Journal Issue
Share
Article

Revisiting the fiscal deficit-inflation puzzle

Author(s):
International Monetary Fund. External Relations Dept.
Published Date:
June 2003
Share
  • ShareShare
Show Summary Details

IMF Survey: Macroeconomic theory tends to associate fiscal deficits with inflation. Why have most empirical studies been unable to uncover a strong connection between the two?

CATÃO: Most of your readers will probably find this as puzzling as we did when we started working on this topic. The inflation literature seems to suffer from a kind of split personality. Well-established theories postulate that fiscal deficits eventually lead to inflation. But living side by side with these well-established theories are empirical studies that show that there is no statistically significant relationship between fiscal deficits and inflation across a panel of countries. At most, these studies have found a very weak relationship. Since we work for an institution that preaches the virtues of fiscal rectitude, we were rather unsettled by this dichotomy. We took it upon ourselves to take a more in-depth look at the issue, building on earlier research we did for a WEO [World Economic Outlook] chapter.

One question we sought to address is why, if there really is a significant and reasonably strong relationship between fiscal deficits and inflation, other studies have not been able to uncover it. We were certainly aware that it was no mean task to pin down this relationship, and one of our main suspects was the econometric methodology.

Fiscal deficits and inflation should not necessarily be related in the short run because governments that run fiscal deficits can, at least temporarily, finance the deficit through borrowing. So, for a while, you may see no relationship at all. But it is bound to show up in the long run, because governments cannot accumulate debt indefinitely. We applied a state-of-the-art technique that allowed us to look at this long-term relationship in a large panel of countries, and that technique is one reason our results differ from others in the literature.

Fiscal deficits and inflation should not necessarily be related in the short run because governments that run fiscal deficits can, at least temporarily, finance the deficit through borrowing.

—Luis Catão

In our mind, the second culprit was that other studies did not allow for the fact that fiscal dominance—when the need to finance government deficits outweighs other policy concerns—varies widely across countries. For instance, fiscal dominance has been very weak in the United States over the past couple of decades or so, and this has also been so in several other advanced countries. In contrast, the fiscal authorities clearly call the shots in several developing couninearity that has been overlootries, although it is important to acknowledge that new institutional arrangements in some of them have managed to tame fiscal dominance quite a bit in recent years. So we separated countries into groups that broadly differ in their degree of fiscal dominance.

The third culprit is the nonlinearity of the relationship between fiscal deficits and inflation. As you know, the higher the inflation rate, the higher the cost of holding cash or sight deposits, so the stock of narrow money [M1] in the economy tends to shrink as inflation rises, reducing the inflation tax base. As with any other tax, the smaller the base, the higher the tax rate has to be to maintain the same amount of revenues. As people hold less and less money, inflation has to be higher to finance a given deficit. So the higher the level of inflation, the stronger the impact of fiscal deficits on inflation. It follows, then, that an econometric model that seeks to pin down the theory must allow for this nonlinearity. The paper proposes a clean and simple way to incorporate this nonlinearity in the estimates: we simply scale the fiscal deficit by M1 rather than by GDP. By doing this, we take into account the nonlinearity that has been overlooked in some previous studies.

IMF Survey: How do your results differ from others?

Terrones: As Luis has already noted, our study differs from others in three important areas: first, it models the long-term relationship between deficits and inflation as nonlinear; second, it uses a new econometric technique that allows us to distinguish between the long- and short-term effects of fiscal deficits on inflation; and, last but not least, it covers a greater number of countries—107 countries from 1960 to 2001, for a total of 3,600 observations—a far larger sample than any other study we know of.

Based on these considerations, our study finds a strong and statistically significant relationship between fiscal deficits and inflation across a broad range of developing countries. In particular, the relationship is quite precisely estimated for emerging market countries, where we found that a sustained reduction in the government deficit by 1 percent of GDP is associated with a drop in annual inflation of 2-6 percentage points, depending on private sector holdings of narrow money.

Another interesting result is that the relationship between deficits and inflation is even stronger if the sample is limited to the 26 countries in our sample with the highest inflation rates over the past four decades. This is what one would expect, given the nonlinear way in which deficits and inflation are related.

Similar to other studies, ours finds no significant relationship between fiscal deficits and inflation in advanced and low-inflation economies, which is consistent with the fiscal dominance story that Luis just mentioned. Indeed, there is increasing evidence that advanced economies adjust their primary fiscal balance in direct proportion to their stock of debt.

But, while the conclusion that fiscal deficits are not a significant cause of inflation in advanced economies is consistent with the findings of previous studies, our results for developing countries are not. For instance, a recent well-known study reports a significant relationship between deficits and inflation only for high-inflation and hyperinflation countries, while we find that this relationship holds across the board, including in countries with moderate inflation. In addition, the effects we get for high-inflation countries are stronger than previously found.

IMF Survey: But presumably fiscal deficits are not the only cause of inflation. For advanced economies, for instance, oil prices may be a factor. For developing countries, a few other factors may matter, too.

CATÃO: Indeed. That’s the reason we considered oil as well as other variables that may influence inflation, such as exchange rate regimes. There is a view, for example, that holds that fixed exchange rate regimes impose discipline on policymakers. These regimes are said to prevent excessively expansionary monetary and fiscal policy, and that, in turn, contributes to lower inflation. The tricky part, however, is to sustain the peg.

In my research on the economic history of developing countries, particularly in Latin America, I have seen the pattern repeated over and over again. Governments are able to keep the peg whenever conditions in the international capital markets are favorable. Capital flows into these countries, reserves accumulate, and the economy expands, while the exchange rate remains fixed. But then, debt builds up and the foreign investor is no longer willing to continue pumping money into the country at the same rate, debt-servicing problems surface, the peg is no longer sustainable, and deficits have to be financed through seigniorage [money creation]. In other words, inflation is back in play. So, in the long run, we don’t really observe a systematic relationship between fixed exchange rate regimes and inflation. In other words, our conclusion that fiscal balances really do matter holds up when the exchange rate regime is included in our model.

The other variable we looked at is openness. We wanted to take into consideration an influential view that followed from David Romer’s 1992 paper, which argued that countries more open to trade tend to have lower inflation. The mechanism through which this relationship is established is, again, policy disci-pline—the benefits of a more expansionary monetary policy are less for countries that are more open to trade, and, therefore, inflation in these countries tends to be lower.

But reality tells a different story. Take the example of two large and relatively closed economies—the United States and Brazil. The United States has experienced low inflation, on average, for a very long time, while Brazil has suffered from very high inflation, on average, for a very long time. And several small open economies have historically experienced relatively high inflation rates. Indeed, a study by Dani Rodrik argues that when a country is smaller, the government tends to be bigger as a proportion of GDP. Bigger governments are not necessarily profligate, but they do have the tendency to want to spend more. The bottom line is that we observe no systematic relationship between openness and inflation.

IMF Survey: Fiscal deficits may reflect other factors pertaining to institutions and macroeconomic landscapes. Wouldn’t these also help explain the different inflation outcomes across different country groups historically?

Terrones:Surely. First, one reason institutions matter is the extent to which countries differ in their capacity to avoid persistent deficits. Normally, countries with weak fiscal institutions and fragile political systems tend to have persistent deficits and, thus, higher of inflation. This has long been acknowledged in the literature.

In countries with fragmented political systems and highly skewed income distribution, the diverse interests of social groups often make it difficult for a government to take corrective measures to eliminate deficits in a timely manner.

—Marco Terrones

For instance, in countries with decentralized fiscal institutions, the so-called province effects—when local governments increase their expenditure and overlook the impact on the consolidated budget deficit because they assume only part of the cost—are likely to be an important source of fiscal imbalances. Also, in countries with fragmented political systems and highly skewed income distribution, the diverse interests of social groups often make it difficult for a government to take corrective measures to eliminate deficits in a timely manner. In both cases, public debt will accumulate beyond sound levels.

Limited domestic financial development also constrains the amount of domestic financing that governments can obtain. Countries with thin domestic financial markets must rely on foreign financing. To the extent that foreign financing becomes costly and international capital markets are periodically closed to them, governments have no option other than to finance their fiscal gaps through money creation, thus producing inflation. The problem is obviously aggravated when the government inherits a large external public debt from previous administrations and the country’s traded sector is small, making it difficult to produce the large trade surpluses necessary to pay the debt. Not surprisingly, a tight positive correlation is observed between high external indebtedness in relatively closed economies and inflation, especially in the 1980s.

IMF Survey: It’s probably fair to say that one motivation for your study is the concern over the negative repercussions of inflation on welfare. But more recently, deflation seems a more urgent concern, at least in some countries. Does your study have something to say about this?

CatÃo: This is an important question. We found that in advanced economies, especially low-inflation advanced economies, fiscal deficits are not significantly related to inflation. This finding suggests that fiscal deficits are not very effective tools for tackling deflation, and I understand that this is consistent with Japan’s experience over the past 10 years or so. Other instruments, such as monetary policy, seem much more helpful, especially if used in a timely fashion to prevent deflation in the first place. At the same time, one important side effect of fiscal expansions is the pressure on long-term interest rates and their potential contractionary effect in the long run. Very large and persistent fiscal deficits that result from, say, a huge tax cut may do a great deal of harm to an economy in the long run without bringing about the intended effect of avoiding deflation.

IMF Survey: Overall, then, what lessons can policymakers draw from your study?

Terrones: The most obvious lesson is that fiscal rectitude helps keep inflation at bay, especially in developing countries with histories of chronic inflation. Devices like exchange rate pegs are not magic bullets and, as recent developments in emerging markets have shown, can have disastrous consequences. Likewise, measures to foster trade openness and participation in regional trade arrangements, although beneficial in many respects, are not substitutes for fiscal discipline.

I should also emphasize that while developing countries have made great strides toward fiscal consolidation over the past decade, significant imbalances remain, and long-term fiscal solvency issues are by no means an issue of the past in many countries. History has shown, time and again, that fiscally dominant governments with sizable debts and shaky fiscal positions are a sure recipe for inflation, even if not immediately.

Another important lesson is that the success of new monetary arrangements, such as inflation targeting, hinges partly on the stance of fiscal policy. This is a direct consequence of the strong interlinkages between monetary and fiscal policies resulting from the government’s budget constraint. Accordingly, the success of institutions and of relatively new arrangements, such as central bank independence in many developing countries, depends not only on well-drafted laws but, more important, on firm political commitment to fiscal discipline.

Copies of IMF Working Paper No. 03/65, “Fiscal Deficits and Inflation,” by Luis Catão and Marco E. Terrones, are available from IMF Publication Services for $15.00 each. See page 166 for ordering details. The full text is also available on the IMF’s website (www.imf.org)

Other Resources Citing This Publication