Inflation targeting and the need for a new central banking framework

ABSTRACT

This article critically analyzes inflation targeting (IT) both theoretically and empirically. IT came into prominence in the 1990s and 1 central bank after another adopted this regime in the 1990s and 2000s. Proponents of IT mainly argued that IT regime was successful on the grounds that it resulted in lower inflation rates and hence better economic performances. However, inflation rates in the world were in a downward trend from the 1980s well into the 2000s, and both IT and non-IT regimes managed to decrease their inflation rates. In addition, focusing too much on price stability through IT paved the way for permanently higher than necessary interest rates and disinflationary “tight” monetary policy periods when inflation rate was above an arbitrarily targeted level. Tight monetary policy can and do affect the real economy negatively and overemphasizing price stability may hurt the economy in terms of lower potential output, decreasing investment and more unequal income distribution. Post Keynesians offer valuable alternatives within the framework of parking-it approach to the existing monetary policy paradigm. Our main conclusion is that central banks should set the policy interest rate as low as possible and keep it there, in line with Keynesian “cheap money” policy.

KEYWORDS:

• Kansas city rule
• inflation targeting
• pasinetti rule
• Smithin rule
• threshold inflation

JEL CLASSIFICATION:

• E31
• E52
• E58

Notes

¹As a case in point, the European Central Bank (ECB) raised policy interest rate twice (from 1 to 1.5%) in April and July 2011 in order to “battle” inflation at a time when Euro area was just beginning to recover from the global financial crisis, which, actually, proved to be very short-lived: Euro area plunged into a deep recession after a short while.

²Implied inflation theory of IT regime is rather reminiscent of Knut Wicksell’s inflation theory, which was formulated as an extension of the quantity theory of money a century earlier and started to attract much attention with the advent of the IT regimes. Wicksell (1898) defined two kinds of interest in his formulation: market rate of interest and natural rate of interest, which is the rate at which supply of and demand for money (capital) and capital goods are equal.

In Wicksell’s formulation, money is perfectly neutral and regarded as just a veil as in the classical quantity theory of money and interest rate is the means that equals demand for and supply of capital goods in the economy in the final analysis. On the other hand, Wicksell assumed that although the natural rate of interest is a built-in factor embedded in the economy, market rate of interest is an exogenous variable, which can be and is set arbitrarily. Moreover, supply of money is regarded entirely endogenous in Wicksell’s formulation. That is, although natural rate and money supply is determined by the market forces, market rate is set by commercial banks and if there is a mismatch between market rate and natural rate of interest, then things go awry in the economy. For example, if market rate rises above natural rate of interest, then supply of savings exceeds demand for capital and as a result of that both credit volume and money supply in the economy dwindle, and because of the falling demand in the face of higher market rate of interest, commodity prices fall. Hence lower inflation. On the other hand, if market rate of interest decreases below of natural rate, then things happen the other way around: Demand for capital exceeds supply of savings and consequently both credit volume and money supply expand, and due to the rising demand emanating from lower market rate of interest, commodity prices rise. Hence higher inflation.

In the bottom line, in Wicksell’s inflation theory, mismatch between natural rate and market rate of interest paves the way for lower or higher demand in the economy and this results in the higher or lower inflation in the economy. The implied monetary policy conclusion from this theory is that if a central bank wants to decrease inflation, then it should increase policy interest rate. And, indeed, this is the very policy advice of the new consensus macroeconomics, and Wicksell’s inflation theory (which can be regarded as an interest theory as well) is very much in spirit with Taylor’s rule, which is the backbone of the IT regime. Former Bank of Canada economist Clinton (2006) stated very clearly how this is the case for numerous central banks around the world in an article revealingly named “Wicksell at the Bank of Canada”: “[T]he Bank of Canada adopted a neo-Wicksellian approach to monetary policy. Since the bank led the way for many other central banks, and since the approach has had wide success, the story has broad, international relevance.” Moreover, then Federal Reserve Governor (and now President) Yellen (1995, p. 43–44) stated at the Federal Open Market Committee (FOMC) meeting in 1995 that following Taylor rule “is what sensible central banks do”.

³The belief of nonneutrality of money in the short run is also held by new consensus macroeconomics, though for different reasons. However, this consensus fails to take into account path dependence or hysteresis and insists that short run nonneutrality of money does not translate into long run non-neutrality. Path dependence, which indicates that the current state of the economy has a certain level of influence on the future states of the economy, is the crucial link between nonneutrality of money in the short and long run as Setterfield (1993) aptly explained, “[w]ith hysteresis, any long-run or final configuration that is reached will depend on the path taken toward it; the evolution of a system may be best described, not by equilibria, but by “contemporaneous” values of variables expressed in terms of their own past history” (p. 361). In particular, hysteresis effect could have a considerable influence on the real economy such that being unemployed for too long can deprive workers of their skills and/or their willingness to work, hence resulting in permanently higher unemployment rate and lower economic activity. That is, as Mota and Vasconcelos (2012) expressed, with hysteresis “the equilibrium state of employment, instead of being unique, becomes path dependent, that is, determined by the history of previous adjustments” (p. 93). In addition, a prolonged recession could result in many unrealized investment opportunities, which may have serious repercussions in terms of never surfaced and permanently lost future investment opportunities. Thereby, an economy could easily find itself in a rather worse economic path than otherwise after a prolonged recession. Beside these, actually, neutrality of money in the economy is quite counterintuitive: one cannot speak of an economy without money and financial system as one cannot understand human body without considering blood cells and circulatory system.

⁴Actually, there are two possible explanations as to the downward trend of inflation rates on a global scale theoretically. First, one can argue that non-IT regimes could have pursued an “IT-like” discretionary monetary policy that overemphasize price stability at the expense of risking the real economy, thereby achieving lower inflation rates, just like IT regimes did. Second, the fluctuations in the inflation rate across the globe could have mostly been the product of global cost-push factors in an increasingly intertwined world. When historical trajectory of average inflation rates and post Keynesian approach to the root causes of inflation considered, it becomes clear that the second explanation is much more satisfactory than the first one. One can rightly argue that global cost-push factors such as oil shocks in the 1970s and 2000s or dramatic rises in food prices due to supply shocks in the 2005–2008 period or steep increases in commodity prices in the 2009–2011 period were the main causes of rising inflation rates on a global scale. The subsequent decline of inflation rates across the globe can easily be ascribed to the very nature of cost-push factors: They come and go. When they come, inflation rates rise. When they go, inflation rates fall gradually to the level of where they were before given that there is no or low level of indexation. So trying to decrease inflation rate that stems from cost-push factors by suppressing aggregate demand can do little good, if any. To make matters worse, such a tight monetary policy can have serious detrimental effects on the real economy, which, ironically, can also feed cost-push inflation. Moreover, the first possible explanation complicates things more than it solves: why does a central bank not adopt IT regime and still pursue an IT-like monetary policy in an era where IT is rather “popular” and could be used as a perfect pretext for tight monetary policy? Because of that, we can safely assume that the stance of non-IT central banks in terms of real economy differs significantly than that of IT central banks on average.

⁵It is well known that empirical literature regarding inflation rate-economic growth nexus in the past mostly assumed that there is a linear between these variables. As expected, the resulting findings in this literature produced mixed results. Although some studies found that there is a negative relationship (see Gylfason and Herbertsson, 2001; Grier and Grier, 2006; Wilson, 2006), others found a neutral and even a positive relationship (see Bruno and Easterly, 1998; Black, Dowd, and Keith, 2001; Mallik and Chowdhury, 2001). However, the negative relationship between these variables could only be observed in countries with very high inflation rates as Bruno and Easterly (1998) put it rather succinctly:

Recent articles in the new growth literature find that growth and inflation are negatively related, a finding that is usually thought to reflect a long-run relationship. But the inflation-growth correlation is only present with high frequency data and with extreme inflation observations; there is no cross-sectional correlation between long-run averages of growth and inflation. (Italics added for emphasis)

And Ericsson, Irons, and Tryon (2001) notably concluded that,

The negative correlation between inflation and output growth obtained by cross-country regression is not robust to changes in model specification. The selection of countries in the sample, the level of aggregation over time, and the choice of dynamic specification all affect the results obtained. … If Africa and Latin America are dropped from the sample, the coefficient on inflation in the growth regression becomes positive and statistically insignificant. For the OECD countries by themselves, no economically important, statistically detectable, long-run relationship appears to exist between output growth and inflation. (Italics added for emphasis)

⁶At this point, one can argue that an IT regime with a higher inflation target could be the right monetary policy. However, IT regime is much more than a mere inflation targeting. Above all, IT paradigm rests on Wicksell’s natural interest rate conceptualization. So, IT regime is seriously flawed even with a higher inflation target. In addition, IT paradigm rests on the belief that inflation is bad, among others. So, there is no room for a growth- and employment-friendly inflation target, if such a conceptualization can be made, in the IT paradigm.

⁷Particularly, the Symposium on The State of Post Keynesian Interest Rate Policy: Where Are We and Where Are We Going?, which was held in 2007, is of paramount importance in this regard.

Additional information

Notes on contributors

Mevlut Tatliyer

Mevlut Tatliyer is with the Department of Economics and Finance at Istanbul Medipol University, Istanbul, Turkey.