An interview with former Peruvian Minister of Education Jaime Saavedra

In his three years as Education Minister of Peru, Jaime Saavedra oversaw wholesale reform of the school system. He not only increased the level of spending on education but also changed how teachers and principals were evaluated, compensated, and promoted. He reformed and simplified the curriculum, led the shift toward a full-day school for secondary age students, put in place basic maintenance budgets for school facilities, built the capacity of the ministry to deliver on its stated objectives, and maintained teacher goodwill throughout the process. The results speak for themselves. Peru’s performance in the Programme for International Student Assessment (PISA) improved dramatically during the period of his tenure from 2012 to 20151 (exhibit). He also pushed for the passing of a new university law and oversaw its implementation creating a new regulatory framework aimed at increasing quality. In this interview, Jaime Saavedra shares with McKinsey how he managed to get so much done in so little time.

Peru’s education system shows a steady improvement in investment and scoring to 2015.

Interview transcript

McKinsey: You came from the World Bank to take on the role of Education Minister in Peru in 2013. What made you make such a change?

Jaime Saavedra: I had been working at the World Bank for ten years on poverty reduction and inequality, and what I found was that the best way for a country to give equal footing to all its citizens—independent of where they were born, their race, or their gender—is education. Peru, my own country, was failing to give its citizens those opportunities. We were not investing enough. Peru was spending under 3 percent of GDP on education, compared to 4 to 6 percent in Chile, Colombia, Mexico, and Brazil. This was reflected in the salaries of our teachers. When my mother was a teacher in the 1960s, teaching was a middle-class career. But with the expansion of the system, as more families sent their children to school, there was a trade-off between quantity and quality. Enrollments went up, but expenditures per pupil went down, and teacher salaries plummeted. By 2010 teacher salaries were about a third in real terms of what they had been in the early 1970s. But it was not only salaries; incentive structures and mind-sets also had to be changed.

McKinsey: You arrived with a clear sense that the system needed to change. Was there broad support for reform across the government?

Jaime Saavedra: When I took the job, it was with a few preconditions. First, that we would need to make a big push on education reform and increase education financing. Second, that I didn’t want any political conditions in terms of staffing. I wanted complete freedom in filling the most important positions in the ministry. The President and Minister of Finance both agreed to these conditions.

McKinsey: So, you had top leadership support. Was there broader impetus for change?

Jaime Saavedra: I was lucky in some ways. The PISA results came out three weeks after I started, so those results were a baseline for us. Peru was last. Last of the 65 countries of PISA. Not bottom 10 percent. Last. And that was a shock. We could have decided to play the results down to say, “But Peru has improved since 2009” (which it had) or “But we are better than many countries that didn’t even take PISA” or “It is an OECD [Organisation for Economic Co-operation and Development] examination that is alien to our culture and priorities.” We have seen that in other countries, and some countries have left PISA after bad results. But we didn’t go down that route. Instead we decided to own the problem. To use these results to say, “Look, we’re not in trouble. We’re in deep trouble.” Then education was on the front page. Education is never on the front page of the newspaper.

McKinsey: How did you capitalize on that moment?

Jaime Saavedra: It was important that we explained the reform to people in such a way that normal human beings could understand both the problems and our approach. When I arrived at the ministry, there was a huge chart from the National Education Plan that had been produced in a two-year process with 25 commissions and 100 meetings throughout the whole country, consulting everyone about how to reform the system. It was, like, 200 balls that linked one to another. “OK, so that’s the plan.” It was detailed and thorough and complete. But who will understand that?

We had to go back to a very basic level and do basic things right. We said, so this is a car, and for this car to move, you need four wheels. The four wheels must move at the same time. One wheel is teachers. One wheel is learning. Another wheel is infrastructure. The fourth wheel is management. Simple: teachers, learning, infrastructure, management. Now, within each one of those pillars, there were extremely complex processes, policies, and decisions. But not all of the public is generally interested in that. That is for the bureaucrats and experts.

McKinsey: That sounds simple, but teachers, learning and pedagogy, infrastructure, and management—those are each big things. How did you break each down into clear goals and priorities and, as you say, focus on the processes, policies, and decisions inherent within each wheel?

Jaime Saavedra: Across each of these elements, we had a few clear priorities. For teachers, we needed to make sure teachers were better paid, that they were motivated and equipped to do their job. Teachers had to understand that their job was making sure that kids learn. That was not happening. For many teachers—not all—the thinking went, “If the kid learns, great. If he doesn’t, I don’t know. That’s not my problem. My job description is to teach, not to make sure that he learns.” These two statements are very different, right?

For learning, we needed to simplify the curriculum and ensure teachers had the right instructional materials, including lesson plans and the right in-school support. We also needed to move to full-day secondary school. That was a big shift. For infrastructure, we were talking about the basics: running water, toilets, buildings that were not falling apart. The quality of the infrastructure was, and still is, very unequal, in a very unequal country. When I say management, I’m talking about two things: management of the system, and management of the schools—selecting the right people to be principals and supporting them so that they could do their job well. What was needed was not rocket science. But implementing that across the whole system is a challenge.

This may sound like a lot, but we were focused on basic education in the first year. Than we added university reform to the menu. Big time. Because of this, I did less than I would have wanted on technical education or secondary rural education, for example.

McKinsey: So, you have your goals—how did you make sure that you had the major stakeholders on your side: teachers, principals, unions?

Jaime Saavedra: Obviously the union was initially skeptical. Here I am, an economist from the World Bank. I must be an imperialist envoy coming here to privatize schools!

I was very clear that my priority was to improve the quality of the public education system, focusing initially on teacher salaries and teacher career paths. I had discussions with them saying, “Look, we agree. You’re telling me that teachers’ salaries are low? You don’t have to convince me that your salaries are low. Let’s work together. Let’s raise salaries—but we need to do that in a way that has an impact on learning.”

McKinsey: It sounds like you got the union on board by raising salaries. What did you ask for in return?

Jaime Saavedra: We increased salaries, and that required intense discussion with the Ministry of Finance, but most of those increases were given dependent upon performance. That was the key. So now we had the problem of how to measure teacher performance. We could not rely on principals, because they hadn’t been chosen meritocratically either (we were just starting that process). So, we devised a teacher’s exam. Are exams the only and the best way to assess teacher performance? No. Not at all. But it was a first step. Over 200,000 teachers took this exam in over 60 cities in the country, all on the same day. The key thing was to make sure that the exam was transparent, that it was fair, that there was no leakage of the answers to the tests—so that we could then award salary increases based on a fair objective measure. We did the same thing with principals. We made them take an exam to pass a certain bar, and after two years, we had 15,000 principals who had been selected meritocratically.

I remember congressmen and politicians coming with lists of people from their region saying, “Look, these are fantastic teachers or fantastic principals. I mean, they are great people. You need to make sure that they are promoted.” This was always how promotions happened in the past. What was game-changing about the tests is that I could honestly say “Look, there’s nothing that I can do. I don’t write the tests. I don’t know the answers. The evaluation of the scores is done automatically. I have no idea who will be promoted. There’s no way that I can do anything, even if you’re telling me that this is a fantastic principal. Hopefully, if he’s fantastic, hopefully he passes the exam, but there’s nothing I can do.” It was an important way to give a signal that this was all about merit.

McKinsey: What did the union, the teachers, and the principals think of that?

Jaime Saavedra: The law that introduced meritocracy in hiring and promotion had been passed in 2012. Obviously, the union had opposed this, as one could expect. But they were pragmatic. I mean, teachers were getting salary increases, and frankly the PISA shock had weakened any possible resistance to the implementation of the law. As for the teachers—we also reached out directly to them. You have to understand that people’s perceptions about teachers had hit rock bottom. You would hear on the radio that “teachers are dumb and unprepared, and it is their fault that our schools are failing.” I would never say things like that because in most cases it is not true (although as in any profession, there are bad performers). I believe that teachers are the key partners and main actors of any education reform. Constantly, constantly, constantly: the key partners of the reform are teachers. We had to change negative perceptions and the reality underlying them: that teaching was the easiest profession to get into, that it was the least demanding career, that anyone with poor grades could always go into teaching. Providing salary increases for high-performing teachers was just the beginning. We had a team working directly on issues related to teachers’ welfare. We put in place a coaching program to raise capabilities, pairing more experienced teachers with newer teachers.2 Similarly, with principals, newly selected ones received training on school management and school leadership, to build their capabilities.

We communicated directly with teachers. At the beginning of the school year, we sent out a text message to 180,000 teachers across Peru saying, “Maria, you are critical for education in Peru. We count on you to make sure that our students are the best. Signed, Jaime.” I remember going to schools in the Andes, and people showing me their cell phones with it. I would say that, in the end, maybe half of the teachers had a good perception of me, because they had a good increase in their salaries, while the other half did not, because they got zero. Such is life.

McKinsey: You mentioned going to schools in the Andes. How often were you on the road visiting schools?

Jaime Saavedra: A lot. I must’ve made 70 trips at least by myself, in addition to those I made with the President. Many of my visits were unannounced, because if you announce that you’re coming, you don’t always see the real picture. The principal was alerted a few hours before we arrived. That way we could learn what was actually happening on the ground.

McKinsey: It sounds like you had a strategy of accountability paired with capability building, for both teachers and principals: identifying and rewarding strong performers but also providing training, coaching, and support. Can you talk more about the scaffolding you provided to teachers to build capabilities in the system?

Jaime Saavedra: Another of the big things we did for teachers was to provide a curriculum that was simpler and more intuitive, then we developed lesson plans to enable teachers to deliver that curriculum. This meant that less capable teachers were still able to provide a basic level of quality. This was again criticized by the academic education community as “hampering professional creativity” but lesson plans were voluntary, and teachers loved them. Many teachers worked half a day as a teacher, then half a day in another job just to make ends meet. They didn’t have time to be writing lesson plans while they were driving a cab or heading off to a private school for their second job.

McKinsey: How about capabilities in the broader ministry and administration? Did you need to build capability for change there as well?

Jaime Saavedra: Oh yes. When I arrived in Peru, the Ministry of Finance staff had a joke about the Ministry of Education. They called it “the bank” because they could store money there across fiscal cycles. They would allocate money to the Ministry of Education, knowing that they would not be able to spend it. Then they could allocate it elsewhere at the end of the year.

It was important for me to build a multidisciplinary team able to deliver on the reform program. Luckily, about 60 percent of the staff in the ministry were not tenured employees, and I ended up changing about 55 of the top 60 positions in the ministry. This didn’t happen overnight—the ministry had to keep on functioning—but over time we brought in a lot of new young talent, and that made a big difference.

I recruited people with a public management background: economists, engineers, a few key hires from the Ministry of Finance. The average age went down dramatically as I hired well-motivated young people in their 30s and 40s from top schools. I didn’t bring in a lot of private-sector talent, because it was critical that my team understood how the public sector operates. I did bring in some though for a few key areas. For example, our programs to evaluate teachers and principals created thousands of lawsuits as individuals sued us for losing their jobs. Every time we considered a new intervention—teacher incentives, salaries, promotions—we needed to check with the lawyers and understand the implications.

McKinsey: How did this impact the pace of reforms?

Jaime Saavedra: We couldn’t slow down. I always told people that in principle we should be here until the end of the government—two-and-a-half years. However, we needed to think always as if we only had six months left. Because this is politics, anything could happen. We could be out tomorrow.

For example, one of the major shifts we made was moving secondary school to a full day. The history of this was that as enrollment rose in the 1970s we didn’t have capacity for everyone to attend school, so we split the school day: one shift from 7:30 a.m. until 12.30 p.m., another from 1:00 p.m. to 6:00 p.m., and sometimes even a night shift.

We decided that we needed to make the investment to switch back to full-day secondary school to give enough time for learning. The first proposal of the staff, backed by some in the academic community, was that we should start with a pilot of 50 schools, then an impact evaluation of the pilot that would take three years, then rolling out and scaling up once we had a tested model. By which time momentum would have stalled. We needed to start with something that was irreversible. We started the first year with 1,000 schools, and then 600 the following years, until we could reach all 8,000 secondary schools in Peru. Now, were we going to make mistakes in the first 1,000 schools? Undoubtedly. But if you do 1,000 schools, then you’ll have to continue. Because parents will start saying “I want that” and the change becomes embedded. Reforms of this type require adaptive learning. You design, implement, monitor, evaluate, learn, redesign, and so on.

McKinsey: Talking of change becoming embedded, to what extent do you think your reforms will be sustainable? You are now back at the World Bank, heading up the Education Global Practice there. Since you left office, there have been four education ministers in under three years in Peru. Will your country continue on its trajectory of improvement?

Jaime Saavedra: It’s hard to say. My administration almost entirely matched a PISA assessment cycle. In that time, we achieved the fourth largest improvement globally, and the largest improvement in Latin America. I remember we were able to say we beat Brazil in math. We cannot beat them in soccer, but we can beat Brazil in math. The other big reform front we opened was university reform. Enrollments had increased dramatically, mostly in private universities, but quality was low and extremely heterogenous. The implementation of the law implied a change in authorities in all public universities and a new regulatory agency.

So how do we make sure that the path continues? Many of the changes were enshrined in law. But the law is irrelevant sometimes, because laws can be changed, or laws be ignored or not complied with. So, laws do not guarantee anything. Ideally, parents would see the positive changes and use their political power to ensure continued progress. But a problem I see is that the voice of parents is not loud enough today. Parents need to speak up more to demand that their children get the best possible education. Parents and society as a whole need to continue to demand continued improvement—only that will hold leaders accountable and ensure sustained change.

Explore a career with us