Scaling up skills

Anindita Bhattacharjee and Md. Kamruzzaman, two researchers of BRAC's Research and Evaluation Division, recently completed a research on impact of skill development on livelihoods of the recipients. It is based on a pilot project initiated by BRAC in 2012: 'Skills Training for Advancing Resources (STAR)'. Leaving aside the question of representativeness of the results for a moment, the paper warrants attention on the heels of heightened demand for skill training for raising productivity of the labour force. The policy implications of the positive results would possibly indicate that such skills should be scaled up across the country.
As we all know, Bangladesh is currently passing through a demographic dividend that is projected to continue till 2040. But despite large supply of working age population in the pipeline, low productivity seems to hold Bangladesh back. Age-disaggregated data shows that about one-thirds of the country's population falls within the age group of 10-24 years but a substantial segment lacks basic education and competencies required for 'growing up as competent adults with high employability'.  It is thus no wonder that to solve this problem effectively, immediate attention needs to be given to skill development of the labour force mostly by promoting vocational training. Arguably, providing them with basic education along with appropriate skill training can help achieve the goal of developing a highly productive work force. However, it is heartening to note that in recent years, government and non-government organisations/institutions alike have put a lot of emphasis on development interventions aimed at 'skilling up' the adolescent population in Bangladesh. Be it for migration within or outside the country or for self-employment, the role of training for raising productivity can hardly be overemphasised. These interventions have focused on various aspects of individuals' lives. Significant achievements have already been attained in the fields of health, nutrition, education, etc. Remarkable progress has been made in increasing primary enrolment rates as well as in achieving gender parity in primary enrolment. But these achievements are counter-balanced by high dropout rates and poor quality of in-class education.
For quantitative surveys, the sample size calculation for the treatment group involved the following considerations: aiming at 95 per cent confidence level and a precision level of ±4 per cent for a total population of 1000 (intervened children in the pilot phase of the project), the required sample size turns out to be about 375. For the treatment group, we rounded up this number to 400. For the purpose of assessing impact of the project, the researchers also needed to have a suitable comparison group. Therefore, they surveyed 600 non-participants from the same communities so that they have a decent size comparison group after using Propensity Score Matching (PSM) method for identifying the best matched/suitable comparison group. In this way, their total sample size for the quantitative survey became 1,000. This total sample of 1,000 children were interviewed with a structured questionnaire in round 1 survey; for the round 2 survey, only about 75 per cent of them could be revisited as the rest had mostly migrated to different areas and could not be located. But majority is from the comparison group, leaving us with two rounds of information on a satisfactory number of treatment and comparison group individuals (treatment- 348, comparison- 399; total- 747). Throughout the report, statistical significance of the difference between the treatment and the comparison groups has been determined using t-test statistics.  
It is found that prior to project intervention, about 28 and 38 per cent of the participants and non-participants were involved with different types of Income Generating Activities (IGAs). In the post-intervention period, the proportions were found to be about 99 and 61 per cent respectively for participants and non-participants. The difference-in-difference (DID) value for IGA involvement is found to be statistically highly significant. This implies that the project has been highly successful in developing the skills of the participants leading to their involvement with different types of IGAs in the post-intervention period. The next set of indicators read by the research shows the types of IGA involvement of the working children in both pre and post- intervention period. Majority of the working children from both the groups are found to be involved with skilled labour activities with the proportion being much higher among the treatment group in the post-intervention period (i.e. in 2013). The proportion of working children involved with self-employment has increased over time for both the groups, with the increase again being significantly higher among the treatment group. For both skilled labour and self-employment, the highly significant positive DID values show promising evidence of the project's impact on the participants' type of employment. Moving on, the proportion involved in services has gone down for treatment group while it has increased for the comparison group over time. The instances of day labouring were highly rare among both the groups in 2012, and became even lower among the treatment group in 2013 while becoming slightly higher for the comparison group.
For an understanding of the project's overall impact on household wellbeing, the researchers analysed information related to household income. Mentionable changes are noticed among the two groups with regard to per capita annual income. In 2012, the average annual per capita income of households from the comparison group was found to be about Tk 3,713 more than that of households from the treatment group. But the direction of the difference was found to have changed drastically particularly in 2013 when the average annual per capita income of households from the treatment group was found to be about Tk 2,529 more than that of households from the comparison group. The DID value of Tk 6,241 is found to be statistically significant, indicating that the average annual per capita income of the treatment households has increased substantially compared to that of the households from the comparison group during the study period.
The last issue to be analysed and discussed directly concerning the participants is their future plans and aspirations that they have shared with the qualitative research team. Most of the children expressed having a strong desire to establish own business in respective trade in future. Some of them want to go abroad as skilled workers, while some others want to continue working in the country but at a better position and with good salary. But they all have realistic expectations and understanding of their own abilities as all of them stated at one point or another that they understand they do not yet have the necessary level of skills to pursue these dreams right away, and that it can be attained only through continuous effort to learn more and to prastice what they have learnt. Many of these children expressed that they want to continue working under the MCPs for achieving more skills. They also mentioned their plans to save money from their earnings in order to initiate something of their own in future. This enabling of the participants to have a dream for a better future and realistic plan to make that dream come true appears to be one of the greatest achievements of the project.
The writer is a former Professor of Economics at Jahangirnagar University and currently Acting Chair, Economics and Social Sciences Department (ESS),
BRAC University.
[email protected]