Skills Development and International Development Agenda Setting: Lessons from an Intervention in Northern Nigeria

Abstract

Skills development remains on the international development agenda but fails to get adequate attention. Based on prolonged fieldwork with a particularly marginalised community of children and young adults in the northern Nigerian state of Kano, this article shows how in contexts of extreme poverty the demand for skills training can supersede that for basic education. Further, by drawing on results of a six-month-long skill-training intervention, the article documents the scope for increased experimentation in the delivery of low-cost community-based skill-training programmes and identifies factors that influence programme completion. It also demonstrates that participation in skill-training programmes can dramatically increase entrepreneurial aspirations among marginalised youth, but that without access to credit most fail to pursue their aspirations. Below certain poverty thresholds, the dire resource constraints make change in aspirations an unreliable predictor of possible improvement in future outcomes.

Le développement des compétences reste à l’ordre du jour du développement international, mais ne parvient pas à obtenir une attention adéquate. Basé sur des travaux de terrain prolongés avec une communauté particulièrement marginalisée d’enfants et de jeunes adultes dans l’état de Kano au nord du Nigéria, cet article montre comment, dans les contextes d’extrême pauvreté, la demande en formation professionnelle peut supplanter celle pour l’éducation de base. En outre, en s’appuyant sur les résultats d’une intervention de formation professionnelle d’une durée de six mois, l’article documente la possibilité d’une expérimentation accrue dans la mise en œuvre de programmes de formation professionnelle à base communautaire et à faibles coûts, et identifie les facteurs qui influent sur l’achèvement du programme. Il démontre également que la participation à des programmes de formation professionnelle peut considérablement accroître les aspirations entrepreneuriales des jeunes marginalisés, mais que sans l’accès au crédit, la plupart ne parviennent pas à poursuivre leurs aspirations. En deçà de certains seuils de pauvreté, les terribles contraintes de ressources font du changement dans les aspirations un prédicteur peu fiable de l’amélioration éventuelle des résultats futurs.

Technical Annex

Variables Influencing the Probability of Training Completion

Using the survey implemented with the youth who took part in the programme, an attempt was made to test particular variables highlighted in the survey (age, trainer quality, type of education, income, etc.) that could potentially affect the probability of a child completing the training programme. The following set of analyses attempt to identify the reasons why youth in the treatment group who received training may have failed to complete the programme. Using ‘completion of the training programme’ as the binary outcome variable, this analysis compares results from probit, logit, and LPM models. Findings do not differ greatly between models. The following tables display all three models.

Section 1: Factors affecting completion of training

Table 5 displays the results using the four variables of primary interest (Schools: Government Primary, Islamiyya, Tsangaya, and ESSPIN) as well as a number of controls which would probably have an impact on whether a child completes the training programme or not. The rationale behind these controls is presented below.

Table 5 Dependent variable – completion of training programme

Age and age2 Age could potentially have an impact on whether students complete the training or not. Younger students may be more determined than older students to learn new skills. However, age often does not share a linear relationship with most dependent variables, which is why we include age2 (the square of the age variable). In our regression model above, it is demonstrated that age shares a significant bell curve relationship with completion rate, i.e. students are more likely to complete the training programme as age increases up to a certain cut-off age, after which point they are more likely to drop out.

Education level Students with different levels of education may have different motivations for completing the training.

Family income level It is possible that as family income increases, so should completion rate.

Enough to eat Whether a child gets enough to eat or not is a reflection not only on income but also health.

Trainer quality Higher trainer quality increases the likelihood of completing the programme. This result is also significant.

Stop work for training This may alter the cost of continuing education.

The 1st, 3rd, and 5th columns show the coefficient or parameter estimates of each of the models and are interpreted as ‘the estimated effect of age on completion is 0.25 in the probit model’, which is often a less clear way of showing the impact. The 2nd, 4th, and 6th columns display the marginal effects. The interpretation of these values is the average predicted difference in the probability of completing the training between individuals born one year apart. Therefore, the average difference in probability of completing the training programme is 6 per cent for each year. As can be seen, the different models do not vary greatly in values, signs, or significance.

Section 2: Trainer's attributes affecting completion of training

Trainer Quality

In addition to the above analysis, it was also deemed fit to investigate if the quality of training, based on a number of characteristics, influence whether the almajiri students complete training courses. The survey had included a number of questions that highlight different characteristics that influence the quality of training. Table 5 notes trainer quality to have a significant impact on completion. This section therefore identifies the particular aspects of trainer quality that may have an impact on this relationship.

Where:

Questionnaire
trainer_bad_workshop	Had a bad workshop
trainer_encouraging	Was encouraging
trainer_good_workshop	Had a good workshop
trainer_knew_skills	Knew the skill techniques
trainer_regular	Regular in holding the class
trainer_scolded	Scolded a lot
trainer_shared_skills	Shared the skills fully
trainer_walking_dist	Located at a walkable distance

Note, Table 6 displays only the results of the trainer characteristics. All other variables from Table 5 are included in the models as controls but are not displayed in the table.

Table 6 Dependent variable – completion of training programme