FRBSF Economic Letter
2001-01; January 12, 2001
Will Inflation Targeting Work in Developing Countries?
Economists are often accused of not agreeing with each other. Believe
it or not, the 1990s may have produced an exception. Beginning with New
Zealand in 1989, a growing number of countries have adopted a so-called
"inflation-targeting" approach to monetary policy (e.g., Great
Britain, Canada, Australia, and Sweden). Although debates about the precise
definition of inflation targeting quickly begin to sound like medieval
scholasticism, the basic idea is that a central bank adopts (or is assigned)
an explicit numerical target range for inflation, and makes achievement
of this target its primary objective. This doesn't mean the central bank
ignores unemployment or the rate of economic growth. As long as inflation
remains within the stated range, the central bank is free (and indeed
expected) to stabilize the economy. However, if at some point inflation
threatens to break out of the permissible range, then the central bank
must make the inflation target its overriding objective.
Although inflation targeting has yet to be recession-tested, economists
do agree that at least so far it has been quite successful. Inflation
targeting countries have experienced low and stable inflation rates, with
no apparent sacrifice in the growth or stability of the economy. (New
Zealand stumbled a bit following the Asian financial crises, but this
is attributed to their unwillingness to let the exchange rate depreciate
enough, rather than to inflation targeting per se.) For once economists
agree on something!
Now, nothing breeds imitation like success, and the apparent success
of inflation targeting has attracted the attention of many developing
countries. The currency crises of the 1990s caused many of these countries
to lose faith in fixed exchange rates as a means to control inflation.
Unfortunately, it is not enough simply to say that you are now going to
adopt a flexible exchange rate policy. Flexible exchange rates are consistent
with a vast array of alternative monetary policies, and market participants
need to have some idea about what this policy is in order to provide an
anchor for their price level expectations. Historically, the anchor most
often proposed was to have the central bank target the growth rate of
the money supply. Experience suggests, however, that this is a poor target
during periods of financial market development and innovation, because
such changes make the demand for money unstable, thereby weakening the
link between monetary policy and the economy. These conditions are likely
to prevail in many developing countries. So, as a result of the disillusionment
with fixed exchange rates and the skepticism about money growth targeting,
many developing countries are seriously considering adopting an inflation-targeting
regime. In fact, Brazil, Chile, and Thailand have already done so.
This Economic Letter evaluates the prospects for the successful
implementation of inflation targeting in developing countries. I argue
that if inflation targeting is to be successful, it is essential to recognize
and account for the structural differences between the economies of developed
and developing countries. Although the defining distinction between the
two groups is in terms of per capita income, I argue that for the purposes
of monetary policy the key differences are in terms of: (i) openness,
(ii) policy credibility, and (iii) financial fragility.
Openness
It is important to realize that while inflation targeting gives the central
bank a goal, it doesn't tell the central bank how to achieve it. Indeed,
this is regarded as an advantage of inflation targeting, as it allows
the central bank to adapt its operating procedures to changing conditions.
However, it does raise the question of what is the best way to actually
hit an inflation target while, at the same time, balancing other objectives.
Most formal analyses of this question are in terms of rules for setting
a short-term nominal interest rate (e.g., the federal funds rate) in response
to changes in inflation and output, the most popular rule being the so-called
Taylor Rule. Taylor's Rule says that the central bank should raise interest
rates when inflation and output rise above their target levels, with the
inflation response being somewhat greater than the output response. Conversely,
when inflation or output falls below its target, the central bank should
cut interest rates.
Although most studies conclude that Taylor's Rule is reliable—and robust
to a host of real world uncertainties—these studies are potentially misleading
when applied to developing countries. This is because Taylor's Rule ignores
the exchange rate, and the exchange rate is especially important in developing
countries due to their reliance on international trade.
Extensions of Taylor's Rule to incorporate exchange rates have been carried
out by Ball (1999) and Svensson (2000). From the perspective of inflation
targeting, exchange rates influence the economy in two principal ways.
First, and most directly, the exchange rate affects the price of imported
goods and, therefore, the inflation rate. Second, the exchange rate affects
the competitiveness of domestic goods on world markets and, hence, the
level of aggregate demand in the economy. Aggregate demand in turn affects
inflation through the so-called "Phillips Curve." Empirical
evidence suggests that the first effect operates pretty quickly, i.e.,
within a few months, since it only requires prices to adjust. In contrast,
the second effect operates with a substantial lag, since it involves buyers
and sellers finding new customers and sources of supply.
Ball and Svensson show that this lag discrepancy can produce poor results
if the central bank is too diligent about hitting the inflation target.
Because the import price effect operates quickly, by frequently adjusting
the exchange rate the central bank can be quite successful at hitting
the inflation target. Unfortunately, these frequent exchange rate adjustments
tend to produce a "whiplash" effect on the rest of the economy,
since they will produce subsequent output responses. Hence, targeting
inflation too closely in open economies can produce instability in growth
and unemployment.
To avoid this whiplash effect, Ball and Svensson show that the standard
Taylor Rule needs to be modified in two ways. First, rather than simply
targeting the inflation rate, central banks in open economies should target
something like a "long-run" inflation rate, which removes any
temporary effects of the exchange rate. For example, if the currency is
regarded as being temporarily strong and likely to depreciate in the future,
then the target inflation rate should leave room for this by being lower
than it otherwise would be. Adjusting the target in this way prevents
the central bank from overreacting to the temporarily low inflation rate
caused by the temporarily strong currency. Second, rather than just use
a short-term nominal interest rate as a policy instrument, central banks
in open economies should recognize the fact that, via asset arbitrage,
interest rate changes produce exchange rate changes that amplify the economy's
response to interest rates. For example, raising interest rates lowers
output and inflation by depressing the level of spending in the economy.
However, in open economies it also appreciates the exchange rate, which
further reduces output and inflation by depressing net exports and lowering
the price of imports. To account for these additional effects, Ball and
Svensson propose the use of a "Monetary Conditions Index," which
consists of an average of the current interest rate and exchange rate.
Policy credibility
Policymakers in all countries are concerned about their credibility.
A policy that would be beneficial if the public believed it would last
could actually turn out to be counterproductive if it were thought to
be temporary. Temporary policies tend to produce intertemporal arbitrage
opportunities, which then trigger hoarding, capital flight, and other
speculative activities. Such speculation can be quite destabilizing.
While credibility is an issue for all countries, it is especially important
in developing countries. Almost by definition, many of these countries
have a history of economic and political instability, which has produced
a legacy of mistrust and suspicion. This makes the task of establishing
and maintaining credibility especially difficult for policymakers in developing
countries.
The importance of credibility to the success of inflation targeting in
developing countries has been discussed by Masson, Savastano, and Sharma
(1997), and analyzed more formally by Kumhof (2000). Masson, et al., point
out that central banks in many developing countries are in a position
of "fiscal dominance." Due to weak tax systems and underdeveloped
capital markets, governments in developing countries must often finance
a significant share of their expenditures by printing money. When this
is the case, the exigencies of the budget take precedence over the control
of inflation. Clearly, in a situation like this, inflation targeting is
not a credible policy.
Kumhof (2000) studies what happens when the public (correctly) believes
that inflation targeting will not last. He shows that when it is costly
to change prices, so that prices are "sticky," inflation targeting
will produce an excessively tight monetary policy. Because the policy
is viewed as temporary, domestic firms will not incur the costs of adjusting
their prices in line with the newly stated inflation target. As a result,
most of the burden of hitting the target will fall on import prices, via
the exchange rate channel. As noted earlier, this tends to produce a destabilizing
whiplash effect on the rest of the economy. In fact, Kumhof comes to the
provocative conclusion that a noncredible inflation-targeting regime is
even worse than a noncredible exchange rate peg!
Financial fragility
One of the lessons to emerge from recent currency crises is the role
of "balance sheet effects" in determining the outcome of a devaluation.
Devaluations that would otherwise be expansionary (because they enhance
competitiveness) can quickly become contractionary if domestic firms have
large unhedged foreign currency exposures. When this is the case, a devaluation
erodes the net worth of firms. If financial market imperfections make
internal finance cheaper than external finance, this decline in net worth
forces firms to cut back investment, which then further depresses asset
prices and net worth.
Like credibility, this so-called "financial accelerator" is
of concern to all countries, since it is widely regarded as a source
of business cycle amplification. However, it is of special concern to
developing countries. Underdeveloped financial markets make internal finance
constraints especially important in developing countries. At the same
time, relatively weak financial supervision and regulation often provide
incentives to incur foreign currency liabilities, since foreign interest
rates are typically lower. This combination makes developing countries
vulnerable to exchange rate changes.
Devereux and Lane (2000) study inflation targeting when domestic firms
borrow abroad while facing collateral constraints. Not surprisingly, they
find that balance sheet effects greatly amplify the economy's response
to external shocks. They also conclude that balance sheet effects diminish
the attractiveness of inflation targeting relative to exchange rate targeting,
although inflation targeting does remain preferable on net. These findings
suggest that countries contemplating a move toward inflation targeting
should also take steps to improve the functioning of their financial markets.
Conclusion
In principle, there is no reason why inflation targeting cannot work
in developing countries. However, inflation targeting is not a "one-size-fits-all"
policy. Inflation targeting rules that work well in the U.S. are not likely
to work well in Thailand or Korea. Developing countries need to be aware
of three key differences. First, the exchange rate is an additional monetary
policy transmission channel, which tends to operate rather quickly; and
unless both the instruments and targets of monetary policy are modified,
inflation targeting in developing countries could produce destabilizing
output fluctuations. Second, inflation targeting won't be credible, and
hence desirable, if the government must rely on the inflation tax to finance
its expenditures. This means that for many countries inflation targeting
needs to be accompanied by fiscal reforms. Finally, inflation targeting
will likely involve larger swings in the exchange rate than many of these
countries are used to. If inadequate regulation and other financial market
imperfections create incentives to incur unhedged foreign currency liabilities,
these exchange rate changes could be destabilizing. For many developing
countries, this means that inflation targeting must also be accompanied
by financial market reforms.
Kenneth Kasa
Senior Economist
References
Ball, Laurence. 1999. "Policy Rules for Open Economies." In
Monetary Policy Rules, ed. John Taylor. Chicago: University of
Chicago Press.
Devereux, Michael B., and Philip R. Lane. 2000. "Exchange Rates
and Monetary Policy in Emerging Market Economies." Working Paper.
Department of Economics, University of British Columbia.
Kumhof, Michael. 2000. "Inflation Targeting under Imperfect Credibility."
Working Paper. Department of Economics, Stanford University.
Masson, Paul, Miguel Savastano, and Sunil Sharma. 1997. "The Scope
for Inflation Targeting in Developing Countries." Working Paper No.
130. International Monetary Fund.
Svensson, Lars E. O. 2000. "Open-Economy Inflation Targeting."
Journal of International Economics 50, pp. 155-183.
Opinions expressed in this newsletter do not necessarily reflect
the views of the management of the Federal Reserve Bank of San Francisco
or of the Board of Governors of the Federal Reserve System. Editorial
comments may be addressed to the editor or to the author. Mail comments
to:
Research Department
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120
|
