1) Why PSIA? Why did you choose PSIA over any other bank instrument or product?
We looked for the PSIA because we wanted to explore two things as part of this experimental pilot in Jordan which we knew was important, however we didn't have the resources in terms of being able to pay for staff, analysis, or even the data. One was looking at the effect of different labor laws and regulations in the labor market and their effect on the outcomes that we observed, and the second was to be able to compare outcomes between young boys and girls, because the Adolescent Girls Initiative (AGI) pilot was created only for women, meaning the financing was for women, the data collection was on women, but we wanted to understand whether the men would have the same outcomes we observed in the pilot. Jordan has a number of labor regulations: for instance if you are on a contract, and have a job for more than 3 months you have to formulize the contract, the employer has to give you social security benefits, which adds a lot to the cost of hiring. In the end we were able to get administrative data from the Social Security, match it to the data from the pilot, and see what exactly was happening in terms of the formulization of employment. In fact what we found was that the employment outcomes we were observing in the pilot when we compared them with the administrative data, were not formal jobs, and that’s why the employers were able to kind of let go of the employees. That allowed us to look at whether the labor market was flexible enough for this class of workers and able to absorb them because from that we basically understood the employers’ perception of the productivity of these workers was not as high as the wage they were obliged to offer under the formal regulation. So that’s why we looked at PSIA, because we wanted to understand this kind of differentiated impacts and the effects of the labor laws on this particular subset of female college graduates.
2) What was the most challenging moment while working on this PSIA?
I think the biggest challenge was actually get access to the administrative data because it’s very rare, at least in the MENA region to access that kind of data. So I think the most challenging thing was making the case to the Social Security Director that this was for a good cause, explaining why we wanted to do it, and how would have been useful for them and kind of convince them that we would treat the data very securely.
3) If you had to do this PSIA over, what would you do differently, based on the lessons learned from this one?
This pilot was the only one where we got resources for a middle-income country. AGI was the broader pilot, designed for low-income countries focusing on adolescent young girls, and we were able to successfully make the case that in this region, given the huge gender inequality observed especially in the labor market, we could do one. We got a much smaller allocation compared to other countries, so we really had to maximize what we were able to do and that’s where PSIA really came in handy because we were very budget-constrained, and the grant had to pay for everything from the design of the experiment to the implementation, including the job vouchers and the data collection. The PSIA really helped us get very critical inputs to be able to understand why we were seeing the outcome we were looking at. We had an awkward situation where we had very short-term results that disappeared and that really needs explanation, and that’s where the PSIA was very critical, so if you ask me what I would do again it’s hard to say because the PSIA was really so valuable.
4) Any recommendations for TTLs working on PSIAs – what are the top three things TTLs should always keep in mind when working on a PSIA?
What I learned personally from the process was that this is a very different PSIA from the standard ones. Typically people are looking at ex-ante or ex-post policy reforms, and this was not looking at a policy reform but at the impact of existing policy regulations. We had to think a bit outside of the box in terms of how to frame it as a PSIA, since it’s true that it was looking at the impact of a certain policy, which is the labor regulation, but it was definitely looking more at the social impact in terms of the gender aspect, than the poverty impact. So it’s not the standard energy subsidy reform for example, which is very straightforward in the line of PSIA. The most interesting thing of this experience was learning how broad this ambit can be, trying to innovate, and looking at different ways of understanding disaggregated impact of policies, of using different data sources, mixing admin data with experimental data.
5) What do you do when you are not doing a PSIA?
I work on country-level poverty work, everything from supporting statistics offices on survey data collection, poverty measurement, poverty assessment, TA, also, I have been part of the regional statistics team for data, impact evaluation, and a bunch of other things. We are also doing some work now on Energy subsidies in Iran.
Nandini Krishnan is a Senior Economist with the Poverty Global Practice. She currently works on poverty in Iraq and the Philippines, on a multi-country survey of host communities and Syrian refugees, regional and corporate initiatives for data and monitoring the Twin Goals, and supports DECRG’s Social Observatory team. She has worked on labor market issues in Egypt, Jordan and the Palestinian territories, on gender in Yemen, the Palestinian territories and the MNA region, and supported impact evaluations of large scale projects and programs in Tanzania, Nigeria, and South Africa. She holds a Ph.D. in Economics from Boston University.