Transcript of a Conference Call Press Briefing by Kenneth Rogoff, Economic Counsellor and Director of Research, David Robinson, Deputy Director of Research and Jonathan Ostry, Assistant Director of Research, IMF

April 3, 2003

Transcript of a Conference Call Press Briefing on the Analytic Chapters of the April 2003 World Economic Outlook
by Kenneth Rogoff, Economic Counsellor and Director of Research,
David Robinson, Deputy Director of Research and
Jonathan Ostry, Assistant Director of Research
Washington DC, Thursday, April 3, 2003

MR. MURRAY: Thank you for joining us today. As the operator said, I'm Bill Murray of the IMF's Media Relations Division and this is an embargoed briefing by Kenneth Rogoff, the Fund's Economic Counsellor and Director of Research. Ken will be briefing today on chapters two, three, and four of the latest World Economic Outlook.

In addition to Ken, joining me is David Robinson, Deputy Director of Research at the IMF, and Jonathan Ostry, Assistant Director of Research and head of the World Economic Studies Division which prepares the WEO. The authors of the various WEO essays are also with me, and as the press summary states, their names and titles are at the top of the press summaries which we circulated in advance of this briefing.

Before I turn the table over to Ken for some brief remarks, let me remind you that this briefing and the contents of the WEO chapters is under embargo until 2:00 p.m. Washington time, or 1900 GMT today.

A transcript of this briefing will be posted on our external Web site after the embargo is lifted.

Chapter one of the WEO will be released on April 9th and we will have embargoed text available for the press 24 hours before Ken's briefing on the global forecasts.

I ask you today to limit your questions to the WEO essays in chapters two through four. Lastly, I expect that when we launched the WEO in the autumn, the Fund's External Relations Department will have a media Extranet site in place to make access much easier for journalists. The site will be a permanent password-protected facility for the press, so you only have to register with us once and will be able to receive a host of embargoed text from the IMF via that route in the future.

Now let me turn the table over to Ken Rogoff. Ken.

MR. ROGOFF: Thanks to all of you who are participating, today. I am going to give a brief overview of the chapters before turning over to David Robinson and Jonathan Ostry, who are here together with the rest of the World Economic Outlook team.

Before providing you with a more concrete feel for each of the three analytical chapters, let me first state what I think the highlights are.

Chapter 2 looks at the effects of bubbles on economic activity. In particular, we find that weak corporate balance sheets have been a drag on corporate investment in Europe and, to a lesser extent, the United States. In Europe, the stifling effect on investment may have been by as much as 2.5 to 3 percent of GDP in 2002, and remains significant. In the United States, the comparable amount for 2002 is about 1 percent of GDP. More worrisome, is the chapter's analysis of equity and housing market booms and busts, particularly housing. Using our statistical measure of peak to peak, UK housing prices are up by almost 70 percent in real terms-that is, a 70 percent price increase beyond inflation-with real housing prices rising in the Netherlands (78%) and Ireland (102%) even higher. In the United States, prices are up 27% in excess of inflation since the last peak. While this appreciation is not as spectacular as in the several other OECD countries, it is notable as it is still greater than during any of the booms we clock for the United States since 1970. In fact, there are many countries, including all of those listed above, where housing price rises already exceed a boom, by our measure. According to the study, on average across the OECD, this implies a forty percent chance of a later bust. Also notably, the study finds that the bust will typically be larger, the larger the increase in prices during the boom.

Chapter 3 puts some numbers on the benefits of what we have often loosely described in past World Economic Outlooks as "institutional reform." While enhanced aid and the opening up of industrial countries' markets to developing countries, especially agriculture, are extremely important, in the long-term, improving institutions that underpin developing countries' economies is ultimately the key to raising world living standards. This proposition has been oft-stated, but here we try to at least loosely quantify the effects.

Chapter 4 also looks at institutional reform, but from the much narrower perspective of labor markets. In this chapter, we find that the equivalent of 4-weeks pay per year for every European worker is essentially being sucked into a black hole to no one's benefit as a result of restrictive labor market institutions. While we readily concede that the chapter only examines the economic costs of such practices as generous unemployment compensation, centralized wage bargaining, strict employment protection, and so on, it shows that the costs are far from trivial. This chapter, together with a large related body of research by academics and other policy institutions, puts the onus on opponents of reform to show that the offsetting benefits of the status quo restrictions, or the costs of achieving a transition, are anywhere near as large.

Let me now briefly say a bit more on the specifics before turning it over to my colleagues and opening it up to questions.

I have already mentioned that Chapter 2 suggests that housing prices in many regions exceed our statistical measure of a boom. Among the countries with housing booms include the United Kingdom (70% housing real price appreciation since last peak), and the United States (27%), but also the Netherlands (78%), Australia (43%), and Ireland (over 100%). Roughly 40 percent of all housing booms are followed by busts. The study also finds that housing price busts tend to be associated with very high GNP declines-even more so than equity price busts. Housing busts tend to be associated with cumulative GDP losses, over several years, adding up to as much as 8%, on average, for OECD countries. Moreover, the study finds that housing busts tend to be highly correlated across OECD countries, most probably because business cycles are highly correlated (if I had to pick a number, the correlation across OECD business cycles seems to be about 50% over the post-1970 period). Most of the previous busts occurring in the years 1980-82, and 1989-92.

Moving to Chapter 3, as I have already mentioned, this chapter puts some numbers on how big the benefits, in terms of per capita incomes and growth, institutional reforms might be for developing countries. We acknowledge that attempting to quantify institutions is as much art as science, but together with many academic researchers in this area, we feel the quantitative results are still potentially quite important. To do this, the chapter explores a variety of quantitative measures of institutional quality reflecting, for example, perceptions of the degree of corruption, political rights, public sector efficiency, regulatory burdens and the rule of law. After taking a stab at quantifying these difficult-to-quantify elements, the results are quite dramatic. For instance, if the institutional quality in sub-Saharan Africa were to increase to levels in developing Asia (one level up in the hierarchy of institutional quality) per capita income would rise by 80 percent from about $800 to $1,400 per annum. If, in turn, developing Asia's institutional quality rose to the average of the entire sample, per capita incomes there would roughly double to $2,500 per year. Beyond these striking one-time effects on a country's income levels, institutional reform can have a similar impact on income growth rates: Sub-Saharan Africa's per capita income growth could be almost 2 percent higher per year if its institutions rose to the world average. Thus, while institutional reforms do not happen over-night and they are admittedly difficult to implement, the potential payoffs are quite large.

Finally, Chapter 4 shows that institutional reform in labor and product market could have a profound effect on raising consumption and investment and on lowering unemployment in Europe. The focus is on continental Europe, where generous unemployment compensation, centralized wage bargaining processes, strict employment protection (e.g., large firing costs) and high taxation of labor income all contribute to unemployment rates far exceeding those in the United States. (It is notable, that countries that have significantly reformed their labor markets, such as Britain and the Netherlands, already have much lower unemployment rates than most of continental Europe.) The chapter finds that if European labor market competitiveness were to converge to the United States level, European unemployment rates would fall dramatically-overall, by over 3 percent in the euro area. Consumption and investment would rise by more than 5 percent over the medium term. If European product markets were similarly brought to United States levels of competitiveness, the gain in output in Europe would double to roughly 10 percent. Needless to say, these are not small amounts. And while they represent only the measurable economic costs to existing labor and product market practices, the numbers nevertheless challenge those who cite various societal benefits of these practices to show that they are at least as large.

Interestingly, we arrived at similar numbers using two very distinct approaches. The first approach was based on comprehensive data on international labor market institutions covering 20 OECD countries over four decades. Importantly, the data base includes many episodes that opponents of labor market reform often cite as evidence that reform is unnecessary, including the 1960s and 1970s, as well as data from the Scandinavian countries, whose labor markets are thought to be somewhat more flexible than continental Europe. The second approach was based on the IMF's new "Global Economy Model" that is being developed to back up and ultimately replace the older framework now in use. The global economy model uses a more institution free approach to measuring the frictions, focusing on markups of prices over costs (40% in Europe versus 15% in the US on average) and comparable markup numbers for wages.

Thanks, and with this I will turn the floor to David Robinson.

MR. ROBINSON: Thank you very much, Ken. I think let's move straight into questions, please.

QUESTION: Hi, Ken. I have a question—there's a couple of statements that seem to be contradictory. Maybe they're not; maybe you could explain them. Both in chapter two.

At the end of page 15, and then going into page 16, and I'll just read you the bit, it says, "While a detailed breakdown is not available for countries yet, investment growth may now stabilize in the United States, as it typical, though it seems to be still falling in the Euro area and Japan." It says investment growth stabilizing in the U.S.

And then on page 28, you say, "The process of deleveraging may have to advance somewhat further before robust recovery is in the offing."

So I'm sort of trying to get a feel as to whether that's the case more so for Europe, or whether, as it says on page 15, dragging over to 16, that investment growth should stabilize in the U.S. So that was the first question.

The second question is your box on page 25 about the probability of recession, predicted with the corporate vulnerability index, is this your index? And you note that the probability of a recession has now fallen to 15 percent. If you could just expand on that a little bit. And a third question, if you wouldn't mind.

You mentioned some data about the drag to GDP by weak balance sheets in Europe and the U.S., and I'm not sure where you got those numbers from within these chapters; if you'd just point me to the right page.

MR. ROGOFF: Mark, let me just first say I gather this is your last time covering the World Economic Outlook and we're going to miss you

QUESTION: Thank you.

MR. ROGOFF: I'm going to turn over the fielding, farming out the answering of these questions to Dave Robinson.

QUESTION: Okay; thanks.

MR. ROBINSON: Hi, Mark. Let me just address the first two a little bit and then I'll ask Jonathan Ostry to talk about the third, the numbers on the balance sheets

I don't think the two statements are really contradictory. I mean, I think the first statement that you cite simply, you know, describes the facts as they are now, and obviously we can discuss that more next Wednesday, if you're still here, when we have the discussion of the current conjuncture.

But the second statement I think refers to the risks, looking forward to a recovery in investments. What we're saying is that leverage is an issue, and in fact—and this was somewhat surprising to us, once we had completed the analysis—more so in Europe, because leverage has increased more there, it's higher, it doesn't appear to have corrected much as yet. And it appears to have a bigger impact on investment anyway, in part because European financial systems are bank-based. So much more investment is financed through borrowing than, say, through the stock market. So that's what I'm say on that one.

The second one is this corporate vulnerability index, our index. It's an index produced within the Fund. As you say, the latest number as of 2-1 2003 is a predicted 15 percent probability of recession, which is certainly not zero but it's clearly much lower than we saw earlier in—was it just around 2000, I think, on the chart, when it was predicting a 50 percent probability of recession and of course it was right.

Let me turn over to Jonathan for the last one.

MR. OSTRY: On the drag to GDP from-

MR. ROGOFF: I just say to say, you probably should introduce yourself each time you speak.

MR. OSTRY: Jonathan Ostry. On the drag to GDP from weak balance sheets, basically what we've done is look at the run-up in leverage in both the U.S. and in the Euro area, and look at the impact on investments, and what we find is that through the end of 2001, which is when the formal part of the study, the data for that study ends, the impact on investment to GDP ratio would be on the order of 1.3 percent of GDP for the Euro area, and a much smaller number for the United States.

We can extrapolate that through 2002, given a further drop in equity prices and the impact on the debt to market value of equity ratio, and there you would get numbers similar to what Ken mentioned in his opening remarks, namely, on the order to 2.5 to 3 percent of GDP for the Euro area, and between three-quarters and one percent of GDP for the United States.

QUESTION: Where are those numbers in chapter two? Specifically which page?

MR. OSTRY: They're not mentioned in the text but you can infer them from the annex on the investment equation. The first set of numbers, you can infer them. You have to do a little bit of arithmetic but they're not explicitly mentioned in the text.

MR. ROBINSON: We've done the arithmetic for you, Mark.

QUESTION: I'll take your word. Thank you.

MR. ROBINSON: Let's perhaps move on to another question, please.

QUESTION: Thank you. Sir, the Fund has said, a number of times, that it stands ready to take necessary measures if the scenario in Iraq is less than optimistic, if things go badly basically, and yesterday The World Bank said that the tools available to international financial institutions are kind of nearing their limit at this point, macro economic tools.

So what exactly, what measures are you ready to take if things go badly in Iraq

MR. ROGOFF: I'd just say we welcome the question but we're not going to answer it till our press conference next week on chapter one. I apologize, but we're just not able to talk about that today.


MR. ROBINSON: Okay. Let's move on to another question, please.

QUESTION: Hi. A question about the chapter on housing bubbles. First of all, do you distinguish between a bubble—in the United States have we had a—some people say we've had a housing bubble. Would you go so far as to say that?

Number two. If in fact we're in a boom, what factors would you look for as being the likely determinants of a bust and are those factors present here in the United States today?

MR. ROBINSON: Thank you very much. I'll say something on the first bit and I don't know if Jonathan Ostry wants to say something on the second.

Well, I think Ken cited the numbers for the run-up in housing prices in the United States, I think it was 27 percent since the trough. That is above the number we define as a boom, but you must remember that's defined in a very statistical fashion. It's basically the top 25 percent of increases in history.

So it's a high number and it's a high number relative to the past experience in the United States as well. Is it a bubble? I don't think I would be ready to say it's a bubble at this stage, in part because there are economic reasons, why housing prices should have increased and will be sustainable.

One of those is simply immigration and the increasing number of people in the United States and another of those is the increasing ability of low-income people to borrow for housing, which means there's simply greater demand for housing.

But I think it's something that we do need to keep our eye very much on in the United States because it is historically high, and I think also in particular regions of the United States where the increase has in fact been much larger than 27 percent, that is also something that needs to be watched quite carefully.

Let me maybe ask Jonathan to address the second part of your question.

MR. OSTRY: Okay. Jonathan Ostry here. I'll basically echo something that David just said. You asked what factors would determine a bust. The study is a purely statistical study, so it's looking at peak to trough declines and ranking them, and choosing the top 25 percent declines and defining these as a bust.

Now historically, of course, you know, a number of factors have been present in previous housing price busts, including a tightening of monetary policy, and others, and obviously the probability of having a bust would depend on what the underlying economic environment is, including monetary policy, and clearly, therefore, the easing of monetary policy in the United States would affect the probability one would assess of the likelihood of a bust.

MR. ROGOFF: I would just add to that, or to reiterate what I said in my opening remarks, that it's notable that the 27 percent rise in real housing prices, that is, prices above inflation from our statistical measure of trough to its current level, is higher than we've seen at any previous time in our data set for the United States.

MR. ROBINSON: Yes. Okay; thanks very much. Let's move on.

QUESTION: If I could follow up on the housing question. It seems from the data that we had a housing bust after both of the last recessions, the '90-'91, and after '82. Am I reading that correctly? Alan Greenspan has been asked this question and he says he does not think that we have a housing bubble in the United States.

Is he seeing something that you're not?

MR. ROBINSON: Well, let me have a stab at a bit of that and Jonathan may want to add. I mean, I think in my answer to the last question, I was quite careful not to say that there was a housing bubble, necessarily, in the United States because there are, you know, fundamental reasons why housing prices might increase.

Nonetheless, as Ken pointed out at the beginning, and just again, it is the highest increase in real housing prices that we've seen in the United States and so that's something that does need to be watched, I think quite carefully.

On the first question you asked, yes, housing busts are clustered around times of recession and I think that is partly because, you know, underling that are a bunch of common shocks which affect the economy and also affect housing and cause these things to happen together. But let me ask Jonathan to elaborate a bit on that.

MR. OSTRY: Jonathan Ostry here. You mentioned '91 and '82. Of course as Ken mentioned at the outset, this is in fact, in our statistical study, the first time that housing prices in the United States have reached a threshold of a boom.

So clearly, in the '91 and '82 episodes, which you pointed out, there were busts. There was no boom in the United States.

In terms of the impact on the real economy, of busts, I think the sense of your question is correct. The impact of housing price busts, when they do occur, is much more significant on the real economy, probably double, on average, the impact of an equity price bust.

So although equity price busts occur more frequently than housing price busts, the impact of housing price busts on the economy is much more severe.

MR. ROBINSON: Okay. Let's move to the next question, please.

QUESTION: Yes. I'd like to pursue the housing question for the U.K. That 70 percent rise that you spoke of initially, what's the correlation between that rise and what was happening on the U.K. equity market and does it follow from that big rise, which I think you said was unprecedented, that the likelihood of a decline in Britain is greater than it is in the States, and as a subsidiary, I just wondered if you looked a bit at San Francisco, because three years into-

MR. ROGOFF: Are you moving there?

QUESTION: Three years into the tech bust, we find that there's been a leveling off but no actual decline, and I just wondered if you might say something about San Francisco

MR. ROGOFF: Well, I think it would be reasonable to infer, that given that we have for the United Kingdom a 70 percent rise from trough to current level versus 27 percent for the United States, that the risks of a bust in the United Kingdom are greater than the United States, and as David—on San Francisco, David commented that we're looking at the United States as a whole, and if you look—clearly, many people have written, if you look at certain regions, the increase in, cumulative increase in prices has been much greater, and therefore they would correspondingly likely be at greater risk.

MR. ROBINSON: Just to add one final point. I think one of the conclusions in the study is also that the, you know, the size of the ensuing fall in prices is often, is related to the size of the run-up in prices earlier. So, clearly, if you have a larger increase in housing prices, there's a greater prospect of a larger fall following.

Let's move on to the next question, please. Are there any other questions?

QUESTION: Hi. Just another follow-up on the housing point. Ken, earlier you said that there was a significant chance, as much as 40 percent, of a bust in the housing market. You also said that there was a great correlation between housing busts across the OECD, and while David said that he didn't see it as a bubble in the United States, et cetera, Ken, the huge run-up in prices in the United Kingdom, if it brings on a bust there, can it bring down the United States' housing market?

MR. ROGOFF: Housing markets are a piece of global interlinkages. There's a significant correlation across the major economic regions in business cycles and economic activity and if I had to throw out a number, something like in the range of 50 to 60 percent, and some of that is due to trade linkages, some of that's due to financial linkages, and some of that's due to common shock such as oil, technology, and our best guess is that reason, these correlations in the underlying economies, are the largest part of why we see the housing market busts tend to be clustered.

They don't exactly happen on top of each other by the month. We were noting they were clustered around three year periods. But that seems to be the most likely reason. That said, housing is a significant component of wealth across countries and given that there are linkages—trade, finance, business cycles, a housing bust in one country will have spill-overs to the rest of the world.

David Robinson, do you want to add anything, or—

MR. ROBINSON: No. I've nothing to add to that. Shall we maybe move to another question.

QUESTION: Another on housing. You talk about the need for monetary authorities to watch this carefully. What can they do about a housing bust?

MR. ROBINSON: Well, I think this is, you know, a part of a much bigger question about how monetary policy can and should react to assets price run-ups and busts, in general, and I think it is a very difficult question and obviously one that has exercised both the United States Federal Reserve and of course is still exercising the Bank of England.

But I think, in general, the answer that we would give is that the possibility of a bubble in a run-up in asset prices is one of the factors that central banks can and must look at in terms of the impact on demand, looking forward, and therefore on inflation. So it's one of the factors that must go into the calculation of monetary policy, looking forward.

But I don't think that we can say, and we wouldn't say, that monetary policy should be entirely directed at asset prices such as housing or equity prices. That would be going much too far. It's one of the factors that they need to consider in making a decision, but focusing, ultimately, on the impact on inflation. But Ken, you may want to-

MR. ROGOFF: Well, David gave a very good answer. First of all, it's a very tough question. I don't think anyone really knows the answer. There's a debate. We've had research papers coming out of the research department at the IMF, giving both sides of it, papers arguing that you should just ultimately be looking at the inflation impact. So if you're concerned about a housing bust, fine, but you need to ask what are the effects going to be on inflation. We've had other papers arguing that if a big run-up in housing prices is accompanied by a large increase in indebtedness, that itself should be something monetary policy should look at directly, and I think the fact we've had papers giving either side of this reflects that there's just a debate on this question and nobody knows the answer, for sure.

Certainly we wrote a chapter about it, it occurred to us to write about it because we thought it was a concern.

MR. ROBINSON: Okay; thank you. Let's move on to the next question, please.

QUESTION: Yes. [inaudible] of some drag on GDP growth from problems in corporate balance sheets, what would be the possible role of the increasing public debt here in the U.S. on the ability of the U.S. companies to recover capital spending?

MR. ROGOFF: I think we have to leave that to next week. I mean, it clearly intertwines with what the effects will be on interest rates. But we'd like to answer that next week.

MR. ROBINSON: Yes. Please ask it next Wednesday.

MR. ROGOFF: Absolutely.

MR. ROBINSON: Okay. Another question, please.

QUESTION: First a technical question. I don't know if this is in the report. But how much of a, what portion of the boom is typically given back in the bust in the housing sector?

My second question is Mr. Ostry said that one precipitating factor in a bust is often tightening in monetary policy.

I'm wondering—at some point the Federal Reserve in the United States will of course have to begin to raise interest rates from their extremely low levels now, and of course the very low level of interest rates right now is one of the things that's keeping housing as buoyant as it is today.

Could that inevitable tightening of policy be the precipitating factor in a bust here in the United States?

MR. ROGOFF: First, narrowly, to answer your first question, what we find is that conditional on there being a boom, given that there's a boom, there's a 40 percent probability of a bust. This is on average, across all the OECD countries in our sample, cause if we weren't looking across, all these countries wouldn't be able to do anything because there are only a couple examples, typically, of busts for each country.

Clearly, as Jonathan mentioned, in the past many housing busts have been associated with increases in interest rates, tightening of monetary policy, but trying to put that question into the whole conjuncture is something, we'll wait till next week.

I don't know if anyone wants to add to—

QUESTION: Ken, my question was more precisely—you have typically like a X percent run-up in prices during a boom. What I'm wondering is what proportion of the X percent is given back in a bust?

MR. ROGOFF: I'm sorry. I apologize.

MR. ROBINSON: I'm not sure we have that number at hand, but if you leave us—if you get in touch with Bill Murray afterwards, we will get back to you and give you the answer because it's a very reasonable question.


MR. ROBINSON: Let's move on to the next question.

MR. MURRAY: Thanks everyone, for joining us today. Again, the embargo is 2:00 p.m. Washington time, 1900 GMT. If you do have any follow-up questions, drop me an e-mail, give me a call, and we'll chase it before the embargo is lifted. Again, thanks for joining us and look forward to seeing you or hearing from you next week at the WEO Chapter One press conference.


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6278 Phone: 202-623-7100