What conditions support well-functioning public-private partnerships (PPPs)? Toulouse School of Economics Professor Stéphane Straub has done substantial research on this topic, which he shares in this episode of the Tell Me How podcast—produced by the World Bank's Infrastructure Global Practice. Stéphane works on issues of infrastructure, procurement, and more generally institutional development in the context of developing countries. He discusses the benefits of PPPs, incentives for investors, key factors in attracting PPPs that serve countries and investors well. He also shares his insights on how government capacity to evaluate projects and to assess, negotiate and implement these partnerships influences their success. You can listen in or read the transcript below.
Roumeen Islam: This is the World Bank’s infrastructure podcast. In this episode, we discuss the conditions under which public-private partnerships or PPPs deliver results.
In public-private partnerships, two very distinct entities join hands over a long time to deliver public services and the results can vary. According to the Harvard Business Review, Australia in 1996 began the Pacific highway upgrade of 408 miles of road and five organizations were involved in this. Procedures and training were established to ensure that the five organizations openly heard and resolved project problems and issues constructively together.
So, when the road builders had unexpected expenses because they had to shore up soft ground, all five partners accepted this reality, even though it lowered profits. The result was that they finished the work seven months ahead of schedule and for $100 million less than the estimate on the concept design. But not all PPPs fare as well, for example, a European union review of nine such projects launched between 2000 and 2014, found seven were late and over budget. Why are there such stark differences? To what extent are the reasons project specific. Can countries do a better job in designing partnerships that deliver on their goals? Let's find out.
Good morning and welcome I'm Roumeen Islam host of Tell Me How and today I have with me, Stéphane Straub, Professor at the Toulouse School of Economics whose done substantial research on the topic of today: public-private partnerships and what makes them work well. Welcome Stéphane.
Stéphane Straub: Hi Roumeen, thanks for having me.
Roumeen Islam: It's very nice to have you here. So, Stéphane shall we begin by clarifying, what do you mean by public-private partnerships in infrastructure? Which definition are we using?
Stéphane Straub: Roumeen, we're going to use the following definition, picture a public-private partnership as a contractor arrangement between a public authority, a national government, or maybe a local government and the private supplier firm. For the delivery and the operation of an infrastructure service. And very importantly, this contract implies that the private part takes the responsibility for maybe building or upgrading the piece of infrastructure, financing or making arrangements for this financing, and then manages this facility. So, the dimension of all of these aspects is key in the public-private partnership definition.
Roumeen Islam: But today, I guess we're going to focus on those that have some financing commitment from the private sector. Not those that don't.
Stéphane Straub: Yes, in general, there is an agreement to consider as a public-private partnership the type of contract, whereas there's some investment by the private parts. So, it's really the investment that mattered. Then it could be financed by the World Bank or government, Barclay, by the private bank. There has to be something beyond the simple management contract for example.
Roumeen Islam: Thank you. So why do governments look for these arrangements with the private sector and how much of infrastructure investments are actually under public-private partnership arrangements?
Stéphane Straub: So, you'd be surprised that, public private partnerships are both rare and quite frequent. So, let me explain. In fact, if you look at the overall investment in infrastructure in developing countries, it's probably a relatively small share, estimations are around 10% of infrastructure investment, maybe 3% of global infrastructure spending when you include developed countries.
But then there are many many PPPs that are active. There are thousands of them. Countries and regions have been very active, like Latin America, for example. So that's the first part, the second thing is why government look for that? Well, the correct answer or the right answer is what they look for in this private participation, because this is supposed to bring some efficiency to the table.
So, over the lifetime of the project, as I was saying at the beginning, from the planning, the building to the operation, you expect that this private partner, because of the characteristic will bring some efficiency to the table and they will do that better than the public sector itself will do. The reason for that, there are several but are really to search into the part where, you know, they bundled is different activities together.
So, if a firm is in charge of planning and building a road or a public infrastructure, and then operating it, then they are likely to invest in innovating at the building stage. So, that operation will be cheaper and more efficient later on, and they can make a bigger profit. And that's really this kind of externality between phases of the project that is key in bringing this efficiency advantage.
Roumeen Islam: That's very interesting. So, you actually just highlighted innovation as a source of gains, but there's also efficiency gains apart from this innovation aspect, there may also be management deficiency gains.
Stéphane Straub: Well, you could say that they also management efficiency gain, but even in management, you know there are innovations.
So, by efficiency gain at that level, we think of the fact that, well, if you know, if this piece has been designed in a more effective way, that the process is more effective than the management will also be easier. And that only comes about if you thought about it from the very early stages of the project.
Roumeen Islam: I see. But I've often heard another justification put forward, and this is about the financing aspect of infrastructure. Could you speak about this, please?
Stéphane Straub: Yeah. So, this is the main point that carries disagreement. When we discuss PPPs is the fact that the government will save on resources by investing that. And I think there's a fundamental misunderstanding between the financing and the funding, because even though it might be financed and operated by the private sector. Ultimately the funding is who is going to pay for that piece of infrastructure is going to be either the users through the fees or the taxpayers.
And so, however, you do it, you know, to probably procurement of PPPs is in the end. It's the same, what changes is at what point you spend the money. So, like thinking that because you do a PPP or you're kind of expanding your fiscal space to use a term that is very common is probably wishful thinking.
Unless you think that you're very constrained today, but in 10- or 20-years’ time, your fiscal situation is going to be much better and then you want to shift the mean, this is most generally not the case.
Roumeen Islam: Right., I suppose.
Stéphane Straub: Yes, absolutely.
Roumeen Islam: So, other fundamental characteristics of sectors that influence the magnitude of the gains or the nature of the gains that affect the trade-off between public and private provision.
Stéphane Straub: Yeah. Sure. So, in the review, we identified a few key aspects. When I say the review, I'm referring to a forthcoming journal of economic literature paper, that is a review of empirical studies on the efficiency of public-private partnerships precisely.
Roumeen Islam: So, you looked at many, many such studies and put together some lessons.
Stéphane Straub: Exactly, we reviewed the whole literature on infrastructure and also education and health for that matter over the last 30 or 40 years and try to see whether there were some general lessons that we could extract from there.
Roumeen Islam: Yeah. Thank you, Stéphane. Sorry, I interrupted. So, please continue.
Stéphane Straub: Yeah, aspects that affect the trade-off. Let me just mention a few, the magnitude of sunk investment. So, when you do a road or when you do a big piece of infrastructure, you have to actually do large-scale investments are large, fixed costs, and these are usually things that are sunk. So, they cannot be redeployed or used for something else in case there is a problem later on. So that carries them the risk of opportunism, maybe renegotiation, expropriation, etcetera. And these risks are high, of course, that goes against the opportunity to do a PPP. So that's the first thing and that varies a lot across sectors. The second thing is I mentioned in the beginning, it's a contract, right? So, if it's a contract you need to be able to contract on something, and these things have to be observable. They have to be verifiable, as we say in economics. So, for example, you want the quality of your services to be contractable, to be observable.
So, when you do you hire a private contractor to do a road, you want a certain quality of service, and you want to put that into the contract, and you want to be able down the line to verify whether the contractor is complying with this contract. It's easy for a road, it might be less easy for a water network because the pipes are underground.
Roumeen Islam: Stéphane, based on what you just said, are there some infrastructure sectors that are more amenable to private participation than others are?
Stéphane Straub: Yes actually. So, if you take what I say and, and apply it to these different sectors, you immediately see that like transport is a sector that fares relatively well, there's competition, there is a generally market test. People pay fees or might be paying fees for the service. Service is contractable and it's easy to verify that the roads are in good shape and that the ancillary services are there. So, apart from the fact that there are large, fixed costs, it works, and you do see PPPs in transport, as a matter of fact.
In energy, you see them mostly in distribution and generation, and maybe we'll come back to that because the energy sector is a complex one, but not in transmission where fixed costs are very large and it's generally a natural monopoly. And finally, you have water. Water is like, it's a bit less favorable than transport because there are large, fixed costs and although you can verify the quality of the water, compliance with the norms, and so on, it's a bit harder to verify the state of the network, for example. Water pipes are underground most of the time there may be leakages, which you don't see in these types of things. So, you do see less PPPs in water as a matter of fact.
And finally, you have the kind of sector that is a bit outside of the scope of this review, which is the ICT sector in general, mobile phones, for example. Mobile phones have all the characteristics that may make them suitable for the private sector. So, what you see there, in general, is complete privatization. So, we're not, not talking about PPPs anymore. And as I say, they're a bit outside of the scope.
Roumeen Islam: I guess what's very interesting is that you see over time that as technology changes and markets change. Sectors where perhaps PPPs would have been harder to do, become easier to do. For example, with better data, you have better monitoring. I think you mentioned this earlier. So, I guess with technological changes, you would agree that certain sectors become far more amenable than they were previously.
Stéphane Straub: Absolutely. I mean, coming back to energy, for example, you start seeing PPPs and actually privatizations in this sector. When there is unbundling of generation, distribution, and transmission and the other set of reforms that came in the 80s and 90s. And of course, the same thing in the ICT sector. We used to think of all these sectors as natural monopolies, and they've been changing quite a lot. As a matter of fact, then you see also the institutional forms evolving together with the technology.
Roumeen Islam: So, moving away from these sectoral characteristics, what are the features of the institutional environment that are important, that affect performance?
Stéphane Straub: Okay, let me just focus on a few of them and I think that will be very intuitive. First, we're talking about a long-term contract. So, two things, if you have better, faster, more efficient contract enforcement, obviously it will be easier to do a PPP. That means rule of law matters and all these characteristics. On the other hand, uncertainty is bad for PPPs political instability, macroeconomic instability, also disaster risk. All these things are going to complicate the stability of these long-term contracts. And then there are some things that are more ambiguous. So, let me group these into like public sector efficiency. So, bureaucratic efficiency or even corruption, because on the one hand, you would think while if there is more bureaucratic efficiency or less corruption, probably you're more efficient at doing PPPs as a public sector, but then think also of the other side of the coin is like, well, if you have a better public sector, more efficient, less corruption, you also better at running public projects.
Then there is a trade-off and it's not obvious that you want to do more PPPs in this case. So, and that's a very important point because it carries to the question of ultimately, what is your objective doing more PPPs or just doing the project more efficiently?
Roumeen Islam: Let me just ask you one question, let me stop you for a second here because you did mention that innovation is a big part of the efficiency gains and I just wonder, even if public sector management is really good. Would they be as innovative as the private sector would be?
Stéphane Straub: They might not be and that might be more important in sectors where the innovation, the part of the innovation potential innovation is more important. So, you would think maybe in ICT or in energy, this is more important, maybe in water, it's less, so then maybe as the public sector becomes more important, you're going to see less PPPs in water. The trade-off would be less stringent in other sectors. I agree. That's a fair point.
Roumeen Islam: All right. So, we talked about these efficiencies, but I was wondering about distributional concerns. Do you think, and have you found, in looking at all the research out there that if you do a PPP, somehow it may be harder to achieve your distributional goals? Reaching the poor, for example.
Stéphane Straub: Yes. I think that's also an important point. Let me put it this way, this is a contract and as economists, we think that whenever we sign a contract, there might be some asymmetry of information between the contracting parties, which means in intuitive terms that while maybe the private operator knows more about its cost or things like that. This means that the contract is going to generate some additional rents to the operator. So, the public sector is contracting with this private firm but has to ensure some additional profits. So, the firm actually wants to go into that contract. And what happens is that if you do that, in fact, mechanically, what is the consequence? The price is going to be higher and given that there is a downward-sloping demand curve for that service, the more expensive it is, the less people are going to consume on the margin. The people that are going to be excluded from consumption are generally the poorest because they have the lowest willingness to pay.
So, just to make it even clearer, there are a distributional concern here, which are that as you do a PPP to build a service that is supposed to include more people, you actually building the mechanism that is such that you exclude the poorest people. And the consequence is that the social acceptability of this project might be low, and we've seen in Latin America, in the 80s and 90s a lot of protests against water PPPs, for example, because prices were increasing despite better efficiency, despite more investment, etcetera, there is this trade-off there. And at some point, there is a limit in that sense. And the additional risk here is that you know, sometimes there's cream scheming as we call it. So, the private sector is only coming in, in sectors that are reasonably profitable. And these are not necessarily the sectors that will serve the poor. Right? So, that might be an issue.
Roumeen Islam: But isn't there a solution to this in that the government can directly provide the amount of subsidy that it wishes to give to these very poor consumers.
Stéphane Straub: I mean sure. But that's a longstanding problem in economics, right? We could always say that we will compensate with transfers. Of course, it's better if these transfers are disconnected from the service itself. So, they don't create distortion at the level of the provision of the service. But on the other hand, we know from a lot of work that has been done at the World Bank and other places that transfers are not that easy to make, they're not easy to target. There are a lot of inclusion and exclusion errors. So, it's also carrying its own problems, but ultimately, that would be the best solution, of course.
Roumeen Islam: So, Stéphane an alternative to subsidizing the consumers directly is for the government to have a contract with the provider, whereby they give the funds to the provider in cases where the consumer is a poor consumer. How would such a contract work? do this work?
Stéphane Straub: I mean, we have evidence that just doesn't work very well. I mean, there's a huge monitoring problem between the government and the firm in terms of how much subsidy is needed etcetera. So, in practice, there's evidence from Colombia that you end up in the low equilibrium trap where this doesn't improve the service. I mean, to make it short.
Roumeen Islam: I see. So again, we have the same problems in terms of doing a contract well and being able to monitor and enforce that contract. This is what you're saying.
Stéphane Straub: Yeah, exactly. It's really about this verifiability of the service and what you're providing and what then the firm gets compensated for, which is tricky in this context.
Roumeen Islam: Let's go back to private sector incentives to invest in a particular sector. How does the remuneration method, whether they get user fees or whether they get taxpayer funds, how does that affect a private investor incentive?
Stéphane Straub: Well, let's again, make it quite simple. If there is a strong market test, then in a sense your PPP might be a way to screen for white elephants.
What I mean by that is that no private operators going to be going into a contract where they know that nobody's going to consume the service and in the end they will be losing money. So, that might be a problem. Now, if there's no market tests, if this is just what we call an unavailability contract, where they get paid by the government from taxpayer money, think of a stadium for example, or some prison or something like that. Then it's really hard to know what's the social value of this investment and so the private investor might go in and say, well, you know, I secure this payment over 20 years and then that's fine. And then maybe the government ended up paying a very high price for a service that has no social value because there's very little demand. So, that, remuneration method, fees versus taxpayer funds will change incentives quite a lot.
Roumeen Islam: So where do you mostly see transport PPPs going? You said that this was a good sector for PPPs. You see them in railroads, buses, ports, airports, where do you see it?
Stéphane Straub: Well, you see the bid everywhere. I mean, the quite a few know PPPs in roads of course. Historically, one of the first PPPs was the turnpikes in England and then in the rest of the world. And they were very successful. nowadays you see a lot of highway concessions in Europe, in France, in Spain, in Latin America, in Chile for example, there has been quite a few experiences with railroads.
Of course, sport and airports as well. I mean, that's quite common even in buses, there are a few projects. So, I mean, these are kind of specific projects, but they tend to concentrate on the main corridors for rows, for example. So, they're quite important.
Roumeen Islam: And then, what about energy? How's the sector different? You spoke a bit about it.
Stéphane Straub: So, energy is complex sector. Where you have a generation. Transmission and distribution. And nowadays in most countries these things have been unbundled, which means that, they are almost separate industries, but of course, they have to work together.
So, you see PPPs, I would say the main lesson here is that it depends a lot. When this process happens, the unbundling process in the 80s, whether this was in mature markets like the US or the UK, or in less mature markets, like in developing countries. So, in mature markets, mostly you saw some privatization in particular, the generation level. In more developing country contexts, you saw some PPPs either in generation or in distribution. I think that's the main dichotomy that you see nowadays.
Roumeen Islam: That you tend to see more privatization in mature markets and less so in less mature markets?
Stéphane Straub: Yes, because in mature market like the US very often you had this wholesale, a market where were generation and distribution firms could sell and buy electricity, etcetera. So, like the underlying institutional context was more conducive to something that is closer to a free market, to make it simple. And I think that's what drove the difference.
Roumeen Islam: So, you mentioned how sectoral reforms have affected the performance of PPPs. Could you explain what you mean? For example, if you take the energy sector, which type of reforms were important?
Stéphane Straub: Yes. So, reforms in the electricity sector was first the unbundling, separating generation, transmission, and distribution segments. Second, changing the regulatory process and that meant mostly moving on to a more inserted type of regulation, like a price cap. We call it the high-power forms of regulation.
Then introducing competition in these different segments whenever possible. And then lastly, introducing private participation. So, this would be the different reforms. So, the regulation part would have a direct impact on the pricing. Right?
Roumeen Islam: Well, I guess introducing competition in the sector would too.
Stéphane Straub: And introducing competition of course would also matter a lot for that matter. So, together, these create very different conditions in these markets. And then from there, there has been an introduction of private agents either in forms of PPP or actual privatization.
Roumeen Islam: So, let's take a step back now and maybe you could just think about some general lessons from your study of all the different sectors you reviewed health education, water, maybe just in short, some general observations.
Stéphane Straub: Here are some general lessons that we highlighted in the review. And I must say in general, the evidence is quite ambiguous in these different sectors.
So, in transport it's unclear, I mean, we have a study from India in the late 19th century where we see that when the railroads became public. So, there were initially PPPs and then they became public again, they became more efficient. So, like not great for PPPs. Turnpikes were very efficient and clearly, giving a boost to the efficiency of the transport sector back in the 18th and 19th century. But then the evidence on modern PPPs is relatively ambiguous again in that sector. True in Latin America and Britain as well for railroads, for example, in energy. Again, it's unclear, I mean there's relatively ambiguous evidence. Finally in water, we see that in developing countries, initially, these PPPs have been quite efficient in the sense that they've extended coverage. In some cases, they've helped reduce child mortality. But then as I mentioned earlier, it's the pricing implication has been more mixed, and the technical efficiency is also less clear
Roumeen Islam: I guess when you say the results are ambiguous. You mean, in some cases they're good and in some cases they're bad and it's hard to, and there are many, and the results depend very much on the things we've already discussed, the institutional environment, the contracting process, etcetera. And this leads me to the next item that I was going to mention, which is renegotiation of PPP contracts, which happens quite a bit. And I was wondering if you could speak about it?
Stéphane Straub: So, renegotiation is probably one of the main issues that people have been flagging with PPPs. I've worked on this issue for more than 20 years now. We generally considered two types of renegotiations. The one that are started by the firms and the one started by the government. So, when it's the firm, it could be that the firm is just facing some unexpected shock. They overestimated demand; they underestimated that cost. They're making losses and then at some point they need to renegotiate the contract. But then also what we see a lot is that there are firms that are very good at behaving opportunistically. And then, so this firm that they are just like trying to win contracts, maybe, underbidding other competitors at the bidding stage because they know that they are going to be renegotiating later on down the line.
So, that’s the first problem, on the government side, we've seen a lot of opportunistic renegotiations as well. Expropriation, nationalization, etcetera. And so, whenever that happens, and it's quite common, in some cases you see rate of a negotiation that are between 50-100%, meaning that, you know, all concessions, all PPPs are renegotiated.
Sometimes several times, we showed that for Latin America, for example, the benefit generally accrued to the firms. Higher tariffs, longer duration, less investment. They accrue to the politicians that are involved because, maybe it's less visible, but that may improve their chance of reelection, for example.
And then there is a third party here that has been invisible, which is the public. And it's unclear that the public is really benefiting from this renegotiation. So that's the dark side of renegotiation. Now there's another side and we have to consider, which is linked to the fact that, you know, when you sign a contract for 30 years, it's unclear what the condition will be in 20 or 25 or 30 years in the general environment. So, there are a lot of unforeseen contingencies, and you have to adjust for that. And it's very hard to write that in the contract upfront. So, at some point renegotiation might be necessary and they might not even be good. And they might just be linked to the fundamentals of this contract, which is the fact that these are very long-term contracts, complex contracts, large pieces of infrastructure, complex services.
And some margin of adjustment down the line is also useful. So that these two aspects in renegotiation, the first one you really want to reduce because it's more the opportunistic type and it's hurting the public. The second type that you may want to understand better because of course you need some flexibility down the line.
Roumeen Islam: We understand that there are cases where you have to renegotiate because of these unforeseen circumstances taking place. And then there are other circumstances where it may just be a result of the contracting environment and things that you could improve upon. So, let's go on to competition, which we touched upon earlier, since we're talking about how well the markets and contracts function with PPPs, how important is competition in the whole process?
Stéphane Straub: So,. The first is clear from the beginning, this is a contract between one public party and one private party. So, x-post, there is no competition. So ex-ante competition for the market is really key in this process. And that's why you generally have a bidding process, and it has to be competitive and it's really important because it helps selecting the most efficient operator. So, this is a theory, of course. An economist has developed this theory of auctions and etcetera for a long time.
Now in practice, we know there are many caveats here because these competitive tenders are not immune to manipulation, to corruption, to collusion, to favoritism. So, there might be a problem in less transparent environment with this process. And there might be cases where you may want to have some degree of intervention or maybe a slightly less competitive procedure, at least before the competitive bidding phase itself.
So, we need to have this x-ante competition, but it's fraught with problems. Second, there is competition x-post. So, as it's clear, you know, in many cases that we won't be any competition, but there is some evidence. And we show that in the review that whenever there is, for example, in energy markets or maybe in transport with model competition, this competition ex-post helps make these PPPs more efficient.
So, as you would expect from economics theory, ECON 101 when there is competition, the surplus is bigger. And that's true also for PPPs to some extent.
Roumeen Islam: And how important is government capacity? Because we spoke about, you know, designing market reforms, we spoke about contracting and, you know, estimating demand. There are so many things that the government and the private sector have to agree on, have to have to do together. How important is government capacity in all of this?
Stéphane Straub: It's absolutely key. And I think if there is a recommendation anticipating on what I'm going to say is that like, very weak government should not be doing PPPs because they won't be able to manage this contract efficiently.
So,and not even talking about political connections, corruption, revolving door issues, and all the government's issue that might actually make this process even less efficient.
So, there is this paradox that some of the governments that you know are more willing to do PPPs because of lack of resources are also the one with the lowest capacity. And then, on the other hand, they should not be doing PPPs because they are less likely to do this well.
Roumeen Islam: So, they really need capacity building.
Stéphane Straub: So, they really need capacity building before anything and that's a long process. I don't think we have evidence of this capacity-building process being a short one. So, it might take decades before they actually in a good position to do efficiently the type of PPPs they want to do.
Now, I still want to say something here that is important, is that the government also they might have different objectives than the private sector. So, the private sector is about being technically efficient and making profit. The public sector might have different objective, social inclusion, distributional concerns, etcetera. And in those cases, it might be important also that they push these objectives, and they actually shape the contract in a way that helps doing that.
Roumeen Islam: Yes. We spoke about this earlier on as well.
Stéphane Straub: Yes, you've seen that in the US for example, and that would sound like a paradox to many listeners, at the turn of the 20th century, the US water sector massively became public, and one of the reasons there were social and distributional concerns, and it was easier for the government to run the service themselves.
Roumeen Islam: Thank you. I'm going to end with this question. Where do you think more study is needed? Now I'm afraid that you're going to answer with lots of areas, but why don't we focus on a couple of big ones?
Stéphane Straub: Okay, big ones. I mean, of course, evaluating the impact is still high on the agenda, you know, and we don't have a lot of very good studies, but let me look forward. And if I can just name two, I think one is the impact of climate change. Climate change means increasing long-term uncertainty, more disaster risk. You can imagine what that does in the context where you want to sign a long-term contract for 30 or 40 years. And we don't understand that very well. So that's very, very important area for further study. You know,
The other one is a, you mentioned it at some point in your question is this big data issue, which kind of changes a lot, you know,, etcetera, and enhance the way we do PPPs as well. And I think that's also an area we have intuitions, but we don't have a lot of precise studies.
Roumeen Islam: So, this might be an area which actually might make it easier for less mature governments to design contracts and monitor them, but this is something that needs to be researched more.
Stéphane Straub: Sure, provided we have the mechanisms in place to do that but yeah, indeed.
Roumeen Islam: Thank you very much. You've given us a lot to think about. Thank you.
Stéphane Straub: Thank you Roumeen.
Well listeners, what did we learn today? Firstly,
Secondly, the way returns to investors are structured, whether through user fees or direct transfers from public coffers, and the way social objectives are fulfilled, are elements that affect the incentives of investors.
Thank you and bye for now.
You can also listen to the full episode clicking on the player below and view all episodes on the Tell Me How: The Infrastructure Podcast Series homepage.