) or some combination of 9 For the end-line survey, the question has been reworded to these. explicitly ask about loans from informal sources. worldbank.org/gender/agi 4 A G I   |   R e s u l t s Serie s FROM PROGRAM RESULTS TO POLICIES The EPAG evaluation provides strong evidence amount equal to the original investment.11 For job that skills training can be an effective policy skills participants this may take as long as eight option for increasing employment among young years. The lower return on investment for the job women in Liberia. The EPAG program increased skills training is not surprising given the limited employment by 50%—a greater increase than has opportunities for wage employment in Liberia. “I am satisfied with typically been found for similar programs. Recent Given these challenges, EPAG is making design reviews suggest that training programs increase the changes to maximize employment outcomes for the training because likelihood of employment by 6 to 12%, and in rare future job skills trainees.12 the training provided cases as much as 30%, especially among young These preliminary results provide strong me with new skills and people and women, but that most programs are not evidence for continued and further investment knowledge which will significantly effective.10 The EPAG expe- be with me until I die. Preliminary Cost-Benefit Analysis rience validates the ... I am now somebody importance of deliv- Type of training Cost per Avg. monthly # Years required participant income increase to cover costs with knowledge, value ering business skills (USD) (USD) among and skills.” training along with trainees job and life skills train- Job Skills 1678 27* 8 EPAG Graduate ing in contexts where Business Skills 1221 75 2 wage employment is Assumptions: 10 percent discount rate calculated monthly; persistence of income increase limited. The positive at a constant rate; income accruing from month 7; exchange rate of 70 LD= 1USD employment results are *Note: the increase in earnings for Job Skills participants was not statistically significant at midline. driven by the business skills trainees. In con- texts like Liberia where opportunities for wage into young women’s livelihood programs in employment are limited, it is important for pro- Liberia. The results will be elaborated in the com- grams to provide basic entrepreneurship skills to ing year, and a final end-line survey will facilitate all participants and to have a realistic assump- further analysis of longer-term impacts. Further tion of how many trainees can be placed in wage analysis is planned to: employment. 1. Determine if the program changed girls’ These impressive results were achieved socio-economic behaviors and outcomes; at a relatively high per-participant cost that 2. Understand if the program expanded girls’ reflects the holistic approach of the EPAG pro- empowerment and agency; gram and the high cost of program implemen- 3. Explore how the results may differ among tation in Liberia. Program costs are approximately different types of girls in the program (by $1221 per girl for business skills training and $1678 age, community, etc.); and per girl for job skills training. Despite the high ini- 4. Gauge if the positive effects for young tial investment, the expected returns of the busi- women also benefitted their families. ness skills training are compelling. Under realistic assumptions, it will take two years for business skills participants to increase their earnings in an 10 Pablo Ibarrarán and David Rosas Shady. 2009. Evaluating the Impact of Job Training Programs in Latin America: 11 The assumptions in this analysis will need to be confirmed Evidence from IDB funded operations. Journal of and examined using the results of the end-line analysis. Development Effectiveness 1(2), pages 195-216. Betcherman, 12 In Round Three, 82% of participants will learn business skills Gordon, Martin Godfrey, Olga Susana Puerto, Friederike and only 18% will study job skills. Prior to training, EPAG will Rother, and Antoneta Stavreska. 2007. “Global Inventory of make formal agreements with select companies to provide Interventions to Support Young Workers: Synthesis Report.” training that is tailored to their specific needs. In return the World Bank. Washington, D.C. Processed. employer will commit to hiring a set number of girls. 1303749 This brief features work supported by the World Bank’s Gender and Development Department and the Africa Gender Innovation Lab, in the Poverty Reduction and Economic Management (PREM) Network . For more information about the AGI please visit www.worldbank.org/gender/agi.