Aggregate supply-driven deflation and its implications for macroeconomic stability.

• Author: Beckworth, David

Deflation is generally considered to be inconsistent with macroeconomic stability. Any sustained decline in the price level is widely believed to be associated with weak to negative economic growth, a lower bound of zero on the policy interest rate, and an increase in financial disintermediation. However, a number of recent studies examining both historical, cross-country experience with deflation and more recent developments find that these concerns are not necessarily associated with deflation (Selgin 1997, 1999; Cleveland Federal Reserve 1998; Stern 2003; Bordo and Redish 2004; Bordo, Lane, and Redish 2004; Bordo and Filardo 2005; Borio and Filardo 2004; Farrell 2004; King 2004; The Economist 2004, 2005; White 2006). They show that the deflationary experiences that shape the modern economic psyche, the Great Depression in the 1930s and Japan in the 1990s, are not truly representative of all deflation outcomes. These studies contend that a broader, historical perspective reveals a more nuanced view of deflation, one that requires taking seriously the possibility of both a malign deflation, a deflation originating from a collapse in aggregate demand, and a benign deflation, a deflation originating from an increase in aggregate supply

Some of these studies further argue that not only is aggregate supply-driven deflation real, but it may be optimal. They contend that by avoiding all deflations instead of avoiding only the harmful form, monetary policy may in fact create economic imbalances that eventually will have to be corrected. Better to embrace aggregate supply-driven deflation today than to deal with the potentially painful correction of economic imbalances in the future (Selgin 1997, King 2004, The Economist 2005, White 2006).

Taking seriously the idea that deflation can be benign, however, is a tough sell in a world where inflation is considered standard and deflation is generally assumed to be harmful. Moreover, embracing this idea is embracing the unknown since aggregate supply-driven deflation has been absent from modern economies. As a result, most observers are simply unfamiliar with it. Aggregate supply-driven deflation, however, is an important issue going forward as the continued opening of China and India and the ongoing rapid technological gains should continue to buffet the world economy with positive aggregate supply shocks.

This article reviews the conventional view of deflation, which assumes that deflation is always harmful, and contrasts it with a more nuanced view of deflation, which makes a distinction between deflationary pressures originating from the aggregate demand and the aggregate supply sides of the economy. We then examine what this more nuanced view of deflation means for macroeconomic stability and how it sheds light on the recent turbulence in the global economy. Finally, the article concludes by considering the implications for policymakers going forward.

The Conventional View of Deflation

Deflation returned as a pressing issue as Japan struggled with its weak economy and falling price level and inflation in many of world's advanced economies declined significantly Between 1995 and 2004, Japan's economy grew an anemic 1.15 percent per year while its price level declined by 0.04 percent per year. Advanced economies saw their inflation rates go from an annual average of 5.83 percent for 1980-94 to 1.97 percent for 1995-2004 (IMF 2008a).

Most of the renewed interest in deflation has been premised on a belief that deflation is harmful to economic activity (Stern 2003). This understanding of deflation is usually attributed to the Great Depression experience of the 1930s, when a severe economic collapse and a dramatic decline in prices led to a "near consensus [that] believed ... deflation was deeply dangerous and to be avoided at all costs" (DeLong 1999: 231). The belief that deflation is economically harmful is not only evident in the deflation literature, but is also borne out in practice: no central bank targets deflation while most central banks either implicitly or explicitly target some rate of inflation. (1)

The conventional view of deflation is often justified by noting several important channels through which deflation can adversely affect an economy. First, given relatively rigid nominal input prices, particularly wages, an unexpected fall in the price level originating from a collapse in aggregate demand will increase real input prices and lower firms' profit margins. As a result, firms will cut back production and employment levels. The more rigid the nominal input prices, the larger will be the decline in output and employment (Akerlof, Dickens, and Perry 1996; Greenspan 2003). (2) Second, unanticipated deflation will pull down nominal interest rates and unexpectedly increase real interest rates, further constricting an already weak economy Moreover, the deflation might drag the nominal federal funds rate down to its lower bound of zero and thereby eliminate the possibility of additional monetary stimulus through the policy rate (DeLong 1999, Svensson 2003). (3) Third, the financial system will become beset with unplanned increases in real debt burdens and unplanned decreases in collateral values as the deflation-plagued economy deteriorates. Consequently, delinquencies and defaults will increase and the balance sheets of financial institutions will weaken causing financial intermediation to suffer (Bernanke 2002, IMF 2003, Furhrer and Tootell 2003). Collectively, these events may reinforce each other in a "deflationary spiral" where expectations of more deflation and additional economic weakness lead to a further fall in aggregate demand and a prolonged economic slump (Fuhrer and Sniderman 2000). Thus, unanticipated deflation--the initial negative demand shock--and even anticipated deflation--the deflationary spiral--can have adverse economic effects. Accordingly, the conventional view is that deflation of any form should be feared and "avoided at all costs" (DeLong 1999: 231).

A More Nuanced View of Deflation

But should deflation always be feared? The aggregate demand-aggregate supply (AD-AS) model indicates deflation can occur for two reasons: a decrease in aggregate demand or an increase in aggregate supply. The first type of deflation is a consequence of a collapse in nominal spending that may, in the presence of nominal rigidities, drive actual output below its natural rate level. This malign form of deflation is consistent with the conventional view of deflation described above and is what most observers invoke when considering the issue of deflation. For example, Federal Reserve Chairman Ben Bernanke (2002) explains the "sources of deflation are not a mystery. Deflation is in almost all cases a side effect of a collapse in aggregate demand--a drop in spending so severe that producers must cut prices on an ongoing basis in order to find buyers."

The second type of deflation, however, is the result of positive aggregate supply shocks that are not accommodated by an easing of monetary policy. Such aggregate supply shocks are the result of positive innovations to productivity or factor input growth that lower per unit costs of production and, in conjunction with competitive market forces, create downward pressure on output prices. Unlike a collapse in aggregate demand, positive aggregate supply shocks that are not monetarily accommodated generate a benign form of deflation where nominal spending is stable, because the decline in the price level is accompanied by an increase in the actual and "natural" level of output.

Figure 1 illustrates these two distinct forms of deflation using the standard textbook aggregate demand-aggregate supply (AD-AS) framework. Here, the malign form of deflation is seen as the result of a negative aggregate demand shock shifting aggregate demand from [AD.sup.1] to [AD.sup.2] and, owing to the nominal rigidities as indicated by the upward sloping short-run aggregate supply (SAS) curve, decreasing output temporarily to [Y.sup.2] from its natural rate level, [Y.sup.1]. Total nominal spending (P x Y) collapses and the price level drops from [P.sup.1] down to [P.sup.2]. The benign form of deflation, on the other hand, occurs as the consequence of positive aggregate supply shocks shifting the long-run aggregate supply (LAS) curve--or the natural rate level--from [LAS.sup.1] to [LAS.sup.2] without monetary policy increasing aggregate demand to stabilize the price level. Unlike with malign deflation, (P x Y) now remains stable even as the price level falls from [P.sup.1] to [P.sup.2].

The benign form of deflation, however, can take on slightly different characteristics depending on whether the advances in aggregate supply originated from shocks to productivity or shocks to factor inputs. Consider first productivity-induced deflation. Profit margins are likely to remain stable since the decline in output prices caused by positive shocks to productivity are matched by a decline in per unit costs of production. Even with relatively rigid nominal input prices, such as sticky wages, productivity gains still lower per unit costs of production and allow for a...