تورم قبل و بعد از استقلال بانک مرکزی : مورد کلمبیا

In this paper we model the Colombian inflation rate in terms of excess demand effects from asset, goods and factor markets. In contrast to previous results for a group of industrial economies, we find that domestic factors are a far more powerful influence on inflation than are external factors. The paper pays particular attention to the potential effects of the Constitutional Reform of 1991, which created a Central Bank independent from other parts of government. We find that the creation of an independent Central Bank did change some of the parameters of the model, as the disequilibria in goods and monetary markets were found to have a smaller effect on inflation after Central Bank independence was granted.

One of the main subjects of concern for policymakers and economists alike is the behaviour of inflation. Empirical evidence for a large number of countries reveals that high and variable rates of inflation are not consistent with sustained economic growth, because they shorten the planning horizon of investors and reduce the rate of productivity growth in the economy (see, e.g. Fisher, 1993 and Barro, 1995). Understanding the main factors that affect the dynamics of inflation is thus crucial to help policymakers design measures to achieve a stable macroeconomic environment, and gain insight about the effects of their policies. Economic theory suggests alternative views to explain the sources of inflation. A first view is associated with the monetarist school, according to which the main cause of inflation lies in expansions of the money supply in excess of real productivity growth. A second view focuses on the external factors that affect the domestic price level in an open economy, either through the transmission of import prices inflation in foreign currency terms into domestic inflation, or through the influence of the exchange rate on prices (via prices of imported intermediate and final goods). A third view lays emphasis on internal theories, which may be further subdivided into labour market theories and excess demand theories. The former highlights the role of the wage, being the result of labour demand and supply interactions, as a component of producers costs, while the latter refers to excess demand pressure effects. There has recently been an increasing interest in studying the effect of institutional factors, like the impact on inflation rate resulting from the degree of Central Bank independence from other branches of government (see, e.g. Alesina and Summers, 1993 and Cukierman, 1992). This paper investigates the determination of the rate of inflation in Colombia in terms of the explanations mentioned above. The basic idea of the paper is that inflation can be associated with excess money supply, demand pressure effects, imported inflation, and wage inflation. The analysis of the main determinants of inflation in Colombia has certainly been a topic of dynamic research throughout the years; see, e.g. Misas et al. (1999) and the references therein. To our knowledge, however, existing literature has not investigated inflation dynamics accounting for all explanations, as it has mainly focussed on only one of the possible origins of inflation. In addition, research on Central Bank independence has typically been addressed within the context of cross section or panel data, but not within a pure time series model as the one we present in this paper.1 Our empirical modelling exercise is implemented in two steps. First, we use multivariate cointegration techniques to test for the existence of long-run equilibrium relationships in three different systems of equations, describing the monetary sector, the foreign sector, and the labour sector. Second, the deviations from the estimated cointegrating relationships are included as determinants in an inflation model. In other words, we expect the inflation rate to adjust to deviations from the long-run cointegrating relationships derived in the first step of the analysis. Modelling inflation in terms of disequilibria from several sectors is important in numerous studies of other economies; see, e.g. Surrey (1989) for the US and the UK, Juselius (1992) for Denmark, and Hendry, 2000 and Hendry, 2001 for the UK, among others. Surrey (1989) and Juselius (1992) find that, within the industrial economies considered, the external influence on domestic prices is a far more powerful influence compared to the domestic influence. Hendry, 2000 and Hendry, 2001 finds that most theories of inflation help explain UK inflation over the last century and a quarter, providing no support for any “single-cause” explanation (in particular, excess money is not found to play a key role). The study of the Colombian case is therefore interesting, because it enables us to determine the importance of internal, external and institutional influences on domestic prices within the context of a developing economy. Our paper differs in one important aspect from these earlier works. The analysis of the Colombian experience allows us to assess the effects of the Constitutional reform of 1991, which radically modified the structure and functions of the Central Bank with the aim of creating an institution independent from the government's executive branch. The Constitution of 1991 created the Board of Directors of the Central Bank, which consists of seven members: the Minister of Finance who presides the Board without veto power; five Co-Directors who are appointed by the President and serve for a minimum period of 4 years and a maximum of 12; and the Governor of the Bank who is elected by the Co-Directors for a minimum period of 4 years and a maximum stay of 12. The terms of the Co-Directors are staggered so that no President can appoint the entire Board at any time. The Constitution stated that the main objective of the Bank was to control inflation, and that it had to coordinate its policies with government macroeconomic policies. The Constitution also forbids the Bank from lending to non-financial private agents—loans to the government are only permitted after an unanimous vote by the Board, something which has not happened up to now. The institutional changes introduced by the Constitution of 1991 have now been in place long enough to allow for some hope of identifying the effects they may have had on inflation. Colombian inflation was on the order of 25% per year in the 1970s and 1980s. This, according to Dornbusch and Fischer (1993), constitutes a classical example of a moderate but persistent inflation process. Since the 1991 reform the inflation rate has been exhibiting a sustained declining path, and is now in the one-digit inflation ground (about 7.5% in 2001). Gómez et al. (2002) and Restrepo (2000) associate this marked change of the inflation rate with the high level of Central Bank independence. Our modelling exercise indicates that the independence of the Central Bank did not affect the autonomous level of inflation, but it changed the response of inflation to disequilibria in the goods and money markets. In particular, the disequilibria in the goods and money markets are found to have a larger effect on inflation before Central Bank independence, suggesting that monetary policy simply accommodated these disequilibria. The paper is organised as follows. Section 2 presents the cointegration properties of three different systems of equations, describing first the monetary sector, then the foreign sector, and finally the labour sector. Section 3 estimates a model for the determination of the inflation rate in Colombia using results derived from the cointegration analysis. The model allows for the potential impact on inflation of the creation of an independent Central Bank. Section 4 concludes.

This paper has estimated an inflation model for Colombia in terms of disequilibria in the monetary sector, the foreign sector, and the labour sector. The model also allows us to study the potential effect on inflation of the creation of an independent Central Bank, granted through the constitutional mandate of 1991. Within the context of vector autoregressive models for these three sectors, evidence has been found for the monetary sector of the existence of two cointegration vectors, which can be interpreted as measures of excess money and of excess demand. The cointegration analysis of the foreign sector provided support for a weak version of PPP between Colombia and the US. And from the analysis of the labour sector we determined a long-run equilibrium relationship among wages, prices, and unemployment. Relating the estimated measures of market disequilibria to the inflation rate, coefficients are estimated with the theoretically correct sign: excess money, excess demand, and deviations from PPP have a positive effect on inflation, while wages above the steady-state do not appear to have an effect. Our findings indicate that disequilibria in the money and goods markets are a far more powerful influence on inflation than are external factors. Another important result was the rather high autonomous part of the Colombian rate of inflation. The study of the Colombian case thus offers an interesting contrast with previous results obtained for the US, the UK, and Denmark, where external factors were found to be the main driving force behind inflation, and the autonomous inflation component was much smaller (or even insignificant). It is also interesting that the estimated inflation equation shows a lower level of inflation persistence than is usual in other estimates, certainly consistent with treating it as an I(0) variable. The Constitutional Reform of 1991, which gave greater political independence to the Central Bank, did change some of the parameters of the model, as the disequilibria in goods and monetary markets were found to have a larger effect on inflation before Central Bank independence was granted. These larger effects suggest that monetary policy accommodated the disequilibria in the goods and in the money markets. Such accommodation is now effectively forbidden by the Constitution, as the Central Bank cannot lend money to the non-financial private sector, and requires unanimity of its Board of Directors to lend money to the government, something that has not happened thus far.