Asian Research Thesis Index

Abstract

This dissertation explores different dimensions of the relationship between inflation and relative price variability in Pakistan. The period covered for this study is from July 2001 to June 2011, as this is the complete period for which data on new base (2000-01) is available. Incidentally, this period has both the low and high inflation episodes. This is the only study in case of Pakistan, which uses detailed data on prices at city level and commodity group level. The dissertation has analyzed relationship between inflation and relative price variability in three aspects. First, we have examined the behavior of price setting agents as reflected in relative price changes in response to demand and supply factors; second, we studied the effect of relative price variability on inflation by estimating fixed effects regression model using panel data of inflation in different cities of Pakistan; and third, we have examined convergence of prices changes in 35 cities of Pakistan, and also looked at how location of cities affects the convergence. The result of our first study suggests that changes in real income have insignificant impact on relative price variability. The results make sense as changes in income (with given preferences) almost evenly affect demand for all consumer items, which may lead to relatively proportional changes in their prices. It can be a case particularly in a developing economy like Pakistan, having a large informal sector, where response of firms is less constrained by wage contracts; and where capacity issues are less heterogeneous. On the other hand, unanticipated inflation, which usually comes from item-specific supply factors, may affect prices of different items unevenly. From the second study, the results show that inflation, both food and non-food inflation, is significantly and positively affected by relative price variability. The results imply that supply side factors, as exhibited in dispersion of relative price changes, are robust determinant of inflation in a developing economy, like developed economies. From our third study, we found that there is bilateral price-level convergence for only food group with speed of convergence (measured by half-life) is around 3 months. On the other hand, prices of non-food commodities have very low speed of adjustment with 20 month half-life. Consequently, relative prices of overall commodities group have half-life of 8 month – a moderate speed of convergence. We have also identified differences in the behavior of relative prices within and across provinces of Pakistan. The relative prices between two cities located in the same province show lower variability compared with cites pair located in different provinces. However, if at least one of city associated with a relative price series is located in one province, standard deviation of relative prices rises in case of overall and food group. While exploring the impact of distance between cities of a pair, we have found that the standard deviation of relative prices increase significantly with the distance. This result accords well with the findings of some previous studies e.g. Engle and Rogers (1996). The policy implications from my study is; as the supply side factors are found to be dominant in affecting economic activity and inflation rate in Pakistan, therefore, monetary authority needs to be careful while taking decisions on monetary policy instrument. For instance, in 2008 when inflation rate was approximately 20 percent, SBP increased discount rate to give a signal of tight monetary policy stance. This badly affected economic activity at that time and GDP growth rate turned out to be zero. Therefore, cost push inflation should be dealt with much care while taking monetary policy decisions. Another implication of this research is that monetary policy may target a narrow measure of general price level. For instance, core inflation can be targeted. Moreover, an index of general price level can be constructed that is in control of monetary policy with minimum control error.

Add/Update Thesis