Loading summary
Transcript4 lines
- [00:00]
A
Welcome everyone to this afternoon's public seminar organized by LSE Ideas, the Kuwait Program in LSE Middle East. We've got a mixed bunch of people here and of course Rafael Espinosa, who I'll say a few more words in a minute, has a nice mixture of policy work experience in the Gulf, but he's also a technical economist. And since we're holding this in 32 Lincolnson Fields, which is where the economics department lives, I thought I'd just say a few words about how we're going to go with this seminar. This seminar is officially a public event and so in that sense, for those of you who are here because it's an economics talk, you will notice that it's different from typical technical economic seminar that we hold in the Economics Department. In public events, the audience is much more polite. In technical economic seminars, the speaker hardly ever gets a word out before he or she is interrupted by questions. I would like to keep to the level of civility that says we are. This is a public event organized by LSE Ideas Kuwait Program and LSE Middle East. So what we'd like to do is to invite the speaker, Rafael Espinosa, to give us a presentation on the topic of his book Macroeconomics of the Gulf. He will speak for 45, 50 minutes and then in the usual public events way, we will then have a question and answer session after that. Rafael Espinosa is an old friend here in the UK. He did his PhD at Oxford. He's been working for the last several years at the imf, from which he has engaged in a whole range of interesting academic and policy research. He works on asset pricing, capital flows, capital controls. But he's also had a direct hand in the unfolding of the world's financial Systems after the 2008 global financial crisis. He has played critical roles in managing IMF interaction with different distressed economies here in Europe at the same time that, as you will see from his experience and his talk this afternoon, he has been greatly engaged with the macroeconomic development in the Gulf region. So he's here to speak about his new book, the Macroeconomics of the Gulf. If I could just then invite you to join me in welcoming Rafael with us. His talk.
- [02:46]
B
Thank you very much, Professor Kua, for the kind introduction. It's really a pleasure to be here to be able to present to this audience some of my work that he's been doing together with two colleagues at the imf, Ghaddafayed and Prasad Anantakrishnan. So this is the work that he's been doing at the fund for the last three, four years. However, it's important for me to indicate from the beginning that these are our views as researchers and not the official views of the IMF of its border or IMF policy. So we put together this research work that was initially done because there were interest in the region on specific kind of topics. And then we realized, okay, this is something that is enough material for telling a story about the region in the last 20 years. And so that's basically why we put this together as a book. And I'm going to present kind of the core chapters of that today. So you will know the region, of course. So this is a little map of the Gulf states. I'm going to be talking exclusively today about the country, the six countries that formed the Gulf Cooperation Council, the gcc, which are Saudi Arabia, Qatar, Oman, the uae, Bahrain, Kuwait. I show here the map here and the numbers nearby are the level of income per capita are the rankings of the countries in terms of income per capita. Per capita, PPP adjusted terms. So I think what is striking when you compare to the rest of the region is really that it's a peninsula of wealth in a region of poverty. So you have these six countries here that three of them, Qatar, Kuwait and the UAE are among the 10 highest income per capita countries in the world. The next three countries, Saudi Arabia, Oman, Bahrain, are more populous countries or countries with less oil revenues recently are around the 40, you know, ranked around the rank 40 in the world in terms of income per capita. Nevertheless, you know, this is dramatically, you know, superior performance to the ones to the countries around the region, you know, ranked 130, 110. So of course we know why, right? We know that oil has been playing a major part in this success. And here there is a bit of a puzzle when we look at this literature on oil rich countries, because this literature has been influenced a lot by this chart which I'm showing to you, which is taken from Saxon Warner. So we normally think that, okay, all rich countries benefit massively from the oil revenues. But there is an academic literature discussing the reverse correlation, which is that all rich countries have done less well than all poor countries. So it's a simple scatter plot I'm showing here, again taken from Sachs and Warner's paper in 2001 on what they call the natural resource curves. On the horizontal axis you have the share of exports of natural resources. On the vertical axis you have the growth performance of the countries per capita between 1970 and 1990. So 20 years of data between 1970 and 1990 and the Gulf states that we are discussing today, they look like extreme case of the resource curse. You have Bahrain, Saudi Arabia not doing very well in terms of growth per capita and fairly high resource intensity. And you have the UAE and Kuwait doing pretty badly in terms of growth per capita. Okay, so we'll come back to that in this discussion today. And the plan of the talk will be centered around the first two parts of the book. First is a discussion of long run growth. Okay, what are the determinants of long run growth in the region? Did the region suffer from Dutch disease? So, you know, if you've been following, you know, the literature on resource curse, there are several potential explanations for research curse. One of them is a so called Dutch disease. The fact that countries with high windfalls of revenues, external revenues, tend to have appreciate real exchange rate and that's harmful to investment in manufacturing in the sector that are more likely to produce long term growth. So it is an issue in the region. A third topic also linked to the resource curse is the idea that these natural resources are accurate to the government and government spending can be inefficient and that could be one of the reason why the growth performance has not been sufficient. And the second part of the book deals with macro stabilization issues, which is quite common of the kind of work that we do at the imf. And that's really specifically assessing the effectiveness and the stance of fiscal and monetary policies. There is a third part of the book that I won't talk about today, which is about the financial sector. Given that a lot of data was coming during the crisis and we were working on these during the crisis, it was useful for us to also discuss the banking sector and understand a bit why there were some trouble in Dubai, et cetera. Okay, so let's look a bit at the determinants of long run growth in the region. First of all, this is a region that has had a massive influx of foreign workers and as well in some countries like Saudi Arabia, quite a large increase in local population entering the labor force, finishing school and entering the labor force. And so we get something like a tripling of the number of workers in the region in the last 20 years. So when we think about long term growth, obviously the fact that labor supply is increasing is actually a big driver. We'll discuss the subject of migration a bit later, but for now just take these as okay. There has been a high increase in the number of potential workers in the region, with the biggest increase coming from the UAE and Qatar in percentage terms and of course in number terms. Saudi Arabia had a huge increase in the number of workers as well. So as economists do when they question long run growth, one of the standard first exercises they do is a growth accounting. And so we just start having a look at the composition of the factors of production. How did they evolve during the last 20 years? So we look at Post, Sachs and Warner. So Sachs and Warner was finishing in 1990. We take the new data from 1991 to 2009. It's already data of better quality. So something we are happy to do with slightly better data. And we also take this 20 years horizon and we first look at, I don't know if that's. We first look at the growth of real GDP per worker. So not per capita, but per worker. So we have to have data on number of workers, but we have that and that's the first column, Delta Y. So per year in Bahrain, GDP per worker would have declined by 1.3%. Okay, let's look at Kuwait. Since we are invited by the luca, the Kuwait program. It's a good example. And I'm going to focus a lot on Kuwait today. But some of the stories are across the region. Kuwait or the UAE minus 3%, minus 3.4% in terms of GDP growth per capita. This total real GDP growth per capita. So we didn't exclude the oil sector. The oil sector is there in volume. Now you may ask why do we care about the volume of oil? Right? We care mostly about the value of oil because that's in pure income. And we could do this exercise looking at growth per capita for the non oil sector in real terms. And we do that as well. And the number don't look as bad. Okay? But it's still, you know, more mostly negative numbers in terms of growth per capita. And then we try to explain this growth per worker as a function of the stock of capital per worker and the stock of human capital per worker. And whatever is remaining is TFP growth, right? From the kind of standard Solaris dual exercise. And what we find is that capital intensity actually declined in three countries in the region. That's a bit surprising because we have in mind countries where the stock of capital has been really exploding. Right? I mean, if you think about uae, you think about new towers in Dubai, you think about new investment projects for racing tracks in Abu Dhabi or universities in Abu Dhabi. There are huge increases in the stock of capital in these countries. But we just said earlier huge increases in the stock of workers, in the number of workers coming as well. It's the ratio that really matters. And actually we find a declining share of capital intensity. When we look at educational variables like the number of years of schooling and we apply the standard formulation coming from Cazzilli's work in the handbooks of economic growth, we actually find that educational outcomes would have, should have increased productivity per worker. So that's the third column, the Delta H. Okay. These of course, taking data on numbers of years of education which actually have increased in these 20 years by about one year. So each worker has one year more of education now than you would have had in 1990 on average. So this would have contributed to this growth. But we don't really know the quality of this education. And people who work on this topic are really worried about, you know, what does that mean to just have one year more education. It's really about the content and the residuals. These TFP that are negative are basically, you know, in the standard literature explained as being some kind of efficiency of the use of these factors of production. And if they are negative, we're worried about it. Let's just discuss some serious caveats to this exercise. It's a very simple exercise. The first one for some of you who are interested in looking at macroeconomic data is to realize that most of these countries don't have good national accounts data, especially for the deflators. So the price of investment goods is actually not the time series that we have for all these countries. We took the series for Kuwait, which had quite a good series going back to 1970. Of course, we thought we're constructing a stock of capital today or in 1990. We have to have a series of investment over a long horizon with its price. And this price has been variating quite a lot. So there is an issue with the data of the price of investment goods. That's one caveat. There are more standard issues as well that come up with other countries, which is what is capital? I just mentioned a racing track in tower in Dubai. What are these capital aggregation issues in this, in summing all these things, and we know that they create strange properties and that's a work of Casaly. Again, types of capital. Also, not all capitals are equally susceptible or complementary with labor. So it's creating some issues. I was mentioning the weight given to schooling, which was quite positive. But it's really a matter of the quality of education as well, and not just number of years, so. And against data on non LGDP which gives different outcomes. When you discuss these results. Okay, you know, one question would be why do we care about tfp? Again, these countries are doing very, very well. We saw that at the beginning. The income per capita are very high and as they produce more oil and as price of oil increase, they're doing very well. Why do we care about total factor productivity? There are kind of two standard reasons that are given and the third one I think is not standard, but I find more interesting. The first one is to say that these countries have to be able to live without oil. So in 56 years time when there is no oil in the country anymore, they'll have to be on their own and they have to have a productive, efficient of resources. I find a different sell to government officials, tell them, look, you have to think about what's going to happen in 70 years. Here is just the depletion rates for oil and gas reserves in the country. Saudi Arabia, Kuwait, the UAE, Qatar have 70 years in front of them without having to worry about not having oil revenues. So TFP is kind of an abstract concept in this kind of situations. So why do we need that? Why do we need to worry about that? So this answer, okay, maybe these countries want to diversify to have a non oil sector that is productive. But if they are anywhere rich with all money, why is it so important? I think there is a second reason which is used sometimes is about the efficiency of government money. At the end, a lot of this all money is government money. Do you want the government to run things efficiently? So if TFP is negative, it's probably an indicator that this spending is not being very well used. I think a third reason which is quite important, that not everybody benefits from the oil money. 60, 70% of the population in some of these countries are foreigners. They are not nationals from the Gulf countries. They don't work in sectors closely related to the oil business. They don't benefit from the rent that the government gives via high wages or subsidies and mortgages or subsidies in education. So it's important for the ones who are not benefiting from the oil rent directly to actually, since they are going to be paid their marginal product of labor to actually benefit from a productive economy. And therefore the lack of TFP growth is important. Okay, when we look at this growth literature and try to explain why growth has been disappointing. So we do this TFP exercise and this growth accounting. But it seems to be also quite useful to look at the growth literature on institutions. You know, these, you know, hundreds of papers discussing the impact of democracy, of, you know, volatility of growth, of terms of trade, of trade, openness of the size of the government of beta convergence, the fact that countries that are poor convert to higher income countries. All this literature has had, you know, quite a lot of results and it would seem a shame not to take into account this literature. So what we did here was to do a very simple exercise and look at what we could call a meta analysis on these different type of topics or what is the consensus of the academic literature of the effect of, let's say, terms of trade volatility on long term growth. So we have done this exercise with seven variables that we think are the most important one in terms of growth outcomes. So Salah Martin and others have been looking at which variables are the more important one. And here we took seven of them for the Gulf countries and for other oil producers. And we found that really the potential causes for low growth would come from high growth volatility. And that's very much linked to the resources literature. And in particular van der Plag has been emphasizing the importance of volatility for the resource curves, maybe also quality of institutions. Okay, here we use one proxy which is I think the corruption index from the icrg. That's the red component in the chart and the green chart, the green bar, which is government consumption, the fact that governments are pretty large. So if you look at this growth literature overall, as, you know, big picture, you get that these are the three potential factors in the Gulf states for why growth would have been possibly disappointing. So whether the region suffers from the resource curse or not, I think that's kind of a matter of vocabulary. TFP has been low, growth per capita has been low, but they have pretty high incomes per capita. So it's, I wouldn't call it a resource curse. Nonetheless, it's important to understand why this growth has been slow and in particular TFP since, you know, that's the part that is unexplained. And the rest of the presentation will focus on three main sections, which is the dash dise explanation of the resource curse, the rent seeking or government efficiency issues with big and rich governments, and the volatility of macroeconomic policy. You know, whether fiscal and monetary policies have been able to lean against the wind and actually contribute to a more stable macroeconomic environment. So let's look at the Dutch disease. So the Dutch disease, just as a quick recap, is the idea that when you have revenue windfalls, they, you know, in the case of these countries, it's oil, but you could think about copper or foreign aid or any kind of external revenues. These will tend to increase the Demand for domestic goods and services. And these goods are not tradable. You can't just use the foreign money to buy, to import them. And therefore you're going to increase the price level in the non tradable economy. And that's an appreciation of the exchange rate. This exchange rate appreciates. It's a real exchange rate. So it's not, it doesn't depend on the exchange regime. The real exchange rate appreciation would reduce competitiveness and potentially the production of non oil exports. And in the literature, the idea that this is harmful to growth either because its primary exports suffer from a secular reduction in their prices, or because these are actually the kind of goods and manufacturing goods that you would really want to produce to have endogenous growth like capacity developments and innovation that are. So when we look at the actual real exchange rate of some of these countries, we find yes, sure, they did have a fairly big appreciation of the real exchange rate. This came through actually price increases, inflation, because they mostly have a fixed exchange rate. So they have maintained a certain parity with the US dollar. It's not true for all the countries. Saudi Arabia had a depreciation of the real exchange rate about long, over the long run. Okay. When we look at data on exports and in particular the share of non oil exports and total exports, we find a bit of heterogeneity here. Some countries have been able to diversify, like the uae, exporting a lot of services. Some of the countries have developed so much their energy sector, like Qatar, that there is no way they would, you know, diversify in the sense of the ratio of non electricity export to oil export increasing. But that doesn't mean necessarily much. So what we look at is more of a theory and empirical model about the impact of oil windfalls on the real exchange rate. So it's basically a model of Dutch disease standard model. But these models can be amended in different ways. For instance, Adam and Beavan have been changing the model by looking at the role of public investment. In this chapter we actually focus on labor and the role of foreign workers. The reason is that foreign workers are actually a big part of the population, going from 20% in Oman to more than 70% in the UAE. Kuwait is also quite high, 60%. And these foreign workers, they come from a fairly, they have a fairly elastic labor supply, right? So they come from Sri Lanka, India, you know, Asia, a lot of countries in Asia, as well as historically the neighboring countries of the Gulf. And as a result, these foreign workers are going to come when the oil windfalls come because the government spends the money, wants to have infrastructure projects and bring in foreign workers. You see that Istak creating some kind of movement correlation between the influx of oil revenues and the influx of migrants. These migrants have two effects. On one hand they are going to increase demand for local services for non tradables. On the other hand, they increase the supply side of non tradables. And so the question is, you know, how is that going to play? We can write that down formally as a little model of Dutch disease with migrants where we write down an expenditure function which is a function of the real exchange rate Q and expenditure has to be equal to income plus oil rent. And so we have the income which is a function of the real exchange rate and the labor supply and the oil rent N and the addition to the standard model is to have the level of labor supply link being a function of migrants m who are themselves a function of the oil revenues because they are tracked by the oil revenues or because the government, when they have spent the money, actually increase the number of visas and receive them to work. It's fairly straightforward to differentiate this equation and get a little result on this demand side and the supply side I was talking about of the effect of oil revenues. So here I have the differentiation. So the effect of oil revenues N DN on the real exchange rate dq. Okay, and an increase in Q is an appreciation of the real exchange rate here. And what we find is that there is a positive demand side effect which is a standard effect we were discussing about Dutch disease. That is, you know, the higher the share of non tradables in total expenditure and the higher the elasticity of demand for non tradables to income, the stronger will be the effect of the oil rent on the real exchange rates. That's kind of the first term eta times lambda. On top of that, migrants come and they consume domestically and the higher they wage GM and the higher the elasticity of the number of migrants to the oil inflow, the more there will be an increase in demand side effect. But on the supply side, it goes the other way around. Workers actually contribute to the productivity of the economy and the higher the inflow, the higher the effect of migrant workers on local services and non tradable supply, the epsilon variable. And the higher the productivity GQ over M, the lower the effect on the real exchange rate. Okay, so that's kind of the parameters you would get from this exercise. And at the end the question is more or less an empirical one, what's going to happen, right on the effect of migrant workers and all revenues on the real exchange rate. So a panel model of the real exchange rate for the region allows us to estimate this. Now the advantage, why is it useful to do panel? Well, something I haven't discussed much is the availability of macroeconomic data in the region. We only really have to 20 years of data of decent quality in the region. Before 1990, there wasn't much. To be honest. If you look at countries like the uae, even very recently, they didn't have a well functioning statistical agency for the entire Emirates. It's a federation for every single Emirates had their own statistical agencies. They don't necessarily coordinate, etc. So getting good data on inflation is a problem. Even so, at one point we were in the country in the UAE and deflators between Dubai and Abu Dhabi were not correlated. So if you look at the inflation of the subsectors, there was no, it looked like Dubai and Abu Dhabi had no relationship to each other. So it's, you know, there were an issue with the construction of deflators. Even it's getting better because the governments are realizing the importance of this data. But if we have only 20 years of good data, it's going to be very difficult. Difficult to go and exercise country by country. And in some cases like in these exercises, it makes sense to put them as a panel. Okay, so the idea that the panel will have enough homogeneity across the six countries to be able to tell a story and that will help with the degrees of freedom. And when we do this exercise of assessing the effect of oil revenues and migrant workers on the real exchange rate, we actually find that oil revenues did not have a significant effect on the real exchange rate. However, the wage of the workers, migrant workers, which is here proxied by their remittances, we don't really have the data, has actually a depreciating effect on the real exchange rate. So it's in line with theory that can be both a direct effect or indirect effect. The direct effect is simply that remittances are really money that goes outside of the country and therefore alleviates the oil inflow. That would be the direct effect and that can be the indirect effect that more generally remittances are a proxy for the wage bill of foreign workers in the economy. We don't have this data which would be the ideal data. So that's basically the idea of this panel. Okay, so overall we find a relatively positive picture for the country that the real exchange rate is not going to be be one of the main issue in terms of the Long run growth. And that's specifically because of their labor policies. So it would change. If there are big change in the setup of foreign workers, which is something that is always discussed in the region for political reasons, we would expect this behavior to change and then we would be more worried about their Dutch disease story. So how efficient is government spending in the region? The second theme in the resource curse is about government efficiency and the fact that all money is well spent. When we look at data and we look at what these Gulf countries spend, we get a bit worried because they look quite different from what standard OECD governments do. So I give the example of Kuwait here. 21% of their budget is spent on the economic affairs outlet. That means support to businesses, et cetera. Okay, on top of that you have $16 billion spent on subsidies to energy, electricity, food, water. That amounts to about 32% of government spending. And that's outside of the balance sheet for in most cases outside of the statements of the government because these are implicit costs subsidies. So you just sell the oil below the price of the international price. So you know, we can almost sum these numbers and find like 50% of government revenues get, you know, spending in support either via business support or subsidies. Quite big numbers. And on top of that, it's not clear that subsidies are, you know, there because of an obvious economic reason like you know, we know that we have to subsidize education because you know, we, you know, it's profitable, you know, from a long term perspective. And people tend to under invest in education and stuff like that. Now here they spend, you know, the reason why these countries spend a lot on subsidies is because they are very all rich and so they can afford it. But that's kind of on the right hand side of the budget constraint. You say you have a lot of money, therefore do anything with your left hand side. That doesn't make much sense. You know, what matters is, you know, your opportunity cost. What can you do with this money? So there is a very strong relationship in the oil producers between the size of the oil sector and the amount of subsidies spent by the government. On top of that, there is also a very strong correlation between the share of investment in the economy and the revenues of the government. So a lot of investment, 40, 50% is actually done by the government. We don't have necessarily very good data on the degradation between public and private investment for the whole world. So what we look is total investment and we relate that to the size of the oil sector and we find a very striking correlation for oil producers between the size of the oil revenues and the share of investment of gdp, which again suggests that these countries are investing because they can afford it, not necessarily because of the returns. And that's worrying because the literature has been quite pessimistic about the role of public investment for long term growth. If we were thinking in general terms about, you know, okay, a country that is very rich and wants to distribute money to its citizens or to residents and you know, investment can be also thought as a public investment can also be thought in general terms as a way to distribute money. Okay, the theory, the static theory we have about that is the optimal taxation theory from public finance. Because we just have to inverse the standard arguments about optimal taxation. We can reverse them and say we have optimal subsidies, we have a government that is very rich, wants to distribute money. What is the optimal way to do it? One way would be to just have checks to everybody with a fixed amount and that would be a lump sum subsidy, a lump sum transfer, and that would be optimal. But if we think that the government doesn't want to do it because it's maybe too politically transparent, maybe there are some reasons they actually want to have industrial policy, they want to have some effect on some markets, then ok, there is. Let's start with the baseline model which is the optimal taxation, and just flip the sign and talk about optimal subsidies. And you know, you will have seen these graphs about, you know, in public finance theory, you would normally want in subsidies the same way you would want in taxation. You would want to subsidize the commodities that are least elastic. Okay? Because the welfare loss, which is triangle in gray in dark gray are some smaller for inelastic commodities than they are for elastic commodities. So the welfare loss is exactly similar to the welfare loss you would have with the taxation. That gives some ideas about giving some policy advice because a lot of these countries actually are thinking about their subsidies strategy. So first outcome of this theory, goods with a low demand price elasticity should be subsidized at higher rates. Okay? That means food, health, services, certainly not energy. Okay? Energy has typically a very high demand elasticity to price around minus one for the region. For what I've seen in the empirical literature, and of course I just said that subsidies are very big part of the energy is a very big part of the subsidies. So that's kind of a first result. A second result would be, well, if you're thinking about desubidizing your economy, which some of these countries are thinking thinking about, for instance, Qatar's industries which takes feedstock from Qatar Petroleum. Actually, the price at which the oil is sold from Qatar Petroleum to Qatar Industries has been increasing in the last years because the government wants to force Qatar industries to actually behave a bit like a competitive company and pay the real cost for its energy. So these kind of exercises are fine, but the things theory tells you care about relative prices, not about individual subsidization rates. And so really you should think about moving subsidies in parallel for a lot of different components. So if you are starting to desubsidize oil, you also should desubsidize gas, maybe you should desubsidize electricity, and maybe you should desubsidize other utilities, for instance. So we get a couple of normative results. That was a static story, but there's also a dynamic story which I want to talk about, which is a lot of these inefficiencies are really dynamic. Really, when the government has, you know, subsidies are so large it can create long queues, people waiting for actually getting these expected government benefits. So I give two examples here. One is a real estate development fund in Saudi Arabia, Arabia that is actually extending interest free loans to Saudi citizens for mortgages. So that's, you know, we're talking about huge subsidies. The balance sheet of these real estate development fund is, you know, almost 100% of GDP or something like that. And the government of course cannot, you know, match the demand. So there is a quantity constraint here and that's creating long queues. So you get in the press reports about people waiting 10 years for getting a mortgage at zero interest rate from the government as opposed to going to the private sector. And that's a huge crowding out of the private sector activities. And you can create a similar queue system with education, for instance. And a famous example is Egypt with college graduates who were entitled to government job. And there was these long queues about people waiting. And at the end Egypt had to remove this kind of scheme. And so, you know, these dynamic efficiencies are quite important. Probably I find they are more significant than the static ones I was just talking about. We actually can, you know, do a little exercise, I think that is relevant for the Gulf because for instance, as you know, with the Arab Spring, some of these countries have been increasing public service wages, okay? And I want to show how this thing is worsening the outcome of the labor market. So if you try and write a little model of queuing for public services where workers can either take a private service job or wait for, you know, finally getting a public service jobs. And we are assuming that they wouldn't move back from a public service job to a private service job later on and they wouldn't be able to move, you know, if they accept a private service job, they kind of lose their position in the queue and they wouldn't get public service jobs. If you kind of make this simple assumptions in dynamic model, you can actually show the conditions under which an increase in public employment crowd out private employment by more than one, which means that the overall effect on employment is negative. So as the government hires public workers, what's happening is that it's creating a higher probability of getting a public service job and higher valuation of queuing and therefore people are even less willing to take a private sector job. So this kind of situation would happen if the public service wage is about 50% higher than the private sector wage, which actually happens in the region quite a lot. So, you know, public employment policies will have a big effect on the private sector participation rate. And again, I find these kind of things, you know, dynamic inefficiencies quite costly. Okay, so now I'm just going to move to the second part of the presentation, the second part of the book, which is specifically about stabilization policies. GDP volatility is very high in oil exporting countries. That's the second to last bar here. And typically much higher than OECD countries and even higher than other developing countries. This is especially true in the Gulf region, probably because they are just more extreme versions of the typical oil exporter. And Kuwait GDP volatility is very high, tremendous, for instance, even after the Kuwait war to go for in the literature on growth and volatility or in the literature on the resource curse is often being highlighted as one of the main factors of the growth of the pro growth performance of exporters. So it's a key element to understand how this volatility is coming about and what fiscal and monetary policy can do. There is some literature on procyclical fiscal policy in emerging markets. That's for instance, Ethan Isleski with LSE has done a paper on that. And you know, we are going to go on this line. The second branch of policy is monetary policy. And here we'll have to discuss about the role of the fixed exchange rate regime in the region. Let's start with fiscal policy first. These are countries that have an exogenous interest rate and normally if they're a closed economy, the Canadian multiplier would be quite high. So the role of the impact of fiscal policy on short term growth will be quite strong at the Same time, they are very open economies with large imports remittances, which could reduce the size of the multiplier. So it's not very clear which way it goes compared to a standard multiplier, which for countries with adjustable interest rate is around 0.5. For countries with a non adjustable interest rate might be closer to one. So there are issues when estimating fiscal multipliers, mostly endogeneity. The fact that the government has automatic response of spending to growth, either because of automatic stabilizers, so really automatic effect, or because fiscal policies are trying to be countercyclical. And the literature has kind of two main ways to deal with endogeneity issue. The first solution is to try to identify exogenous increases in spending and government spending, something that is not related to, to the business cycle. Okay, so there is Rummer and Romer in the US doing a lot of that here. We know we don't have this kind of detailed information on how the budgeting is being done. And it's kind of complicated, good science. But we have a specific case of a good instrument. In Saudi Arabia, the lunar history calendar which is used to pay public servants is actually not the Gregorian calendar on which the private sector is being paid and actually the public servants seven months every two or three years. So this can be used as an instrument. It's really an exogenous increase in public spending that they can see in the data. And there are two drawbacks of using this instrument. The first one is that the degrees of freedom are quite limited because it's 20 years of data, 25 years of data. And we know IV is biased in this kind of setting. And also the increases are fully anticipated, which actually should, it's going to mean underestimate the fiscal multiplier. But it's worth doing because the effect is quite striking in shaded areas. You see here the years in which the Gregorian calendar adjusted to the history lunar calendar. And that's a gray shaded areas. And the wage bill of the government is a gray line which really spikes every single time. So there is no doubt about that. The wage, wage bill is increasing quite massively in these years by about 15% every two or three years. And you can see how GDP growth, the black line actually does have these little bumps that kind of follow these things. And when you do that, you find the IV instrument tells you the estimated player for Saudi Arabia is around 0.4. So it kind of give us a little idea of what the baseline of the broad area we expecting this first count player to be on the second solution, which is again standard in the literature, is to use actually a ver and to try under some assumptions, such as fiscal policy is not contemporaneously reactive to GDP because maybe policymakers don't have high frequency data, then the standard VR procedures can be used to estimate fiscal multipliers. On top of that, we're going to be able to talk about procyclicality and contracyclicality of the fiscal policy and we're going to be able to decompose the growth cycle of these countries. We find multipliers of the range, fiscal multipliers of the range of 0.4 to 0.7 for different countries for different years. So it's a significant impact of fiscal policy in these countries. Despite the trade openness, fiscal policy is mostly pro cyclical in the region. Okay, so you see, Kuwait, you know, these are the impulse response, the effect of a shock to GDP on fiscal policy. And these are quite procyclical responses. And that's coming really from the fact that the governments don't have medium term frameworks for their budget planning or are not really using them, you know, and also from the effect of oil revenues. One exception that the oil revenues that accrue to the government and at the same time stock stimulate the economy. Saudi Arabia is a bit of an exception. And here we have faith on these results because actually they have been always mentioning the fact that they were doing fantastic policy and a big example where in the 2000 years when the price level was very, very low, the central bank really decumulated reserves for the government to spend. So actually it makes sense. But fiscal policy is pro cyclical, which is again quite typical of emerging markets and especially of oil producers. You know, this VR allows us to decompose the growth cycle. In Kuwait, of course, we have, you know, the first Gulf War, which is this huge collapse in GDP and bump right after. But we use dummies to control for that and we get an idea of, you know, the contribution of the different type of shocks to GDP growth. And we find that government expenditure shocks are actually a big component. Again, remember, this is a pro cyclical fiscal policy with a fairly big multiplier and has been contributing to big growth outcomes. In 2004, the period of overheating for the Gulf region also has contributed to the modest growth in the 1998, 2000 years when the price level was very low. And again, fiscal spending was actually contractionary in 2010, 2011, when actually you would have wanted to be more stimulative. Okay, so we've talked about fiscal policy and I'm just going to conclude on monetary policy, these are countries with a fixed exchange rate. Okay, Coed is a bit of an exception. It has a basket and the basket is not, I think is not public, but it's a component of the US dollar and the Euro. You can guess it by looking at the data. So mostly these are countries with a fixed exchange rate and they're interesting to study on their own, but also because they are kind of specific cases compared to advanced economies, they're actually quite common for emerging markets and in less developed countries. So they have a fixed exchange rate regime. The central banks do not target an inflation rate, obviously. And as a result, central banks try to operate in quantities, stabilization of foreign capital flows, but also attempts at affecting interest rates and the supply of credit even within the fixed exchange rate framework, which is kind of surprising. But actually they are having some policy space there. When we want to assess the fixed exchange rate regime, there is two ways to think about it. One way is to say, well, the price level overall has been quite stable and that therefore it has been quite successful as a monetary policy regime. I find a bit strange an argument because yes, it's a monetary anchor, but you know, saying that there is no hyperinflation, you know, is not such a huge success. A lot of countries don't have hyperinflation. Now that we know a bit more about, you know, how to control monetary aggregates, we shouldn't be surprised about that. What I worry more is like this big volatility in the inflation rates. Okay. And if you think about the standard macro models, this is actually more costly than anticipated price increase. Right? So we worry more about unanticipated change in the price level than about anticipated change in the price level. And the voltage of inflation is actually potentially more costly than a trend of 6,7% inflation constantly across many years. So I wouldn't say that that's my personal view, again, not the view of neither my co authors or my colleagues at the fund, that I wouldn't say that the exchange regime has been very successful in terms of stabilizing inflation. You can see here the fairly high volatilities. Okay, now the question is, okay, the alternative which is moving to a flexible exchange regime, is that a good alternative for these countries? And one classical argument is to say, well, you know, the banking system is not competitive enough, not developed enough, that the transmission mechanism of monetary policy is actually relatively weak. So I show here in this, in the, in the, I don't know what kind of color is it. In this brown line you have the three month interbank rate, which is the one at the central bank is most likely going to be able to affect via quantity operations or even by interest rate setting if they were doing that. And then you have the lending when the deposit rates shown in different colors. And yes, there is a broad movement and you can find some co integration and stuff like that, but overall you don't have a strong transmission mechanism. And that's true for koe, that's true for the rest of the region. So therefore having your own inflation targeting regime with setting up interest rate would not necessarily be a very efficient thing. On top of that, central bankers are actually quite happy that they do have some leeway in affecting interest rates even in the fixed exchange regime. And that's obviously because assets are different. As an international macroeconomist you think interest rates are exactly equal if you have exactly if you have a fixation age regime. But in fact it's about the type of assets that you have and if they are imperfect substitutes, there is no reason for this trade equality to hold. A good example was the global financial crisis in which suddenly US assets and emerging market assets, even with the fixed exchange regime were not considered to be identical. And you can see these bigger gaps in the spreads between the US dollar Libor rate and the interbank rate in Kuwait. Okay, so it's useful, you know, when you want to discuss the effectiveness of monetary policy to attempt to estimate a monetary ver. What is done in most advanced countries is okay, let's put this data together, let's identify monetary shock and let's get the impulse response, okay? And it's just a very, very difficult exercise in the region. So we did it and we get some reasonable results. But it's very difficult. The first reason that monetary VRs, the identification strategies are normally predicated on high frequency data. The assumption that ideally you want monthly data, maybe quarterly data, because you really want to say that the central bank is just waking up and doing something, a shock to the interest rate, not anticipating or not taking into account current events to counter react. Because if you do that you will underestimate the effect of monetary policy. And we don't have these high frequency data in the region at the same time, policymakers don't have it either. So the econometrician here is not less data than the policymaker. So it's possible that the monetary policy is not very reactive and therefore that the VR identification strategy actually works. Of course you have A fixed exchange regime, which means that you actually also want to control for the US monetary policy. Okay, so what we do is to extend the kind of two country VR model done by Minion and Rogers and the gmcb. And because it's two country panel VR, we're going to be able to disentangle the monetary shock from the US from the monetary shocks from the Gulf region. And we pull this as a panel again because of the issue of degrees of freedom. So we have 168 observations and what we find is that tightening of US monetary policy, the Fed fund rates, is actually a significant impact on inflation in the Gulf. That goes directly, that's the impulse response, but also indirectly via commodity prices, global growth, et cetera. When we look specifically at the Gulf, we actually do, we don't find an effect of monetary policy in the Gulf on output growth. So it's not significant here, this impulse response. So monetary policy is not very effective in terms of affecting growth. We do find an impact of monetary aggregates on prices. But that's really the quantity theory of money. I mean, you know, we just said earlier that the short term inflation volatility was quite harmful. And actually that's exactly the horizon over which there is very little effect on monetary policy. So when you look at a couple of quarters, you actually don't get, or even up to eight quarters. So two years you don't get an effect of monetary policy on inflation in the Gulf, you just get it over the long run, which is not surprising, but also not very policy relevant. Okay, so I just want to conclude on the presentation to say that, you know, the big findings are that the growth performance in real terms and after filtering for the capital accumulation and the, you know, education, increase of education and the increase in the labor supply is actually a bit disappointing. You know, so this TFP has been declining. The Dutch disease might not be the main issue because actually the real exchange rates have not appreciated that much or at least not across all countries in the same way. And one possible explanation is that the labor market is quite different from what you would have in a standard Dutch disease model. So other potential explanations for this low performance would be that the public sector is very, very large and doesn't have good incentives to spend money efficiently. On top of that it creates rent seeking and all the kind of behavior that will affect the private sector. And I was giving the example of the mortgage market for instance in Saudi Arabia. And other explanation for poor long term growth might be that growth and fiscal spending volatility is very very high. And that has been one of the factors typically expl using the literature to explain poor performance of resource rich countries. So fiscal policy has not been very helpful. It has tend to be procyclical. And monetary policy is not really affecting growth. It can affect, you know, I mean, it's being good at anchoring long term inflation, not so much about affecting short term growth or about stabilizing inflation in the short term.
- [54:45]
A
Thank you very much.
- [54:47]
B
Thank you very much for the invitation.