This post is coauthored with Alaka Holla
The rigorous evidence on vocational training programs is, at best, mixed. For example, Markus recently blogged about some work looking at long term impacts of job training in the Dominican Republic. In that paper, the authors find no impact on overall employment, but they do find a change in the quality of employment, with more folks having jobs with health insurance (for example).
We get similar results in a recent paper with Kevin Croke on a training intervention in Nigeria, but the “quality” of the job is where it gets interesting. We set out to look at occupational sex segregation and how training might overcome it. Given what we heard from talking to firms and women during initial fieldwork, a big issue for us was to understand the pre-existing biases of the applicants themselves.
In order to measure these biases, we turned to the implicit association test. A small but growing group of economists have used this tool from social psychology but for those of you who don’t know what this is, the IAT is a computer based sorting task that measures the time required to associate different groups with different concepts. Using one of our examples, if it is easier to mentally pair pictures of men with words associated with an office and women with words associated with the home, then subjects should be able to make these pairings faster than the opposite pairings (women and the office and men and the home). Because the differences in sorting times are often less than a second, these associations are considered implicit and automatic, or beyond conscious control. Indeed, they often differ from explicitly expressed attitudes or biases. Females, for example, often exhibit stronger implicit attitudes linking males with career and females with family than males, despite reporting weaker explicit attitudes. For our work in Nigeria, we measured not only the office vs. home biases, but also office vs. trading and professionalism vs. unprofessionalism. If you’d like to try a different version of this test for yourself, this site will let you take sample tests.
Before getting to the results, a bit about the intervention. We were working with a World Bank project which was funding a number of activities to help jump-start the nascent ICT/BPO sector in Nigeria. For the evaluation, we focused in on a training program which used standards and curricula developed in India to train and certify recent university graduates for jobs in this sector. We worked with the training providers to randomize applicants into a treated (trained) group and a control group.
So what did we find? As indicated above, the training program on average had no significant impact on the likelihood of being employed or the levels of earnings. But it did induce a fair amount of sectoral switching – trained individuals were 1.7 percentage points more likely to have a job in the ICT sector. This might seem small, but remember this is a nascent sector – only 6 percent of the control group works in IT. So some switching.
But now for the biases. We found that women who were implicitly biased against women’s being professional switched at a much higher rate – 3 times higher than neutral women. In contrast, explicit (i.e. what they said in response to survey questions) biases against professionalism had no significant interaction with treatment. And other potentially more fundamental sources of bias, like whether women are better suited to be traders or should even be in the workplace at all, also showed no significant impacts. So the training really worked to overcome participants deep seated or subconscious biases.
What do these results tell us? In addition to imparting some skills, training can clearly open up the possibility set of participants to new careers. While in this case it had an effect on average, it had a much bigger effect on women who might have talked themselves out of this type of job in the first place. It’s important to note there were no special gender themes in the training or in the recruitment, so it was simply the expanded horizons brought on by the training that enabled these biased women to switch into a male dominated (in Nigeria) sector. How exactly this changed their minds is a topic for future research.
The rigorous evidence on vocational training programs is, at best, mixed. For example, Markus recently blogged about some work looking at long term impacts of job training in the Dominican Republic. In that paper, the authors find no impact on overall employment, but they do find a change in the quality of employment, with more folks having jobs with health insurance (for example).
We get similar results in a recent paper with Kevin Croke on a training intervention in Nigeria, but the “quality” of the job is where it gets interesting. We set out to look at occupational sex segregation and how training might overcome it. Given what we heard from talking to firms and women during initial fieldwork, a big issue for us was to understand the pre-existing biases of the applicants themselves.
In order to measure these biases, we turned to the implicit association test. A small but growing group of economists have used this tool from social psychology but for those of you who don’t know what this is, the IAT is a computer based sorting task that measures the time required to associate different groups with different concepts. Using one of our examples, if it is easier to mentally pair pictures of men with words associated with an office and women with words associated with the home, then subjects should be able to make these pairings faster than the opposite pairings (women and the office and men and the home). Because the differences in sorting times are often less than a second, these associations are considered implicit and automatic, or beyond conscious control. Indeed, they often differ from explicitly expressed attitudes or biases. Females, for example, often exhibit stronger implicit attitudes linking males with career and females with family than males, despite reporting weaker explicit attitudes. For our work in Nigeria, we measured not only the office vs. home biases, but also office vs. trading and professionalism vs. unprofessionalism. If you’d like to try a different version of this test for yourself, this site will let you take sample tests.
Before getting to the results, a bit about the intervention. We were working with a World Bank project which was funding a number of activities to help jump-start the nascent ICT/BPO sector in Nigeria. For the evaluation, we focused in on a training program which used standards and curricula developed in India to train and certify recent university graduates for jobs in this sector. We worked with the training providers to randomize applicants into a treated (trained) group and a control group.
So what did we find? As indicated above, the training program on average had no significant impact on the likelihood of being employed or the levels of earnings. But it did induce a fair amount of sectoral switching – trained individuals were 1.7 percentage points more likely to have a job in the ICT sector. This might seem small, but remember this is a nascent sector – only 6 percent of the control group works in IT. So some switching.
But now for the biases. We found that women who were implicitly biased against women’s being professional switched at a much higher rate – 3 times higher than neutral women. In contrast, explicit (i.e. what they said in response to survey questions) biases against professionalism had no significant interaction with treatment. And other potentially more fundamental sources of bias, like whether women are better suited to be traders or should even be in the workplace at all, also showed no significant impacts. So the training really worked to overcome participants deep seated or subconscious biases.
What do these results tell us? In addition to imparting some skills, training can clearly open up the possibility set of participants to new careers. While in this case it had an effect on average, it had a much bigger effect on women who might have talked themselves out of this type of job in the first place. It’s important to note there were no special gender themes in the training or in the recruitment, so it was simply the expanded horizons brought on by the training that enabled these biased women to switch into a male dominated (in Nigeria) sector. How exactly this changed their minds is a topic for future research.
Join the Conversation