Why did you decide to target mid-level officials in the Peru reform?
In Peru, 50,000 schools with half-a-million teachers provide a service to nine million kids every single day. They are working in collaboration with local authorities to provide the right structure and all the inputs that those schools need, and those local authorities are working with the minister. At different levels, each official has to do their job for the whole [education system] to operate smoothly. So, we asked, how can we design a structure where additional resources incentivize local governments to complete their administrative work based on certain standards and on time.
How did you decide on which results, or "performance commitments", local officials should try to achieve?
The goal was to improve the quality of the delivery process. Identifying areas where we had to make progress and then how much progress to ask local authorities to make as a condition of the grant is kind of an art. It's not simple.
It was not "first they had to achieve learning results"…no. It was a little more humble than that. For instance, every year, local governments must write up contracts for temporary teachers. This has to happen a couple of months before the start of the school year for those teachers to be in school. So that became a condition: write these contracts and hire those teachers on time. If the local governments complied, they got a monetary incentive.
Another challenge was that textbooks had to be in schools before the school year started in March. For that to happen, books must be designed, printed, and distributed at the provincial and district levels, then distributed to schools, and teachers have to use them. That was not happening – sometimes the books arrived in August. So those processes had to be fixed.
Explain your approach to setting targets for the results you chose.
For all results-based financing, you target certain levels of improvement – your improvement indicator. The delta has to be ambitious but not impossible. For instance, Peru needs to increase teacher attendance in schools. But if attendance is 50 percent, you cannot ask a province to move to 90 percent in three months. But you can ask them to go to 70. In this first semester, you need to improve that; in the second semester, you need to improve more.
How did you link funding to results?
If the local government only achieved some of the results, they received a smaller share of the additional grants [on top of the usual education funding they received]. If they achieved all, they got a larger pot. Local governments could use those resources freely for education. Money that can be used at the discretion of local authorities or principals is extremely valuable. Typically, in lower-middle-income countries, most funding goes to [teacher] salaries and there's no flexibility. In this case, they could choose to buy a printer or train teachers or refurbish the school – whatever they wanted. That was a very powerful incentive. To have results-based frameworks work and be a real incentive for local governments even a small part of the resources must be flexible.
Results-based finance doesn't work in a vacuum. Can you talk about elements that made Peru's reform successful?
The first condition for any results-based framework is a very good information system, because RBF relies on people doing something they can show. If we're talking about writing teachers' contracts on time, you need to know who's been hired and when that contract is signed. Peru is not a gigantic system, but if you're hiring thousands of contract teachers, it's not trivial to get that information flowing. If you don't have that, it's impossible to condition any resources upon that result.
What about capacity-building?
You have this carrot in one hand, and in the other hand, it's not a stick but support. Local governments in poor areas might have smaller administrative capacity to improve delivery or comply with administrative procedures. You need to support them because, at the end of the day, you want everyone to comply with better standards.
What happens after they meet the targets?
A good results-based framework will move to other, more sophisticated indicators. Going back to book distribution, let’s say you manage to get books delivered on time, which is not an easy feat – but we managed to do it in Peru – let's then worry about other outcomes. Maybe condition resources aimed at making sure kids do not drop out, and later start to think about linking more resources to learning outcomes.
What's your message to the Bank's government partners, ministers, and others about how RBF can support their efforts to ensure children get a good education?
A key challenge in education is to improve the quality of the service. We know how a good school works. You need to have teachers who are well trained and who have a curriculum, equipment, the right material and books, and the right engagement with kids. The challenge is you need to replicate that in thousands of schools every day.
Using a results-based framework helps to instill a culture of service that complying with procedures matters, complying with certain intermediate outcomes matters, complying with the final results you're looking for matters. So that is the final objective – that everyone in a complex, large, difficult-to-manage system internalizes that they are there to provide a high-quality service so that all kids can have a good education.