Estimation of Medium Term Inflation Target for Pakistan

Transcription

1. SBP Working Paper Series No . 104 February, 2020 Estimation of Medium Term Inflation Target for Pakistan Fayyaz Hussain Dr. Muhammad Rehman STATE BANK OF PAKISTAN
2. SBP Working Paper Series Editor : Sajawal Khan The objective of the SBP Working Paper Series is to stimulate and generate discussions on different aspects of macroeconomic issues among the staff members of the State Bank of Pakistan. Papers published in this series are subject to intense internal review process. The views expressed in these papers are those of the author(s) not State Bank of Pakistan. © State Bank of Pakistan. Price per Working Paper (print form) Pakistan: Rs 50 (inclusive of postage) Foreign: US$ 20 (inclusive of postage) Purchase orders, accompanied with cheques/drafts drawn in favor of State Bank of Pakistan, should be sent to: Chief Spokesperson, External Relations Department, State Bank of Pakistan, I.I. Chundrigar Road, P.O. Box No. 4456, Karachi 74000. Pakistan. Soft copy is downloadable for free from SBP website: http://www.sbp.org.pk For all other correspondence: Postal: Editor, SBP Working Paper Series, Research Department, State Bank of Pakistan, I.I. Chundrigar Road, P.O. Box No. 4456, Karachi 74000. Pakistan. Email: wps@sbp.org.pk ISSN 1997-3802 (Print) ISSN 1997-3810 (Online) Published by State Bank of Pakistan, Karachi, Pakistan. Printed at the SBP BSC (Bank) – Printing Press, Karachi, Pakistan
3. Estimation of Medium Term Inflation Target for Pakistan Fayyaz Hussain 1 & Dr. Muhammad Rehman 2 Abstract As a part of its strategic goal, State Bank of Pakistan (SBP) plans to switch to flexible inflation targeting regime. In this regard, the first and most important question is about setting appropriate medium term inflation target. This study aims to search for the medium term inflation target that is conducive for socioeconomic welfare. Seeking guidance from the international best practices, we have used broadly two procedures to search for appropriate inflation target and its band for Pakistan. First, we have analyzed the target setting practices of advanced and emerging inflation targeting economies. Second, we used a variety of technical methods. Results show that inflation level above 8 to 9 percent is harmful for the society. On the lower bound inflation below 4.0 percent is also undesirable. Keeping in mind these findings, inflation target range in the similar emerging economies, history of inflation volatility in Pakistan and importance of exchange rate, we recommend inflation target to be set at 5.5 percent with a band of +/- 1.5 JEL Classification: E310, E520, E580 Key Words: Price Level; Inflation, Monetary Policy, Central Banks and Their Policies Acknowledgments Authors would like to thank Waqas Ahmed, Omar Farooq Saqib, Mahmood ul Hasan Khan, and anonymous reviewer for the thoughtful comments on an earlier version of this paper. Contact for correspondence: Fayyaz Hussain Joint Director, Monetary Policy Department State Bank of Pakistan I.I. Chundrigar Road Karachi 74000. Email: Fayyaz.Hussain@sbp.org.pk 1 2 Joint Director, Monetary Policy Department, State Bank of Pakistan, Karachi (Fayyaz.Hussain@sbp.org.pk) Joint Director, Monetary Policy Department, State Bank of Pakistan, Karachi (Muhammad.Rehman@sbp.org.pk)
4. Non-technical Summary As a part of its strategic goal , State Bank of Pakistan (SBP) plans to switch to flexible inflation targeting regime. In this regard, the first and most important question is about setting appropriate medium term inflation target. Government’s annual CPI inflation target is largely based on recent price trends and keeps on changing both up and down and sometimes by big margins. Such frequent changes in inflation targets are not conducive for anchoring inflation expectations. This study aims to search for the medium term inflation target that is conducive for socioeconomic welfare. Seeking guidance from the international best practices, we have used broadly two procedures to search for appropriate inflation target and its band for Pakistan. First, we have analyzed the target setting practices of advanced and emerging inflation targeting economies. Second, we used a variety of technical methods. Global practice shows that most of the inflation targeting central banks of emerging economies are targeting inflation in the range of 2 to 6 percent. Advanced economies set inflation targets in the range of 1 to 3 percent. Analysis of inflation growth nexus shows that inflation band of 5.7 to 10.4 percent is conducive for economic growth. Conduct of monetary policy in Pakistan as measured by Taylor rule indicates implicit inflation target of 8.9 percent. Analysis of impact of inflation on price dispersion shows that inflation band of 4.3 to 9.1 percent corresponds to price stability. When economy operates close to its potential (zero output gap), inflation averages around 7 to 8 percent. Inflation in the band of 4.0 percent to 5.0 percent corresponds to the lowest wage dispersion in the economy. Keeping in mind a) these findings, b) inflation target range in the similar emerging economies, c) history of inflation volatility in Pakistan and d) importance of exchange rate, we recommend inflation target to be set at 5.5 percent with a band of +/- 1. 5. Page 4 of 24
5. 1 . Introduction In line with the general trend across the globe, over the last decade price stability has become one of the most important, if not sole objectives, of monetary policy in Pakistan. Despite this increasing importance, however, there is no consensus on the specific inflation target which is consistent with the definition of price stability1. As a result, government’s annual CPI inflation target is largely based on recent price trends and keeps on changing both up and down and sometimes by big margins (Figure 1). Such frequent changes in inflation targets are not conducive for anchoring inflation expectations. The importance of price stability consistent inflation targets has become all the more important with the State Bank of Pakistan (SBP) strategic goal of implementing flexible inflation targeting. In this backdrop, this study aims at exploring the inflation target band that is consistent with the definition of price stability and can be used as a starting point for medium term inflation target2. Figure 1: CPI Inflation and Government's Inflation Target Inflation Inflation Target 18 16 Change in percent 14 12 10 8 6 4 2 2019 2018 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1996 1995 0 Rest of the study is planned as follow. Section two briefly describes the theoretical debate on the meaning of price stability and global practices on setting inflation target. Section three discusses the empirical approaches of estimating inflation target band and replication of those approaches for Pakistan’s case. Section four recommends an appropriate inflation target along with tolerance band and time horizon for Pakistan and the last section concludes the study. European Central Bank (ECB) defines price stability as “a year on year inflation in the Harmonized Index of Consumer Prices for the Euro Area of below 2 percent”. Similarly, an inflation rate of 1 to 3 percent corresponds to price stability in Advance economies while in transition economies inflation in the range of 4 to 5 percent would correspond to price stability (Jonas. J and F.S. Mishkin, 2003). 2 Though there are different measures of price indices such as Consumer price index (CPI), wholesale price index (WPI), sensitive price index (SPI) and GDP deflator, we opted for CPI based measure of inflation for many reasons. First, all the inflation targeting countries are using this index for setting their inflation targets. Second, Government of Pakistan has already been setting inflation targets based on this index. Third, shortcomings of competing price indices also make CPI index the most suitable candidate for analysis. For instance, WPI does not include prices of services that constitute more than one half of the economy. Likewise, data on GDP deflator is available on annual basis only and SPI covers prices of only kitchen items. 1 Page 5 of 24
6. 2 . Background 2.1. Theoretical Debate Over the last six decades, Phillips curve has been used as a guiding principal in the monetary policy thinking for setting inflation targets. In the late 1950s and early 1960s, negative trade-off between inflation and unemployment was treated as menu for policy makers to choose from long run objective of price stability and unemployment. Empirical evidence showed that inflation can be reduced by tolerating higher unemployment rate or unemployment rate can be reduced by tolerating higher inflation3. Major blow to this paradigm came from higher inflation decades of the 1970s and 1980s. In 1970s and 1980s, monetary policy formulation remained under the influence of Milton Friedman’s work (1968) showing no trade-off between inflation and unemployment in the long run (vertical Phillips curve), though short run trade-offs remained. In this framework, central bank can choose the low and stable rate of inflation along the vertical line but it has no impact on the permanent rate of unemployment. With little control on NAIRU4, price stability or very low inflation became the overriding objective of monetary policy in this regime. In 1990s despite very low and stable inflation in United States and Canada, annual unemployment rate moved in a wide range raising questions about vertical Phillips curve in the long run. This inflation puzzle was solved by Akerlof et al (2002). They found that at higher level of inflation Phillips curve is vertical while as inflation falls below certain threshold it increases unemployment. Between these two thresholds, Phillips curve is negatively sloped. They attribute this increase in unemployment to downward nominal wage rigidity and near rational neglect of low inflation. Their findings implied that price stability corresponds to some positive inflation level. 2.2 Global practices on setting Inflation target Inflation rates above and below the certain ranges does not bode well for economies. There is plenty of literature on the costs of high inflation (for detail see O,Reilly 1998). Most prominent among these costs of high inflation include financial disintermediation (into Real Estate and Gold) and consequent decline in financial savings and investment contributing to lower growth. The literature on adverse impact of very low inflation, however, is relatively scarce. Frequently cited detrimental impact of very low inflation includes: (a) risks of growing deflationary concerns; (b) at very low inflation, downward nominal wage rigidity makes it difficult to cut real wage rate (in case of adverse demand shock) that adds to unemployment and hurts economic growth; and (c) limited room to cut real rate below zero in case of adverse shock. Keeping theses costs of very high and very low inflation in view, central banks across the globe target positive inflation rate. As per Jonas and Mishkin (2003), there is a general consensus in the literature that inflation rate in the range of 1 to 3 percent corresponds to price stability in the advanced economies. This is also evident from inflation targets of advanced inflation targeting countries (Figure 2). 3 Phillips (1958), Samuelson and Solow (1960), Lipsey (1960), Eckstein and Wilson (1962), Perry (1964) and Kaliski (1964). 4 Estimates of NAIRU differed across countries and time periods and so did the optimal inflation. Page 6 of 24
7. Figure 2 : Inflation Targets Levels and Bands in Advanved Economies as of June 2018 5.0 4.0 3.0 2.0 1.0 Norway Iceland Australia Korea United Kingdom Sweden New Zealand Japan Israel Canada Czech Republic 0.0 In case of emerging economies, analysts argue that emerging economies should target somewhat higher inflation than their developed counterparts because of Harrod-Belassa-Samuelson effect5, measurement error in inflation, price level convergence, low credibility and low independence.6 For instance, Škreb (1998) argues that owing to measurement error, inflation in the range of 4- 5 percent would correspond to price stability in transition economies. The hypothesis of price convergence suggests that emerging economies with lower price level set a higher inflation target to reflect the expected price convergence toward their developed counterpart. Likewise, with the less credible central bank, emerging economies can manage inflation expectations less effectively than developed countries. This makes them set inflation targets above the targets set for developed economies. Lastly, in emerging economies central banks are more prone to government influence. This limits their ability to keep inflation at a lower level and they choose a relatively higher inflation target than that of developed countries. Figure 3: Inflation Targets Levels and Bands in Emerging Economies as of June 2018 12 10 8 6 4 2 Ghana Argentina Ukrane Uruguay Kazakistan Turkey Uganda Brazil Moldova South Africa Paraguay India Arminia Guatemala Dominica Rep Russia Serbia Georgia Philipines Indonesia Mexico Hungary Colombia Chile Albania Romania Poland Thailand Peru 0 In line with the above mentioned arguments, most of the inflation targeting central banks of emerging economies are targeting inflation in the range of 2 to 6 percent (Figure 3), higher than range (1 to 3 percent) of the developed economies. It is also worth noting that inflation target bands in the emerging 5 Countries with high productivity growth also experience high wage growth, which leads to higher real exchange rates. The Balassa-Samuelson effect suggests that an increase in wages in the tradable goods sector of an emerging economy will also lead to higher wages in the non-tradable (service) sector of the economy. The accompanying increase in inflation makes inflation rates higher in faster growing economies than it is in slow growing developed economies. 6 Clinton (2000), however, argue that rapidly rising labor productivity in emerging economies does not support targeting relatively higher inflation in these countries. Page 7 of 24
8. economies are wider than those of the advance economies . This probably reflects the more volatile macroeconomic environment in the emerging economies. 3. Search for medium term price-stability consistent inflation target for Pakistan Empirical literature on the selection of inflation targets can be broadly categorized into three distinct approaches. First are the macro econometric models. Most prominent among these are studying the impact of inflation on aggregate output,7 and estimating implicit inflation target from Taylor rule.8 Statistical methods such as relative price variability (RPV) and output gap can be categorized as the second approach9. The impact of inflation on RPV is considered an important channel for real effects of inflation. The RPV analysis suggests inflation thresholds that have smaller impact on RPV. Output gap analysis search for the level of output where aggregate demand meets aggregate supply and given all else constant, inflation gravitates to its long run value.10 Besides macro econometric and statistical approaches, a third type of contribution has been purely microeconomic method.11 This approach studies the impact of inflation on relative real wage variation and finds that inflation-induced increase in wage variation that is harmful for economic growth. Seeking guidance from the above international practices, we use macro econometric, statistical and microeconomic methods to search for appropriate inflation target for Pakistan. These methods and their findings are discussed in greater detail below. 3.1 Macro econometric models 3.1.1 Inflation growth nexus Central banks across the globe are usually mandated with the dual objective of price stability and maximum sustainable level of growth. These two objectives are interlinked. Not only level of economic growth vis-à-vis its potential influences inflation trends but inflation environment also impacts economic growth. A plenty of literature has consensus that high inflation has negative impact on medium to long-term growth.12 Many researchers have tried to explore the level of inflation above which it is inimical to growth. For instance, Khan and Senhadje (2001) found such inflation thresholds in the range of 1-3 percent for developed economies and 7-11 percent for developing economies.13 This implies that price stability consistent inflation target should be set below the inflation threshold that hurts economic growth. Regarding Pakistan, a simple data analysis suggests that inflation rate in the range of 5 to 7 percent appears to be beneficial for economic growth. Specifically, using annual data on inflation and economic growth from 1976 to 2016, we computed average growth rates against different bands of inflation. Maximum economic growth corresponds to the period when inflation remained in the band of 5 to 7 percent (Table 1). 7 This approach finds the inflation threshold above which it is harmful for real growth. Most influential contribution in this regard is that of Khan and Senhadji (2001) and Rober Barro (1997). 8 Main contribution in this regard is from Clarida et al. (1998) 9 Most influential paper is by Park (1978) 10 Mishkin (2007) 11 A good example of this approach is the work done Groshen and Schweitzer (1996; 1999). 12 See for example Barro (1997), Khan and Senhadji ((2001), Fischer (1993) etc 13 Robert Barro (1997) using a panel of 100 countries found that inflation rate above 15 percent is definitely harmful for growth. Page 8 of 24
9. Table 1: Relationship between Inflation and Growth (FY76-FY16) Inflation range Average growth upto 3% 4.7% 3% to 5% 4.9% 5% to 7% 5.9% 7% to 9% 4.9% 9 % above 4.7% Based on the growth strategy, empirical studies on Pakistan suggest that inflation thresholds lie in the range of 4.0 to 9.0 percent. Specifically, Mubarak (2005), Hussain (2005) and Iqbal and Nawaz (2009) found inflation thresholds at 9.0 percent, 4 to 6 percent and 6.0 percent respectively. In a more recent study, Arby et al (2017) estimated inflation threshold in the range of 6.3 to 8.9 percent. However, these studies on Pakistan have not addressed important econometric issue of inflation endogeniety identified in the literature. For instance, Barro (1996) mentioned that inflation is endogenous variable that may respond to growth or other variables related to growth. Likewise, Khan and Senhadje (2000) acknowledged the issue of inflation endogeniety and checked the robustness of their results by applying instrumental variable approach. In case of simultaneity problem, the coefficient estimates of above studies on Pakistan would be biased. This study attempts to address the potential endogeniety of inflation by applying threshold vector autoregressive (TVAR) model. Before estimation of TVAR model, we have applied Likelihood Ratio (LR) test of Hansen (1999) to test linearity against the threshold(s) between inflation and growth along their other control variables. LR test basically compares the co-variances of different regimes: