Interview with F. Halsey Rogers and Nazmul Chaudhury on Absent Teachers and Medical Workers in Developing Countries

Thank you very much. We appreciate this question and all the good questions we have gotten. It's not a simple matter of rule of law, but that's certainly part of the issue. One of the problems often is the more senior staff who are absent, those are the people whom we would expect to enforce better performance, but if they're absent, you don't have them setting a good example or enforcing attendance.

Another problem is that in some cases expectations are low, even though the law is there. Some of our colleagues have done a report on Peru recently, which has shown that really expectations have declined over the past 20 years as salaries fell, and so supervisors don't feel they could enforce good behavior and good performance anymore. They don't feel they have legitimacy to do that.

And another problem is you often have union rules in place that are designed to protect teachers and medical workers from political interference in their jobs, but they have the effect of making it impossible to discipline teachers or doctors for poor performance. So, in that case, you have the law, you have the rules, but one of the rules is that you have to go through a very lengthy and difficult procedure to actually discipline staff who are not performing well.

It's mostly the poor and rural poor who are the biggest sufferers when teachers don't come to school, when doctors have and health workers don't show up to clinics. So, there's thousands and thousands of poor students and the parents and patients who face this problem on day-to-day basis, and now the issue is looking what can they do about it.

Most times they have institutional mechanisms, say, in India, where they have the right to join local government and express their concerns, and local government is then supposed to channel that back up to higher authorities. Sometimes that works, but most of the times it really doesn't. Some of the direct things that local communities have done in response to this problem is by leaving a lot of the dysfunctional public sectors in these rural areas in particular. If you look at--even in lots of parts of the world, there is creation of private schooling, it's mostly in urban areas, but again this is a signal that a lot of people are leaving the public school system. If you look at healthcare, say, in India, 80 percent of the treatments are taking place not in public health centers, but with private practitioners of varying quality.

So, one thing communities have done is by voting where their feet, and leaving the public sector, going to providers of their choice, and others by trying to bringing to bear their direct powers over providers.

Interestingly enough, in a study in Bangladesh, where we actually found that in communities that have educated mothers, and there are educated parents, their absenteeism amongst teachers was much lower, and this effect was stronger than initial supervision. So communities can play a powerful role if they get engaged.

We are very much in favor of giving teachers a reasonable salary and having them treated with respect, and this is certainly part of the answer, but it's not the whole answer. Our studies show that a big part of the problem is lack of monitoring and lack of discipline and performance, a lack of either pay or promotions based on performance. So our sense is because not all teachers are equally dedicated to their work, if you simply raise salaries, but don't improve monitoring or discipline or raise expectations of performance, you will get some teachers just taking advantage of that -- taking the higher pay but not performing any better. Certainly teachers have to be paid well, but you need to be monitoring and helping them with high expectations at the same time.

But on the question of respect, I think you have hit on something, that nonsalary factors are probably important in determining attendance. When a teacher decides whether to go to school each day, the decision is not just based on what he or she earns, but also what the work conditions are like. Does the school have decent infrastructure and equipment so that the teacher will be able to teach? Are the students motivated? Is it easy to get work? We believe that some of these factors can be quite important, and respect is probably an element in this. Unfortunately, it's not something that government can easily change, and so we are going to focus on some of these more tangible factors like the quality of the infrastructure and the school, the pay, and the monitoring.

The simple answer to this is the direct responsibility lies with the policy makers and politicians who are responsible for providing these services, health and education services in the national governments. They're the primary people who are responsible for that. Then there is responsibility in the larger sense with their international development partners, with bilaterals and multilaterals that they work with. But again, the responsibility is ultimately with the politicians who are responsible for providing these services.

Now, the thing is that in terms of who it most effective in solving this. As is pointed out, this is a highly complex problem, and there is not one magic bullet that can solve this problem. It will take a collective effort of governments, local government, and active support from international donors and other multilaterals donors.

But ultimately, what is required is recognition. If this is a problem that's recognized on a national scale, if this is a problem that is talked about in the national press, talked about local government in the sense that if there is discussion there, then at least we could try to talk about what has been tried in other place, country, what has worked. It's not about blaming anybody per say. It's first identifying the problem and see what it will take to resolve this.

And the issue about do we think it's about top management, as my friend Halsey pointed out, a lot of times it's the headmasters in school that are absent. A lot of times it's the medical officer in charge at the clinic who is absent.

So, these are the people who are not assuming their responsibility. Who is responsible for these high level staff? Is it the people in the local ministry, in the finance ministries, and that is something we are grappling with on an operational basis day to day.

This will be a question that will be in the minds of many readers. Before answering it specifically, let me pick up a theme. We don't want this to be about blame. In fact, one of the things that struck us in studying these various countries is the striking number of teachers and medical providers who do go to work each day under sometimes very difficult conditions, given that the expectations seem to be so low, that there is so little performance management, so little monitoring of the quality of their performance. And so I want to reiterate that sometimes we understand teachers and doctors are absent for reasons that they can't control, and sometimes--and often it's really the system that's at fault that doesn't have expectations of high performance, so here is one reason why people might be absent through no fault of their own. HIV/AIDS infections may either have made them ill or made a family member ill or has led to the deaths of family member or friend so they need to attend funerals. We hear these as explanations for absence particularly in Africa.

Returning to the question, the answer is no. Uganda is the one country that has a rather high prevalence rate among our six countries, and so that's the only country where we examined the issue. In that case, our colleague who did a draft paper on Uganda did an analysis and found that by and large, the illness rates that were reported were not high, and not many of the absent teachers were said to be absent because they were ill, and also the pattern of absences both regionally and across age groups didn't match very closely what we know about the pattern of AIDS in Uganda.

On the other hand, we have colleagues here, one of my colleagues in the research department and his co-authors have done a nice detailed paper on teacher absence in Zambia, and their conclusion is that probably HIV/AIDS is one of the reasons behind absence in Zambia, which is a very different setting from the six countries that we looked at. So the bottom line would be that we need to look at this on a country-by-country basis, and we need to recognize it may be a major reason for absence, but we didn't find it in the countries we are looking at.

This question comes from India and also stresses the role of local government as well as other forces. First at issue is local government. Constitutionally, in India, local governments have been at least delegated with the responsibility over primary schools and primary health clinics. But the thing is that when you dig deeper, and say you actually go down to the local government and you ask local government officials who are in charge supposedly of these primary schools and local health clinics, what is it that they have responsibility over. We find out that actually it's very little. Even though the rhetoric says they are responsible over these local facilities.

For example, take teachers. A lot of times it said that local governments have responsibility over local primary schools, but the teacher's salaries have nothing to do with it, and you have no control over what you pay the teacher, the government teachers, how you can discipline him, then in a sense responsibility is vacuous because you really don't have any real rights that you can exercise at the local level.

Same thing goes with local health centers. You don't have rights about doctors who are posted there, about their salaries, about their performance measure, so in a sense that again rhetoric doesn't match reality, and even in times when sometimes local governments are empowered with some very tangible rights, like let me give you the example of the state in India by the name of Uttar Pradash. Their local governments have the right of certifying whether the teacher is in school or not. They have to certify yes, this teacher came to school 75, 80 percent of the time, and please, now release the salary. A lot of times local governments and communities end up signing these on a pro forma basis without ever going to the school and checking if the teacher is there.

So, again it goes back to what is it that ultimately drives this. Is it giving more hours, is it acting upon those powers, this right balance between federal, national, local government. It's a very complex issue, but we welcome the role of local government as long as they are given real powers, and then the basis of which to evaluate with. There is a lot of state level initiatives going on in India which is supported by the World Bank in which, especially in the education sector. That's looking at giving better incentives and better performance measure for teachers if they prove improve their performance, come to school regular leer, and teach better, but this is still in the process of being related. Whatever initiatives we take, they have to be properly evaluated and put emphasis on learning and what works and what doesn't work.

Just to pick up on the answer Nazmul gave a moment ago, there are experiments going on, he mentioned some in India. Also on the education side we had a recent experiment carried out by one of our co-authors, Michael Kremer and some co-authors of his in Kenya, where they did a randomized experiment looking at what happens if you pay teachers a group incentive for the school as a whole -- if they're able to increase test scores, what happens to attendance, what happens to test scores. And their experiment shows it's very hard to design these experiments in a way that they will have lasting effects, and this underlines the point that Nazmul made about evaluating carefully this type of policy.

We hear about pay for performance, and it makes sense in theory, but when you look at it in practice and how it's going to be implemented, you recognize it's going to be just part of the solution, and you have to do it carefully. What Michael and his co-authors found is that when they paid for performance, they were able to lift the test scores along the dimensions that were being rewarded, but they found two things. First of all, this better performance of students disappeared very quickly once the reward was taken away; and secondly, that it seemed to be a case of "teaching to the test." The teachers knew exactly what to teach so the students would pass the test and they would get their bonus, but there didn't seem to be any more general learning going on. That's why it's very important that you implement these policies carefully. And that's why we would argue it's not just a matter of implementing this type of pay for performance (although that's probably part of the answer), but it's better overall monitoring, involvement of communities, and better working conditions -- so that teachers feel a desire to teach better, so we are not just trying to motivate them with carrots and sticks but also appeal to their professionalism, because that's a driver in the performance of these teaches and medical providers.

So, we are not going to solve this with a neat technical solution like a pay for performance contract, but a battery of policies that move in the direction of better accountability and better performance.

At one level this is already started to happen, regardless of any change in current public policy. For example, I mentioned earlier that studies show in most South Asian countries, at least, a majority of the people don't seek healthcare in public health facilities. They're already using private providers. So, what that means is that they have in a sense given up on the public sector or they're combining the public sector with other private providers.

Similarly, in schooling, we are seeing an explosion in private schooling in lots of parts of the worlds, lot of parts in Pakistan in India, in Bangladesh. It's not only an urban phenomena, but it's moving beyond urban areas.

But there is an issue of the fact that patients and students, just moving to the private sector doesn't solve this issue, and this is something we want to stress, that this is not something that is a natural conclusion of our study, that we are saying there are problems of governance, accountability in the public sector does not mean that now automatically you go to the private sector.

Of course, the government could provide financing and allow people to go to the provider of their choice, but always there would be an issue of quality.

In our study, we didn't compare explicitly differences in absenteeism across private and public providers. In some countries we had that opportunity to do so, say, in India, where we did compare both private schools and public schools, and also in Bangladesh, where we compared private and public schools. There we actually found there was hardly any difference in absenteeism across public and private providers in lots of cases.

So, again, it's not an issue of public or private because there are larger horses that are there in terms of subsidies in the economy and so one thing is that it's accountability and governance, and the public sector could play a role in improving accountant and governance in the public sector as well the private sector.

Just to finish off, I have given this example India of a lot of people turning now to the private sector. Again, a colleague of ours who is involved in the study, Jeffrey Hammer and another colleague in the research department, Jishnu Das, have looked at the quality of private providers in Delhi, and there is a huge variation ranging from the top health-care providers in the world to quacks who operalate on the street and perform all types of surgery under bridges.

So, it's not just going to the private sector. It's again about quality and accountability throughout the system.

This is an excellent question. That's why we like to stress the reason for the absenteeism is part of a more general problem in performance and general problem of governance. And, in fact, we found that if you look across Indian states, it's the states that have the highest absence rates that also have the lowest rates of teaching among the teachers that are at the school. So, we find that in probably the states with the poorest governance in India, we find that 40 percent of teachers absent at any given time, but we also find that a very low share of the teachers who were in school are actually teaching at any given time in those poorest governance states.

So, in some states we have as few as 20 or 25 percent of teachers who are supposed to be on duty actually teaching at the time that our survey teams come around, and so clearly merely getting people in schools or in the workplaces in the clinics is not going to solve all the problems of accountability and performance.

On the other hand, we do have some interesting and suggestive evidence from Professor Duflo at MIT and a co-author of hers who have done an interesting experiment in India in one of the states that has high absence rates. What we have done is work with the local NGOs. They have given cameras to the teachers and required them to take pictures of themselves with the studnets at the beginning and end of each school day, and there is a date and time stamp on each picture. So, this is an attempt at a technological solution to the problem, and what they find is that then they do spot-checks of attendance, and they find that actually absenteeism drops pretty substantially -- because these teachers now understand that their pay is based on the number of days that they are in school. But not only does absence drop, which we would expect, but also the performance of students as measured by test scores rises. So, that suggests to us that at least you're likely to have at least some effect if you get more staff into their workplaces. While it may not deliver a huge improvement in policy, if you don't take other measures to improve performance, there probably will will be some gain.

So, while we think absenteeism is just part of the problem, it IS part of the problem, and if you fix it, it goes a way toward a solution.

To answer this question in a reverse fashion, first this issue of these absenteeism rates stand out for both urban and rural areas.

First of all, our study that we did, the multicountry study, our focus has been predominantly, almost exclusively on rural areas. Within rural areas, we do have variations by poverty, and even for remoteness. There, we find that yes, in general, as you go from more remote areas, their absentee rates increase. In poorer regions, absentee rate increases.

In Peru, the issue is it's predominantly a rural problem. It's a problem of the poor. And in general this is something that we find across the study. Something that we did specifically in the Bangladesh case where we have a better geographical dispersion of this issue in the health sector, well there as you go to sort of rural communities in which there is a town center and then health clinic is based there, there we find about the same rate of absenteeism, about 40 percent that we found in India and other states in India. But, as you move to remote rural areas, there the absenteeism rates jumps to almost 70, 80 percent, so yes, we have indication that this problem is more severe as you go to more remote interior regions of the countries, more rural areas.

Going back to the first part of the question is why does it turn out that better paid staff are also more absent? One think I could say like in the health sector, they're the best paid staff, and these clinics are doctors, which doctors have lots of parallel labor market opportunities. They could work in the public health clinic. They could have their own private practices, legal or illegal, in the local community, in some other community. So, there is a huge demand for their services, and something that we have been now discussing with is how to specifically address this problem, in terms of should there be more explicit recognition that doctors should be allowed to practice, even in these clinics, after hours. Maybe in some coordinated fashion. At least for part of the time they are providing services because as this study finds out that this problem is really most severe in terms of sheer magnitude of numbers in the health sector.

That's a very, very good question, and to do justice to this question, let me first start with very dry answer.

From the World Bank's perspective, who we interact with primarily are state governments and federal governments. They're sovereign entities. There is only so much that we can do in our dealings with them. The rest is ultimately up to the federal or the states to provide these services, who they are responsible for, their citizens.

Now, in a more larger sense of the question, yes, definitely. There is a role for lots of agents to be involved in this. As you mentioned, there is the role of NGOs. There is the role of civil society in community. But one thing we are trying to grapple with, not talking about the Bank in general, but as a larger research agenda and question that the global research community is trying to address this, is what is it that makes politicians responsive to. Are they voting in terms of what is the voting preference of citizens?

Say, for example, in health and education, really things that are really important to people's lives, why don't they vote for local politicians and state and national politicians who would deliver their services? How do politicians respond to voter behavior? What platforms do politicians stand on? Do they stand on a clientless platform in which they're giving services to their ethnic groups and their kin and what not, or are they standing on public goods platform in which there is a large general discussion in a society about these issues and which they're responding to?

And this is something that is still in a very nascent stage of research at least, and there are colleagues in our unit such as Stuti Khemani in the Research Group, Phil Keefer in the research group, and Ken Leonard at New York University. Lots of people in developing countries are also looking at the issue.

But ultimately this is an issue that the responsibility lies with the politicians and policy makers at least as the Worlds Bank is primarily concerned.

Only in situations in which there is severe conflict, there where the state is not functions that sometimes we go directly with NGOs or circumvent the state, but as a mandate we need to go with the state, and it's making the state more responsive either through local citizens demand or from a larger and concentrated effort from the donor community.

This is an interesting question because as Nazmul has mentioned, part of the problem we think is it's hard to get teachers and doctors out of the rural areas where the living conditions are not what they are accustomed to. In particular, doctors come from higher-income strata in society, come from urban areas, have been trained in urban areas, so rural areas don't often have the amenities they are looking for, such as education, good education for their own children. So they may keep their homes in urban areas and travel out to rural areas to do their work, but that will mean often that they don't get out there, that they're often absent.

So, the question is might we be better off recruiting more teachers and doctors locally, and especially improving the quality of education and training out in those communities? And I think the answer is very much yes. Obviously, we face a bit of a catch-22 here because we are saying we need to educate students in rural areas better so they could become teachers and doctors, but at the same time we have a problem with the quality of education that we are trying to solve here.

However, one part of the response may well be to recruit teachers in particular locally in these areas, in the more rural and poorer areas perhaps without the same qualifications as their urban counterparts.

So, the question is, is it worth making a tradeoff if we accept marginally or somewhat less qualified teachers, will we get better performance because at least they're likely to be in the schools in the rural areas where they could make a difference.

Our analysis really doesn't let us look at that tradeoff, but it does let us look at whether you are going to get in a gain in attendance with locally recruited providers. We look across these six countries, and we do find on average teachers who are born in the district where the school is located do attend work at higher rates. It varies from country to country, but in a number of the countries we find this effect, suggesting that the part of the solution may be to recruit these local teachers and be willing to accept some tradeoff in quality if necessary. If they attend work, then you make sure you have somebody in the schools to teach children, because no matter how well-qualified a teacher is, if he or she doesn't show up to work very often, students are not going to learn.

So, this is something worth looking at and experimenting with more, and we think it is part of the solution.

We actually did look at the question of gender. We looked at a lot of variables, a lot of factors that might be an influence in absence, because there are a lot of stories when you go around to these countries and talk to many people, there are stories about why people are absent. One story is women may be more absent because they have more family responsibilities that keep them away in work. In fact, we don't find that. When we look across the six countries, really in none of the countries are women more absent on the teaching side, and only in India, I believe, are women more absent among medical providers.

The more general point is that gender is not a strong determinant of absence, and other individual factors, like age or experience, don't seem to be all that important. What matters more really is the conditions they're working in, the quality of the governance, whether the system is working. It's not really about some people being more absent than others just because of something in their nature, in what we could tell. It's really about the accountability, the extent to which better performance is demanded of them, and the extent to which they're supported in better performance.

Let me just take this opportunity to direct people toward our papers which has been posted online because we really looked to a lot of factors and a lot more than we can discuss here in an hour. We encourage people to follow the links posted with this chat, and go to the project Web site on provider absence where we have a number of papers on specific countries. We have a summary paper that summarizes across the six countries we are looking at, and we have some news articles that summarize our findings.

So, for people who want to know more, I would encourage to you go look at that Web site.

That's a great question, Larry, and I want to let you know there have been lots of people asking similar questions, and that's something we, as an institution, also talk about and discuss in a very regular manner. What is the World Bank going to do about this?

First is that, as an institution, these results that you will find as you go to the link and look at the different studies, the basic figures of absenteeism across these countries wherein countries have been reported in the World Bank's flagship publication which is the World Development Report 2004, Making Services Work For Poor People. So, as an institution, this has come out as a problem that we at least have highlighted.

Now, going beyond that, in terms of just highlighting an issue, what can we do about it? The World Bank is a rather large institution, there are a lot of units that are directly and indirectly involved with this issue. Most direct involvement is from the human development network. There, in the education unit, and I can say for in South Asia, there we have actually engaged a lot of governments in at least, again, this discussion. We have presented these results at the federal level, at different state levels, at different countries, Pakistan, Bangladesh, India, and so forth. There is at least discussion when that with our counterparts, and certain places there is actually initiatives now, as Halsey was mentioning. In states in India, policy makers are experimenting with incentives to make sure teachers show up to work and do their work more properly. It has to be evaluated to see if it brings real changes on the ground.

But also Latin America, where Halsey wants to say a little bit more, but there is interest in from the education sector and other units that deal with larger planning and economic issues. This has been highlighted at very prominent places in the Bank. Beyond that, I can't really comment, but something that I can say is overall where there has been a lack of discussion in the Bank has been in the health sector, where it be within the Bank, whether it be with our other health issues in world health organizations, whether it be our local counterparts in government, where this problem especially among doctors is so blatant. It hits you on the face, and we still have not seen any real initiatives from anywhere to address this problem.

Thank you for that. Yes. I didn't want to let this one hang there unanswered. I think the basic answer is that clearly we do not think that democracy is responsible for poor performance or poor standards. If you look across countries, including the OECD countries but also within the developing countries, it's the wealthier countries that tend to have better performance on these measures, and they tend to be more democratic. So we know over time, over the long term, increasing democracy and increasing better performance tend to go together.

So, we wouldn't want to blame poor performance on democracy. And within our sample, we have a variety of countries, all democratic. If we look in India, for example, and Peru, both of which are vibrant democracies, they have very different performance on this measure of absence -- from 11 percent of teachers absent in Peru to 25 percent in India and 40 percent in some Indian states.

You mentioned upper-level functionaries who own the roost. Certainly they are going to have power in any system. The difference is that when they perform poorly in a democracy, there is at least a chance of citizens holding them accountable and influencing them. As Nazmul has already said, the problem is that in democratic systems, voters don't always seem to focus on this issue, and they focus often on another issue, so it doesn't get fixed as quickly as we like -- but by and large democracy is going to be the way in the long-run to hold government workers accountable.

More generally, let me finish by saying that democracies, as has been pointed out, are laboratories of experimentation, and we are in a period of experimentation here. Lots of interesting work is going on, being carried out by both our World Bank colleagues and academics, and you have a lot of governments facing up to this issue of performance, moved in part by the World Development Report 2004, which brought a lot of attention to this issue as Nazmul mentioned. It's really an exciting time to be working in this area, and I encourage all of our readers and participants here to stay tuned to the World Bank Web site, and the research Web site, because there are a lot of interesting developments going on, and I'm sure they will be appearing here in the months and years to come.