Inflation Targeting: Insights from Behavioral Economics

Abstract

In many emerging and transition economies, inflation targeting (IT) has been a successful framework for monetary policy in the areas of improved central bank transparency, higher real GDP growth, and better collection of tax revenues, but has been controversial in other respects. This article reviews the performance of IT in transition economies by focusing on its day-to-day policy conduct from the perspective of behavioral economics. The experience of the Czech National Bank (CNB) is used as a representative example of how IT is implemented in a transition economy. Our review of IT as practiced in the Czech Republic suggests that the CNB’s theoretical assumptions may not be fully consistent with the economic behavior of economic agents as described by behavioral economics and therefore lead to biased forecasts. In addition, we find that CNB forecasts are not always turned into policy on the basis of a well-behaved monetary policy reaction function. We discuss the policy implications of our findings.

Keywords:

• behavioral economics
• forecast bias
• inflation targeting
• monetary policy
• perceived inflation

JEL Classification:

• E3
• E5
• F4
• O2
• P22

Notes

¹ Mishkin (2008) evaluates the performance of many emerging economies that have adopted IT and concludes, “Inflation targeting has been a highly successful monetary policy strategy for many emerging market countries” (p. 13). Later studies also indicate positive findings. For example, Akyurek, Kutan, and Yilmazkuday (2011) find that the IT regime has been effective in Turkey. Krušec (2011) reports that monetary transmission mechanisms improved in transition economies that adopted an IT regime. Lucotte (2012) finds that IT adoption in emerging economies improved public revenue collection. Mendonça and Galveas (2013) report that the adoption of an IT regime in Brazil in June 1999 improved central bank transparency. Ayres, Belasen, and Kutan (2014) find that IT adoption increased real GDP in many developing countries in Europe, Latin America, and the Middle East. Dancourt (2015) provides evidence that implementation of an IT regime and accumulation of foreign-exchange reserves in the early 2000s have both allowed the Peruvian central bank to improve and maintain macroeconomic stability.

² Vickers (1999) provides a good discussion of issues regarding forecasts, monetary board objectives, and the board’s use of information not included in the forecast from the standpoint of U.K. practice, and his analysis has applicability well beyond the case of the U.K.

³ For nontechnical surveys of behavioral economics concepts as they apply to public policy, see Reeson and Dunstall (2009) or Dawney and Shah (2005). Ranyard et al. (2008) provide an excellent review of the literature on perceptions and expectations of price changes and inflation from a behavioral economics perspective.

⁴ Barberis, Huang, and Santos (1999) show how this concept applies to asset pricing, and Quattrone and Tversky (1988) demonstrate the implications for political choices about, inter alia, inflation.

⁵ See Tomarken et al. (1989).

⁶ Thus, Vickers (1999) describes the decisions of the U.K.’s Monetary Policy Committee as being entirely in the public interest and beyond the sway of material incentives. He writes, “Quite apart from the obligation to fulfil our statutory duty, we have the strongest professional and reputational incentives, which in my opinion are incapable of being enhanced by financial incentives to get as close as we can to the inflation target” (p. 6).

⁷ Van Pelt (2008) reports: “In one study of annual reports by public companies, researchers found that executives tend to attribute favorable outcomes to factors under their control—such as corporate strategy or innovation initiatives; while unfavorable outcomes are much more likely to be attributed to uncontrollable external factors, such as inflation or weather. Their self-confidence leads them to believe that they’ll be able to avoid or easily surmount potential problems in executing a project. And their misattribution bias leads them to take credit for lucky breaks.” The experience of one of the authors of this article in advising governments on economic policy suggests that the same observation could be made of politicians.

⁸ Many recent studies forecasting inflation in emerging economies use either versions of the Phillips curve or alternative time-series models. For example, Kim (2008) uses a structural cointegration approach to model inflation in Poland. Zhang (2013) and Andiç, Küçük, and Öğünç (2015) employ a hybrid New Keynesian Phillips curve to forecast inflation in Turkey and China, respectively. Kaya and Yazgan (2014) follow a long-run structural-modeling approach and employ a vector error correction model for forecasting inflation in Turkey. Camacho and Perez-Quiros (2014) use a regime-switching model for Latin American economies. He and Fan (2015) use a Bayesian vector autoregressive model (VAR) model for modeling inflation in China. Inflation forecasts based on such a variety of models certainly produce different forecasts of inflation and hence provide an important challenge for central banks to choose the best model for inflation. Additionally, structural breaks in inflation forecasts are common when central banks switch to a new forecasting model. This also has been a problem for CNB forecasts, and we discuss this issue further in this section.

⁹ For an early treatment of monetary policy in the Czech Republic and other transition economies, see Kutan and Brada (2000).

¹⁰ See Antal, Hlaváček, and Horváth (2008). They conclude: “Inflation prediction error has decreased over time. We further point out that the GDP growth and interest rates were respectively above or respectively below the forecast most of the time, even in the situation of systematic undershooting of the target. Thus, undershooting then cannot be explained with the help of standard demand mechanisms and positive supply impulses were admittedly underestimated in the past” (pp. 58–59).

¹¹ De Carvalho Filho, and Chamon (2008), Yörükoglu (2010), and Arslan and Ceritoğlu (2013) also report significant inflation forecast bias for a group of emerging economies: Brazil and Mexico, and Turkey, respectively.

¹² Pintilescu et al. (2014) find that higher inflation in transition economies brings about significant inflationary uncertainty. Benczúr and Rátfai (2014) compare private consumption volatility patterns among emerging, developing, and advanced economies and find that transition economies in Central and Eastern Europe have relatively much higher consumption volatility than the other country groups.

¹³ It is well known that external developments in emerging economies heavily influence macroeconomic fundamentals, including perceived price changes. For example, Maćkowiak (2006) reports that a major source of change in the aggregate price level in the Czech Republic, Hungary, and Poland is driven by external shocks. Łyziak (2009) finds that energy, as a frequently purchased key item, affects perceived inflation for Polish consumers. Investigating the relative importance of domestic and foreign shocks in explaining inflation dynamics in the Czech Republic and seven other transition economies, Globan et al. (2015) report that foreign shocks are more important in the medium run (more than a year), while domestic shocks dominate short-run (6- to 12-month) movements in inflation. Experience of other emerging markets is not any different (for evidence for China, see Tang, Wang, and Wang 2014; Zhang and He 2015). In addition, Wan and Jin (2014) provide empirical evidence that emerging economies tend to recover from financial crises at a relatively high speed, which causes further fluctuations in output and inflation in a relatively short time period. Overall, what we learn from these studies is that a good inflation forecast in an emerging economy requires efficient projections of external shocks, including global financial crises, and their duration. However, this is a challenging task, which may also explain why it might be easier for local agents to forecast domestic developments better than external ones.

¹⁴ Hałka and Łyziak (2015) report that perception of price changes by Polish consumers is heavily driven by price changes of frequently purchased goods and services, reflecting their purchasing patterns. In addition, Polish consumers tend to disregard negative price changes of these items.

¹⁵ For more details on the model used by the Czech National Bank, see Andrle et al. (2009).

¹⁶ By monetary policy, we mean interest rates. The basic interest rate set by the CNB is the 2-week repo rate (so-called repo rate), and the discount and Lombard rates always follow the repo rate, so that the discount rate is the repo rate minus 100 basis points, and the Lombard rate is the repo rate plus 100 basis points. Nevertheless, following the global financial crisis, the repo rate was set at 0.75 and the discount rate at 0.25, because it cannot be zero, since many penalties in the financial market are set as a multiple of the discount rate. The Lombard rate was thus 1.75.

¹⁷ We treat extraordinary meetings of the bank board as meetings that are not in accord with staff recommendations, because by definition, the staff could not give a full-fledged recommendation based on its model. However, this need not mean that the staff opposed the rate change at the very moment of the extraordinary meeting.

¹⁸ Although in absolute terms, the number of meetings at which the Board refused to cut was higher than the number of meetings at which it refused to hike, we consider a relative indicator (number of refusals to cut or hike/number of proposals to cut or hike) a better measure.

Additional information

Funding

This article was supported by Grant No. GACR P402/10/0289, “Instability of the Financial Markets and Effectiveness of Their Regulation,” financed by the Grant Agency of the Czech Republic (http://gacr.cz). An earlier version of the article was presented at the annual meeting of the Association for Comparative Economic Studies (http://www.acesecon.org). The views presented here are solely the personal views of the authors and do not necessarily reflect the views of the institutions with which they are affiliated.

Notes on contributors

Josef C. Brada

Josef C. Brada is at the W.P. Carey School of Business, Arizona State University, Tempe, AZ.

Jan Kubíček

Jan Kubíček is an Economist at the Czech National Bank.

Ali M. Kutan

Ali M. Kutan is a research fellow at the Jiangxi University of Economics and Finance, China, and Professor of Economics in the Department of Economics and Finance at Southern Illinois University, Edwardsville, IL.

Vladimír Tomšík

Vladimír Tomšík is a Vice-Governor at the Czech National Bank.