"From Here to Eternity: The Outlook for Fiscal Adjustment in Advanced Economies" By Carlo Cottarelli

May 20, 2013

By Carlo Cottarelli
Peterson Institute for International Economics
May 20, 2013

As prepared for delivery

Thank you very much Adam for these nice words and for inviting me here. Today, I am going to talk about a topic that has been keeping us busy for the last few years and, as the title of my presentation suggests, will keep us busy for a few more years, as, whether we like it or not, high public debt is going to be with us for quite some time.

My presentation will be based on our Fiscal Monitor, the IMF publication that we introduced in 2010 to monitor fiscal developments around the world. I will address four main questions

  • First, how bad is the fiscal position in advanced economies? It is pretty bad. Indeed, one could argue that, in many countries, it has never been as bad as it is today.
  • Second, what are the risks arising from the poor state of the fiscal accounts in these countries? High for some of them. For others, the immediate risks are not so high. But the situation would be difficult for all in the long run, in the absence of further adjustment.
  • Third, what is the appropriate strategy to address these fiscal challenges? We continue to believe that fiscal adjustment should proceed, but at a moderate speed and with a medium-term orientation, at least for countries not subject to severe market pressures.
  • Fourth, is this strategy of gradual fiscal adjustment working? I will argue that it is, although major challenges still lie ahead.

One issue that I will not discuss, for brevity, is the merit of what we can call non-orthodox approaches to fiscal adjustment, like inflation or debt restructuring, but I will be happy to take questions on this issue, and I am sure there will be some.

So, let me start with the first question: How bad is the fiscal position in advanced economies?

Some of you may have already seen this figure depicting the public debt to GDP ratio for advanced economies since 1880. The ratio is currently at about 110 percent, close to the peak reached during the Second World War. Public debt is also at a peak with respect to private sector wealth, although the available time series here is much shorter.

But the outlook is actually worse than at the end of the Second World War for four reasons:

  • First, most of the fiscal adjustment then consisted of cuts in military spending, which are easier to implement from a political perspective.
  • Second, spending for pension and health care will be on the rise over the coming decades, and by several percentage points of GDP. So countries in need of fiscal adjustment will have, so to speak, to swim against the tide of rising spending pressures.
  • Third, long-term growth prospects are not great, which is bad as growth erodes the debt-to-GDP ratio. Long-term growth will be lower than in the post-war period when the labor force was rising rapidly. We now project a decline in the labor force in advanced economies.
  • Fourth, financial repression in the 1950s and 1960s, in the form of interest rate ceilings and investment requirements in government paper, facilitated the financing of public debt. These restrictions are no longer in place.

Not all advanced economies, however, are equally in trouble. Leaving aside some smaller countries, fiscal problems are particularly severe in 10 countries, where debt ratios are high and still rising: United States, Japan, UK and seven euro area members (France, Italy, Spain, Belgium, Greece, Portugal and Ireland). They represent about 40 percent of world GDP so what happens to their fiscal policy has major implications for the rest of the world.

Let’s have a closer look at these countries. This figure will show on the horizontal axis the increase in the debt to GDP ratio since 2007 and on the vertical axis the level of the deficit to GDP ratio also since 2007. I am going to show you first the other advanced economies: these economies started, on average, with a small surplus in 2007; deficits and debt increased in 2008 and 2009, but since then things have improved rapidly with declining deficits and essentially no increase in debt ratios. This contrasts with the behavior of our ten countries. Here is the average, with deficit and debt rising sharply in 2008, 2009 and 2010, and with declining deficit but still rising debt ratios since then. And here is the behavior of individual countries which clearly belong to a different class.

The magnitude of the fiscal challenges faced in the years ahead by these countries can be illustrated by our standard long-term adjustment scenarios, which focus on the degree of fiscal tightening that will be needed in the future to lower public debt from the current level to, say, 60-80 percent by 2030. Let’s clarify by showing you one of these scenarios. In our scenarios the adjustment is gradual. Here you see the primary balance of the government—that is the difference between revenues and non-interest spending—improving gradually by 2020 and then stabilizing at the required level for another ten years. The debt ratio keeps rising for a while, but then declines, hitting its target, 60 percent of GDP by 2030. We run these scenarios for each country. Before showing you the results it is important to recognize that the difficulty of implementing this adjustment can be measured by two indicators: first, the size of the adjustment following 2013; second, the level of the primary balance that will have to be maintained over time.

The results of these scenarios will be reported in this chart: on the horizontal axis you find the first indicator—the needed improvement in the primary balance over the next seven years—while on the vertical axis you will see the level of the primary balance that will have to be maintained until 2030. Each country is represented by a set of points, not a single point, because we run scenarios that vary in terms of the specific debt target, 60 or 80 percent, and the assumptions on interest rates and growth, which are key determinants of debt dynamics. The legend tells you exactly what the assumptions are for each scenario.

The cross country ranking is fairly clear: France and Belgium are in a relatively stronger position, with required adjustment of about 2 percentage points of GDP after 2013 and targets for the primary balance levels of 2-3 percentage points of GDP. Most other countries are in this central and more challenging pack, with two exceptions: Italy, whose “solar system” centers around the vertical axis, implying that not much further adjustment is needed, but at a higher level than other countries, implying that the level of the primary surplus it needs to maintain for many years is very high; and Japan, which is the clear outlier.
So, there are important differences across countries, but running these large fiscal surpluses over time is likely to be challenging for all of them. So one question could be: what happens if instead there is no further adjustment and public debt remains as high as it is now?

In other words: what are the risks arising from the poor state of the fiscal accounts?

Concern about high public debt has focused on two very different aspects.

The first one is the effect of high public debt on long-run economic growth. This concern is most often linked to the names of Carmen Reinhart and Ken Rogoff, who found an historical correlation between debt exceeding 90 percent and low growth. Leaving aside the issue of the recently found problem in their spreadsheets—I am sure you have heard about this—the focus on the Reinhart and Rogoff results has been excessive, in my view, from the beginning. First, their work did not take into account that the correlation between growth and debt can run from low growth to high debt and not vice versa. Second, their results were based simply on statistical averages rather than on econometric estimates of how high public debt affects growth.

However, several econometric studies have focused on the relationship between debt and growth and, with one exception (the Panizza and Presbitero paper), find evidence of a relationship that likely goes from high debt to low growth. Without getting into much detail, this chart does show a negative relationship between debt levels and growth, noting that here growth is growth in the five years following each debt level observation, thus at least suggesting a causality that goes from high debt to low growth.

The second reason why we should be worried about high public debt relates to the risk of a fiscal crisis. In theory, the risk of fiscal crises caused, in particular, by self-fulfilling expectations is high for high debt countries because, when public debt is high, an increase in interest rates can really make a difference between what is sustainable and what is not sustainable.

In practice we do see that high debt does matter for the perception of risk. In this chart you see for each country the deficit and debt ratio in 2012 and each country is represented by a balloon whole size is proportional to the size of the CDS spread: as you can seen, almost all large balloons fall in this area which corresponds to a combination of high deficit and high debt: so, large fiscal imbalances do matter for the perception of risk. They are not a sufficient condition to be in trouble—you see many small balloons even for some high debt countries—but they are at least a necessary condition to be in trouble.

Of course, as I said, some high debt countries are still enjoying record low interest rates, like US, UK and Japan. This benign market attitude can be explained at least in part by the surge in base money and the consequent monetization of public debt that we have observed since 2007, as this figure shows for our three countries.

This surge in base money has not been inflationary because it has been matched by an increased demand for base money. Over time, as conditions normalize, the share of public debt financed by printing money will have to decline—the so called exit strategy problem—which will tend to raise interest rates on government paper. While this normalization of interest rates is expected to take place in the context of higher growth, the net impact on the differential between interest rate and growth, which is what matters for the dynamics of the debt ratio, is likely to be positive, thus putting pressure on high debt countries.

Altogether, maintaining public debt at high levels over the long run does not seem to be a great idea. There is probably broad consensus about this, although there are clearly differences on why high public debt is bad, on exactly how high it needs to be before bad things happen, and on the urgency of the case for lowering public debt.

Let me therefore come to the third question of my presentation. What is the appropriate strategy to address these fiscal challenges, including in terms of urgency of the fiscal adjustment? In this austerity debate, we have the view of the fiscal doves, who argue that there is no immediate fiscal problem and that there is even room to expand, and the view of the fiscal hawks who believe that the price to be paid for committing the sin of fiscal profligacy will be very high. Our position since 2010, when Olivier Blanchard and I published a blog post entitled “Ten Commandments for Fiscal Adjustment in Advanced Economies,” is an intermediate one between these extreme positions.

Let me summarize the four main elements of our position:

  • First, for countries that are not under market pressure, fiscal adjustment should proceed at a moderate pace within a medium-term fiscal adjustment plan to enhance credibility.
  • Second, these medium-term plans should be defined in cyclically adjusted terms so as to allow the automatic stabilizers to operate if growth differs from initial expectations.
  • Third, countries under market pressure should proceed more rapidly but also for these there is a speed limit beyond which fiscal adjustment can be counterproductive.
  • Finally, this fiscal strategy should be supported by structural reform to boost medium-term growth and by relaxed monetary condition to cushion the contractionary effect of fiscal tightening and to reduce the risk of self-fulfilling fiscal crises.

We have often argued that this position is based on the recognition that there are costs if you tighten too much—essentially not enough short-term growth—and there are costs if you do not tighten enough—arising, for example, from the risk of a fiscal crisis. However, this, per se, does not justify taking an intermediate position. What is critical, instead, is the existence of nonlinearities that arise from taking extreme positions. In other words, the cost of excessive austerity and of excessive profligacy can be particularly high. In this case the total cost function has a minimum for intermediate policies, as illustrated in this chart.

Let’s look more closely at the costs of fiscal austerity and the costs of fiscal profligacy. As discussed in a recent blog by Olivier Blanchard and Daniel Leigh, the costs of fiscal tightening are now high because:

  • First, fiscal multipliers are particularly large now reflecting the downward inflexibility of nominal interest rates, the difficulties banks have in granting credit, particularly in Europe, the large share of households that are credit rationed, and the fact that, in a recession, cuts in demand affect more real output than prices, a point discussed in the April 2012 Fiscal Monitor. These factors are likely to operate to produce relatively higher multipliers regardless of the composition of the fiscal adjustment, whether it is done on the revenue side or on the spending side. Governments should still calibrate the composition of the adjustment to minimize its short-term and long-term growth impact—for example by avoiding cutting investment spending and avoiding tax increases in countries where tax rates are already high. But it would be unrealistic, in the current circumstances, to hope to engineer an expansionary fiscal contraction focused on spending cuts of the kind discussed by Giavazzi and Pagano in their famous 1990 paper because the mechanisms that would stimulate growth following a fiscal tightening—a drop in interest rates or an exchange rate depreciation—would not operate now for most countries in need of fiscal adjustment.
  • The second reason why fiscal tightening has a particularly high cost now is that the cost of an output loss is larger when the economy is already growing slowly. We tend to dislike extremes (indeed that is the basis for running countercyclical policies). We have some evidence that markets dislike extremes too. Laura Jaramillo and I wrote a short note on the effect of large fiscal tightening on interest rates. The results are illustrated in this figure, showing that when the multiplier is sufficiently high and the tightening sufficiently strong, interest rates may rise, not decline, following a fiscal tightening, This happens because markets get worried about growth as well as about the rise in the debt to GDP ratio that may accompany the fiscal tightening when GDP declines, an effect noted in a recent paper by Luc Eyraud and Anke Weber, two economists from the IMF’s Fiscal Affairs Department.
  • Finally, there may be higher than usual hysteresis effect, as noted by Brad de Long and Larry Summers, because of the length of the current recession and the related high number of long-term unemployed.

Let’s now look at the other side, the costs arising from the fiscal profligacy. I already mentioned that these arise from the long-term growth effects of high public debt and the risk of a fiscal crisis. Let me therefore just underscore three issues regarding these costs.

First, the risk of fiscal crises could be eliminated if there were a commitment technology that would make sure that fiscal tightening took place later, in good times. In this case promising fiscal adjustment would be enough. Unfortunately such a technology does not exist. Things are made more difficult by three factors. First, in the past advanced economies did not use their good times to lower public debt which was, it is worth recalling, already at record levels in 2007 before the crisis. Second, this fiscal twist (expand now, tighten later) was already danced in 2008-09: so countries that had promised to tighten later are supposed to tighten now. Third, gaining credibility by implementing long-term pension and health care reforms—which affect long term fiscal trends at low cost for output today—may not help much because, unfortunately, interest rates do not seem to be much correlated to pension and health care spending increases. Altogether, just promising fiscal adjustment is unlikely to be enough.

Second issue related to the cost of fiscal profligacy: we tend to underestimate the cost of fiscal crises before they occur for two reasons. First, fiscal crises are typically nonlinear events that are not easily detectable before they happen in full swing. The timing of these crises is highly uncertain—as we often say, it is a situation of Knightian uncertainty. Second, these fiscal crises do involve large output losses but these costs are observed only after the crisis materializes and so we tend to focus on the observable costs of fiscal austerity. Altogether, there may be a tendency to ignore these costs. Yet, one can expect that the risk of a confidence crisis becomes increasingly large the larger the fiscal imbalances, which is an important source of nonlinearity.

Third issue: the risk of turmoil on the government paper market is not the same for all countries. A surge in interest rates due to an increased risk premium, which would be quite dangerous in the presence of high debt, is a very unlikely event for countries like the United States or Japan given their strong investor bases. It would be a “black swan.” But, black swan events do happen and their costs can be pretty high. It is a tail risk that cannot be ignored and that will persist as long as public debt remains high. When debt is high even a modest increase in interest rates has severe effects, particularly in countries, like Japan, where banks have invested massively in fixed yield government paper. Again, I am talking about low probability events, but their cost, if they materialize, would be huge.

The bottom line is that, faced with high costs from extreme solutions—too much austerity and too much profligacy—the most sensible course of action is to proceed with fiscal adjustment at a relatively moderate pace, of course taking into account that for countries that are already under market pressure the risk of a fiscal crisis is of course larger, which calls for a faster adjustment.

Given this approach, which countries are moving at the appropriate speed, and which are not? The Fiscal Monitor identifies some cases where fiscal adjustment is not proceeding in line with our prescriptions:

  • On the one hand Japan is moving too slowly. Its deficit remains stuck at about 9 percentage points of GDP and its debt keeps rising fast. Japan has become an outlier in terms of fiscal performance, as evidenced by this three-dimensional chart showing the deficit, the primary deficit and the net debt ratio for 2013. It was not an outlier in 2010, as you can see. Japan needs to define quickly a medium-term adjustment plan to lower its public debt. It will have to go well beyond the currently envisaged adjustment measures, such as the increase in VAT from 5 to 10 percent.
  • On the other hand, the US is adjusting too rapidly this year. The fiscal cliff has been avoided but the fiscal tightening this year will still be the largest one in the last three decades. A more moderate pace of adjustment, in the context of a medium-term adjustment plan, would be preferable.
  • Finally, a few European countries that have suffered from chronically weak private sector demand could at least consider smoothing the pace of the adjustment. But in most European countries the underlying pace of adjustment is appropriate in terms of measures. The main risk to avoid is targeting headline deficits regardless of growth developments. In this respect, we see more flexibility on the side of European institutions. While the Stability and Growth Pact targets are still set in terms of headline figures, these targets have been revised significantly over time, as illustrated by this chart.

Let me address the last question: is this strategy of steady fiscal adjustment working and is it realistic to continue it for the future?

There has been considerable progress since 2010 in fixing the fiscal accounts. Two milestones will be reached in 2013. First, the average deficit of advanced economies will be about half of what it was in 2009, from 9 percent of GDP to little more than 4½ percent. Second the average debt ratio—weighted by GDP—will stabilize, indeed it will slightly decline. So fiscal adjustment, while costly in terms of output, is yielding some results

There is no doubt, however, that major challenges still lie ahead. Let’s go back to the chart showing the additional adjustment needed for each country to bring down debt to 60-80 percent of GDP by 2030. How challenging are these targets? To answer, we have looked at what countries did in the past when they really tried to lower their debt ratio. Most specifically, for each advanced economy we have looked at highest primary surplus sustained over a ten-year period during the last sixty years. This is the corresponding distribution which has a median of 3¼ percent. We have also looked at the largest improvement in the primary balance over a seven-year period (as in our scenarios countries still have seven years ahead to complete the adjustment) following a three-year period of adjustment (as countries have already been adjusting for about three years, so they may feel some fatigue). And this is the distribution with a median of about 5 percentage points of GDP. We have then used these distributions to identify in our summary chart areas of increasing difficulty corresponding to increasing quartiles of the distribution. So areas closer to red are more difficult to achieve.

Most countries are now to the left of this vertical line representing the median of the distribution of seven-year fiscal improvements, implying an additional required adjustment that is below the median of what was achieved when countries in the past really tried hard. So not impossible, although obviously still hard, as we are comparing the required adjustment with what were the best past performances. Most countries are, however, above the median in terms of primary surplus level which would have to be maintained over time. The bottom line is that, while the required levels of primary surplus are now not out of reach, maintaining those primary surpluses over time will require a special effort.

But will it be necessary to maintain a very large primary balance for such a long time? Let me conclude with a note of optimism. I suspect that most investors are more concerned about seeing the debt ratio put on a downward path than they are about the speed with which the debt ratio declines. Once the debt is on a downward trajectory, whether a certain debt target is achieved in 2025, 2030, or 2040 is less important, particularly if fiscal institutions are strengthened to ensure that fiscal plans are perceived as credible. If this occurs, real interest rates may turn out to be lower than those that we assume in these scenarios and countries will have some room to reduce their primary surplus without disturbing the trend of falling debt ratios.

So, to conclude, getting debt ratios back to where they need to be will remain a long process, but hopefully it will not last from here to eternity.

Thank you very much for your attention and you can find on our website more information on the world’s fiscal accounts from our Fiscal Monitor. Thank you.

IMF COMMUNICATIONS DEPARTMENT

Public Affairs    Media Relations
E-mail: publicaffairs@imf.org E-mail: media@imf.org
Fax: 202-623-6220 Phone: 202-623-7100