V Policy Response
- Taimur Baig, Jörg Decressin, Tarhan Feyzioglu, Manmohan Kumar, and Chris Faulkner-MacDonagh
- Published Date:
- June 2003
Given the costs of deflation, it is desirable to prevent it from taking root, rather than deal with it once it has become entrenched. This is a particular challenge for policy in a low inflation environment. A number of conceptual and practical issues related to both prevention and cure are examined below. Strictly speaking, monetary policy should, by itself, be able to stave off deflationary expectations. However, the central bank may underestimate the risks of deflation, or be unable—because of political economy reasons—to undertake the necessary steps. When monetary response is insufficient or inadequate, fiscal policy and structural measures can play an important role.
Policymakers need to be attentive to four channels through which deflationary forces can be propelled and through which policy can act:
Exchange-rate channel. An open trading system can serve as a buffer, while fixed exchange rate systems have, in the past, transmitted deflationary shocks.
Asset price or portfolio rebalancing channel. Deflation affects the relative trade-off between assets, and cash generates a risk-free rate of return (equal to the rate of deflation), discouraging risk taking.
Expectations channel. Entrenched expectations help determine nominal wage demands and ex ante real interest rates. Once deflationary expectations set in, they help to drive up real wage costs and real interest rates.
Credit channel. When deflation reduces the value of collateral, banks find it difficult to discriminate credit risks, and the external finance premium rises. Distressed banks may curtail lending, further driving down output and prices.
A vigorous policy response may be required if one or more of these channels are not available. For example, if deflation is setting in, banks are restricting credit, and labor markets are inflexible with sticky wages, the central bank may need to be particularly aggressive in easing policy and changing expectations—so as to compensate for its reduced ability to act through the other channels.
Policy Before the Onset of Deflation
Preventing deflation in a low inflation environment requires preemptive and possibly aggressive action. Past episodes suggest that sustained deflation can be unanticipated, even as inflation and nominal interest rates fall close to zero. The challenge is evident from the recent experience of Japan. In the mid-1990s, household and business surveys—along with government and corporate bond yields—showed that markets expected moderate inflation, right up to the onset of deflation. Furthermore, as Ahearne and others (2002) show, ex ante, the monetary policy stance was appropriate or even loose (using Taylor-type monetary policy rule). But ex post, policy proved to be too tight, reflecting lower inflation than forecast. Given such circumstances, with baseline forecasts still projecting inflation, policymakers could be hard-pressed to justify implementing measures against deflation.
Monetary policy therefore needs to set targets to provide a buffer against the risks of deflation. Under inflation-targeting regimes, policymakers should not set an inflation target that is too low, but instead provide a “buffer zone” to obtain insurance.32 As
Figure 6 illustrates, the likelihood of hitting the zero bound on interest rates declines significantly as the inflation target increases (see Box 5). With flexible monetary policy regimes, clear communication with the markets is important, as is a willingness to act quickly, even though easing may appear unnecessary ex ante. Model simulations suggest that in Japan lowering short-term interest rates by another 200 basis points at any point between 1991 and early 1995 would have helped avoid deflation (Ahearne and others, 2002). A proactive stance is particularly necessary if financial markets and institutions may be at risk of coming under stress, say following the bursting of an asset price bubble. (Aggressive easing by the U.S. Federal Reserve in early 2001 can be seen in this light.)
Figure 6.Inflation Target and Zero Bound
Source: Hunt and Laxton (forthcoming).
1 Probability of zero bound on interest rates being hit with a given inflation target. These probabilities are derived from Multi-mod simulations for a large open economy.
Box 5.Inflation Targeting and Deflation Risks
The benefits of low inflation are well known. But given risks of deflation, is there a case for targeting a somewhat higher rate of inflation, say 2 to 3 percent, rather than a lower or zero rate? There are three relevant considerations:
First, a small positive rate would reduce the likelihood that countercyclical monetary policy becomes constrained by the zero floor on nominal interest rates. A number of studies suggest that the probability of the zero interest rate floor becoming binding is small for inflation targets down to around 2 percent, but then increases for lower targets, and especially below 1 percent (see Hunt and Laxton, forthcoming; Kieler 2003; Reifschneider and Williams, 2000; and Yates, 2002). It has been argued that these findings may underestimate the likelihood of the zero constraint becoming binding given the non-linearities in the system.
Second, the potential dispersion in trend inflation across members of a currency union may be large. Inflation in some countries may have to be compensated for with deflation in others, if the inflation target for the union is low. For the Balassa-Samuelson effects, a variety of studies suggest that the steady state inflation in Germany could be significantly below the euro area average, and perhaps below 1 percent if the ECB aims for 1.5 percent for the aggregate.1
Third, inflation may facilitate relative price and wage adjustments in an economy where agents are averse to nominal wage or price cuts.2 Downward rigidities may stem from elements of money illusion, and employer concerns about the impact of wage cuts on worker morale. Akerlof, Dickens, and Perry (1996) argue that steady-state unemployment in the United States would be considerably higher at rates of inflation below 3 percent, and especially below 2 percent. However, the coexistence of low inflation and falling NAIRUs in a number of industrialized economies in the 1990s—including the United States and the United Kingdom, and several euro area countries—suggests that the reduction in inflation to roughly 2 percent has not had any detrimental impact on the smooth functioning of labor markets.1 See for example, Canzoneri and others (2002), and De Grauwe (2000). The Balassa-Samuelson model has not received unequivocal backing in the empirical literature. For example, Rogoff (1996) concludes that across industrialized countries, there is long-run convergence to PPP, the Balassa-Samuelson effect notwithstanding. Kieler (2003) shows how average inflation rates might differ across the euro area countries over the long-run depending on the how fast remaining differences in price levels are reduced. The results suggest that trend inflation rates could vary from around 1–1¼ percent in Germany and France to 2–3¼ percent in Portugal, Greece, and Spain, assuming an inflation target of 1.5 percent.2 See Keynes (1936); and Tobin and Dolde (1971).
Fiscal policy can play an important role in supporting incomes, relieving pressure on firms’ and households’ balance sheets, and underpinning confidence. In an environment of deflationary expectations taking hold, and with the economy in a liquidity trap, stimulative policy—beyond the automatic stabilizers—may be necessary. Measures to boost return on capital investment that would have efficiency gains while signaling authorities’ commitment to prevent a generalized decline in prices have an added attraction (see Rogoff, 1998). In the case of Japan, Ahearne and others suggest that a moderate amount of additional fiscal loosening during the first half of 1990s could have assisted economic activity and lowered the risk of deflation.
An open trading system offers a measure of protection from deflation, although this protection is much greater under flexible rather than fixed exchange rates. Countries with deficient domestic demand face incipient downward pressure on tradable goods prices and interest rates. With flexible exchange rates, and a liberal trade and capital account regime, the downward pressure on interest rates would lead to currency depreciation and aid exports. Under such circumstances, expectations of deflation are less likely to take hold. But policy still needs to be alert to the risk of deflation spreading from imported deflationary impulses in particular sectors, especially if the sectors are large with relatively inflexible structures. Where exchange rates are fixed, including in a currency union, the adjustment to adverse demand shocks would be considerably more protracted, as it would operate through domestic prices rather than the exchange rate.33 This makes it more likely that expectations of deflation develop, with further deleterious consequences for demand.
Lower barriers to trade can help in spreading the adjustment to demand shocks. Instead of a large adjustment by a single country, a small adjustment globally could help to redress country-specific imbalances.34 However, globalization may also contribute to raising the symmetry of economic shocks across countries. The recent equity price bubble is an example of how increased financial linkages might contribute to spreading “irrational exuberance” and a subsequent correction (see Rogoff and others, forthcoming, on the costs and benefits of financial globalization).
With flexible exchange rates, the threat of an internationally transmitted, global deflationary shock is small. This is because countries at risk of deflation could expand their money supply and create inflationary expectations. If the policymakers’ anti-inflation credentials are high, there may be little risk that these expectations will be difficult to reverse. (This also underlines the case for the transparency in long-term objectives that provide broad guidelines for inflation over the medium-term—see Rogoff, 2003.) In the case of a global demand shock, if every country expanded its money supply in response to a deflationary impulse, exchange parities would change to the extent that the impact of the shock and the monetary response differed across countries. But in general the stock of money would be higher, reducing the likelihood of deflationary expectations from gaining ground.
Nevertheless, the global effects of a depreciation in a large country can be significant. A depreciation in the U.S. dollar, for example, would lead to higher U.S. exports, offsetting foreign production, and put downward pressure on foreign prices. It would also have income and balance sheet effects, and if the dollar decline were severe enough, foreign balance sheets could come under significant pressure, aggravating deflationary pressures there with effects that can rebound on the United States.35
Policy Once Deflation Has Set In
Policymakers face a difficult challenge under deflation, because the traditional relationships between policy instruments and economic variables, as well as the transmission mechanisms, may be impaired or altered. For example, when deflation has set in, even if nominal rates are low or zero, monetary policy may not be loose enough to spur activity. This may be especially constraining given that in a liquidity trap, the economy’s equilibrium real interest rate may well be negative.
The problems may be compounded if the banking system is undergoing difficulties, say following a sharp slowdown in activity or the bursting of an asset price bubble. An increase in banks’ bad debts is likely to reinforce their unwillingness to take risks, curtailing credit. In such a situation, efforts to boost base money may have limited effect if the additional liquidity largely ends up as banks’ excess reserves.
In these circumstances, historical experiences suggest that policymakers may need to consider “unorthodox” policies. Monetary policy can generate inflation, independent of the structural impediments in the economy (Rogoff, 1998 and 2002). As Bernanke (1995) observes: “. . . if the price level were truly independent of money issuance, then the monetary authorities could use the money they create to acquire indefinite quantities of goods and assets. This is manifestly impossible in equilibrium.”
For instance, the monetary authority, if credible, could lower interest rates along the yield curve by promising to hold short-term rates at zero for a specified period (along with an explicit inflation target) or by purchasing longer-term bonds. The bond purchase operation could be underpinned by announcing target yields and committing to purchase securities consistent with that target, thereby lowering rates along the entire term structure.36 In addition, the range of assets eligible as collateral for bank borrowing from the central bank could be widened, reducing term and liquidity premiums and lowering the cost of capital. The central bank could also extend its range of purchases to equities (exchange traded funds), real estate (real estate investment trusts), or foreign assets (foreign currency or bonds).
Once deflationary expectations have taken hold, policies will be successful only to the extent that they manage to break expectations. Monetary policy helps to stimulate the economy because a lower policy rate helps to flatten the yield curve, lower long-term rates, and encourage contemporaneous consumption and investment. This role does not change under deflation; the central bank must choose steps that have the largest effect on expectations. For example, if the credit channel were not severely impaired, the central bank may be able to lower long-term interest rates by purchasing government bonds along the yield curve, which would help encourage bank lending. However, if the credit channel were impaired or the balance sheets of firms and households were distressed, the central bank may need to take more drastic steps, such as purchasing private assets.
The “unorthodox” policy prescriptions can pose operational challenges because the broader impact of these policies may be uncertain. Even during inflationary times, the lag structure and impact of changes in the monetary stance may be unclear. With unorthodox policies, there may be little guidance to the appropriate extent of easing. There is also a risk that the central bank’s balance sheet could incur sizable capital losses if the value of the acquired assets falls. Nevertheless, these uncertainties and risks are small compared to the costs associated with entrenched and deepening deflationary expectations. The monetary policy response should also not be held hostage to structural reforms: one of the key lessons of the events that led to the Great Depression in the United States was the misplaced and ultimately catastrophic emphasis on reforms at the expense of monetary easing.37
Fiscal policy financed by monetary expansion can play an important role in this situation. Indeed, for an economy in a liquidity trap, the use of unorthodox monetary operations noted above could entail a significant fiscal element.38 Stimulatory fiscal policy will boost activity, help address problems in the credit and asset channels, and result in an increase in prices of goods and assets. For liquidity-constrained consumers, traditional fiscal multipliers may actually be higher under deflation than inflation because the credit channel is weakened. If deflation is due to an especially large demand shock, beyond the automatic stabilizers, discretionary fiscal policy can play a particularly useful role in supporting incomes and spending. A well-timed tax cut will increase disposable incomes, helping to encourage consumption. Higher government spending could also help boost production and relieve unemployment. When banks are distressed and the credit channel is impaired, the government may need to inject capital into the banking system, and under deflation, it may need to act faster than it would otherwise.
However, fiscal policy needs to be tailored to credibly boost aggregate demand. The stimulus should not be wasted on projects for political economy reasons that benefit too few to have an effect at the aggregate level. Furthermore, to limit Ricardian effects, spending programs and tax relief should target liquidity-constrained consumers and, as noted above, projects that boost the return to private investment. It is also important to ensure that fiscal policies do not give rise to a permanent increase in the level of expenditures. This suggests that measures to reallocate demand inter-temporally, for instance, through temporary tax credits for investment or reductions in consumption taxes that automatically lapse after a period of time, would be most appropriate (see Feldstein, 2001 and 2002).
Structural reforms can also play an important role in combating deflation. By increasing flexibility in factor markets, they may facilitate reallocation of resources. They may also allow for a more rapid disposal of bad assets and promote enhanced intermediation in the economy, improving the effectiveness of the monetary transmission mechanism. In addition, measures aimed at promoting deeper and more liquid capital markets can help reduce frictions generated by deflationary shocks; with the banking sector under stress, corporate financing will be able to take place more readily if there is access to alternative markets. By helping raise the long-term return on capital and enhancing the economy’s growth potential, structural reforms can thus be a key component of an anti-deflation strategy. Even in the short run, a credible and well-articulated package of reforms can underpin confidence and, by raising expectations of future growth, mute deflationary pressures. Structural reforms should not, however, be regarded as a prerequisite for the monetary and fiscal response.