There are a number of monetary policy choices facing low-income countries. In many ways, the academic research would suggest there is no single best monetary policy for all countries. Low income countries face a unique set of challenges when adopting monetary policy. Primary problems include a lack of credibility, technical expertise and country-specific shocks. For many of these countries, alternative regimes such as a pegged exchange rate to a close trading partner might be most appropriate. While I recognize this possibility, this essay will not seek to address the relative merits of all monetary and exchange rate policy choices available. Rather, in the instance were a nominal anchor is preferable and implementable, this essay will seek to consider the relative merits of monetary targeting (MT) and inflation targeting (IT) for low income countries.
The goal of monetary policy should be to provide an environment in which macroeconomic stability fosters long-term economic growth. In light of this, the best framework to evaluate monetary policy is in its ability to provide both inflation and output stability. I will compare MT and IT by considering why OECD countries have moved away from monetary targets to inflation targets and then I will consider if the shocks low-income countries (LICs) face and the monetary transmission mechanisms in these countries make monetary targets more appropriate.
Many OECD countries that had explicit monetary aggregate targets have phased these policies out. For example, in 1980 the UK adopted an M3 growth target but the relationship between M3 and nominal incomes broke down causing the target to change to M0 and eventually abandoned altogether. Similarly, Canada abandoned its monetary targets in 1982 with the central bank governor, Gerald Bouey, famously saying, “ We did not abandon monetary aggregates, they abandoned us.” The reason for this move away from monetary targets is two-fold. The first reason is that it can be shown that IT dominates MT in response to demand shocks. That is to say, targeting inflation provides greater stability to both inflation and output. A demand shock moves inflation and output in the same direction. A negative demand shock drops inflation and vice versa. As a result, leaning on policy levers to return inflation to the target range should move output in the correct direction as well. Targeting inflation directly provides a more stable inflationary environment without the cost of higher output stability in comparison to an M-based framework. Indeed, the adoption of inflation targeting has resulted in lower inflation volatility, but also lower output volatility. (De Gregorio, 2007) The second reason is that the demand for money is not constant. As a result monetary aggregates are an indirect and flawed indicator for future inflation and other goal variables. Rapid developments in financial innovation have resulted in the velocity of money becoming an increasingly complex , endogenized and volatile variable. This was the primary reason for the failure of monetary targets in the UK, US and Canada during the 1970s and 1980s. (Mishkin, 2000) If most OECD countries have moved away from monetary targets, why might they still be appropriate for LICs? There are two broad reasons: the types of shocks and monetary transmission in LICs
LICs may face supply shocks more regularly than demand shocks. Many low-income countries have concentrated consumption baskets. In particular, food forms between 40-60% of the CPI basket for most low income countries.
Fuel too is an important input and even in crude producing countries, refined petroleum products are almost exclusively imported at world prices. The result is that CPI is disproportionately weighted toward traded goods. In contrast to demand shocks, supply shocks move output and inflation in opposite directions. Consider a supply shock such as the Arab Spring or the poor Russian wheat harvests of 2011. Under strict inflation targeting, policy-makers should raise interest rates to limit CPI inflation. Higher oil and food prices will reduce disposable income and an increase in interest rates will further decrease domestic absorption. The result is that strict inflation targeting may result in large levels of volatility in output for a country that regularly faces supply shocks. A monetary target by contrast would not require an interest rate hike to the same degree in response to an external supply shock.
The solution that an inflation targeting regime can adapt is to target core inflation. Core inflation measures the change in the price level of non-traded goods and excludes items such as fuel and food. While monetary policy cannot impact the inflationary impact of imported price shocks, the secondary effects can be influenced. As a result, if inflation expectations and wage bargaining are more closely tied to headline inflation then it will be difficult for the central bank to focus exclusively on core inflation measures. This may be especially true in LICs without a track record of credible nominal anchoring and LICs in which locally manufactured goods have a high proportion of their inputs in imported items such as fuel. Sustained levels of above-target headline inflation may result in inflation expectations losing their anchor to the target range and in a self-fulfilling cycle resulting in both core and headline inflation moving above the target range and prove persistent as the central bank struggles to re-establish credibility. This would suggest that, at least initially, LICs would have to target headline inflation more vehemently and as a result, may suffer higher output volatility than necessary in order to build a credible inflation targeting track record.
The second reason monetary aggregates may be appropriate in LICs is that their monetary transmissions mechanisms are different from developed economies. monetary transmission mechanism is a complex and much debated topic. The traditional transmission mechanics of monetary policy takes place through central bank intervention in discount rates and short-term government securities which impact interbank interest rates. In turn, through competition, this should lower interest rates banks offer individual and corporate customers. Thus an easing of monetary policy would increase domestic absorption. Firstly, consumption would increases as intertemporal substitutability allows consumers to shift more consumption forward as the relative cost of credit to time-preference decreases. Corporates would also increase investment until the marginal return to capital matches the lower interest rate. Indirectly, an easier monetary policy should weaken the exchange rate and thus boost demand for exports. Transmission mechanisms, however, are considered to include more complex and imperfect avenues through which monetary policy impacts output. These include consumption effects as a result of changes in property or equity asset markets. Furthermore, bank competition may be imperfect as a result of regulation or oligopolies and this may limit the transmission from lower bank funding costs to the impact on bank lending to both individual and corporates. The monetary transmission mechanisms described above rely on a number of effective arbitrage channels between different domestic securities, foreign securities and asset markets. These channels are typically not present unless there exists a well-functioning and highly liquid interbank market for reserves, government securities and for equities and real estate. Furthermore, it requires that there is a high degree of practical international capital mobility. (Mishra et al. 2010) These assumptions may not apply to LICs. But while much has been said about monetary transmission mechanism in developed economies with these features, there has been little research on the monetary transmission mechanisms of LICs.
Mishra et al (2010) list 12 ways in which LICs differ substantially. These can be summarized into financial development, institutional environment and financial integration
First, formal financial institutions are not as prominent in LICs. The ratio of deposits-to-GDP is 0.32 for LICs which is significantly lower than the 1.24 ratio for developed economies. In addition to the lack of penetration , the concentration in the banking sector is higher in LICs than in developed economies. This further weakens the transmission mechanism between banks and agents in the real economy. Many LICs have tiny bond markets or private bond markets simply do not exist and thus the transmission between central bank actions, foreign capital and bank funding rates are limited. Furthermore, the lack of liquid stock and real estate markets limit the indirect transmission mechanisms through asset markets that monetary policy might have.
Second, institutional and legal deficiencies in the market make the costs of financial intermediation substantially higher. One possible consequence is the inter-bank market is under-developed. The result of the lack of transparency and contract enforcement may result in mutual distrust that leads banks to all hold chronic excess reserves thus limiting the role an inter-bank reserve market might play. Saxegaard (2006) Finally, LICs are financially more isolated from the global financial system than developed economies. As a result, the arbitrage channel between domestic and foreign assets is limited and the impact of monetary policy through the exchange rate channel will be subdued.
The outcome of monetary policy in LICs has an interesting pattern given the difference in the financial sector. The correlation between changes in monetary policy rates and money market rates is significantly lower but it is consistent in the short-term and in the long-term. (Mishra et al. 2011) The opposite is true of advanced economies where changes in monetary policy have a greater long-run impact on money market rates but that impact changes over time as the transmission works its way through various channels. What does this mean for monetary policy? It indicates that in economies with complex financial systems, the demand for money is both endogenous and variable and thus targeting monetary aggregates is an unnecessarily complicated target when trying to achieve a more general goal of macro-economic stability. But for LICs, this complexity is not necessarily as prominent. Thus in an economy dominated by supply shocks, the cost associated with monetary targeting may not be a particularly problematic
Ultimately, some of the drawbacks of inflation targeting will diminish overtime. Hard earned credibility will allow greater flexibility between headline and core inflation. Many economists argue that it is the commitment to price stability that has allowed central banks to focus on core inflation during the recent oil and food prices spikes without letting inflation expectations rise as they did in the 1970s. The figure below indicates this phenomenon for the US.
Indeed, this has been the experience for many industrialized nations. Inflation targeting did not provide instant credibility and credibility was earned with the cost of below-normal output during disinflationary adjustment. But output returned to normal levels once target ranges had been achieved. The period following this was generally followed by higher growth and lower output volatility. (Mishkin, 2000). While the benefits of inflation targeting will grow, the advantages of monetary-based targets will diminish as financial development gathers speed and increasing innovation in micro-credit, insurance and asset markets unpredictably changes the transmission dynamics. Even in the slightly more developed emerging markets, the problem of instability of the money-inflation relationship has been found by Mishkin and Savastano (2000.) Furthermore, the instability in the relationship between inflation and monetary aggregates may work to decrease the accountability and public communications abilities of central banks in LICs. So in conclusion, I believe that where individual country circumstances do not justify alternative monetary policy choices, the base case should be a monetary policy of inflation targeting.
Batini, N. 2007 Monetary Policy in Emerging Markets and Other Developing Countries
Nova Science Publishers, New York.
Hammond, G., Kanbur, R., Prasad, E. 2009 Monetary Policy Frameworks for Emerging Markets
Edward Elgar Publishing Ltd, UK. The Bank of England.
Mishkin. F. 2000. From Monetary Targeting to Inflation Targeting: Lessons from Industrlialized Countries. Columbia University & NBER
Mishkin, F., Savastano, M. 2000, Monetary Policy Strategies for LatinAmerica
National Bureau of Economic Research Working Paper No. 7617, March.
Mishra, P., Montiel, P., Spilimbergo, A. 2010. Monetary Transmission in Low Income Countries
IMF Working Paper, Research Department and European Department.
Mishra, P., Montiel, P., Spilimbergo, A. 2011.How Effective is Monetary Transmission in Developing Countries? A Survey of the Empirical Evidence
Centre for Economic Policy Research. Discussion Paper 8577
Saxegaard, M., 2006, Excess Liquidity and Effectiveness of Monetary Policy: Evidence
from Sub-Saharan Africa IMF Working Paper No. 06/115