الگوی اقتصادسنجی اقتصاد مالاویان

This paper has estimated a small-open economy macroeconomic model for Malawi. The structure of the model consists of production, expenditure, government, monetary, employment sectors and prices. The estimated parameters of the long run version of the model were used to perform simulation experiments to determine the model's tracking performance of the historical data and to assess the effects of changes in selected exogenous variables on key macroeconomic variables. The dynamic simulation results indicate that a sustained devaluation of the Malawi kwacha improves the real trade balance, but leads to higher inflation and reduces real GDP growth. Bond-financed increases in government consumption expenditures are less inflationary, lead to higher real GDP growth, but worsen the real trade balance position. The short run version of the model was estimated using the cointegration estimation technique.

The model developed in this paper utilizes the small-open economy IS-LM aggregate supply framework and takes into account the unique features of the Malawian economy.1 Although this framework has been associated with analyzing dynamic macroeconomic phenomena in developed economies, this view is changing. Indeed, such models have been estimated for a number of developing countries, both for individual countries and groups of countries (see, e.g. Elliott et al., 1986 and Haque et al., 1990). It is in this spirit that the present model of Malawi is developed and used for simulating the effects of alternative policies on the economy of Malawi. The current model has been estimated in line with the co-integration technique. This involved, as a first step, specifying the form of the long run relationship and collecting the appropriate data. The second step involved determining the order of integration of the relevant time series. This was achieved by the use of augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests. The variables in the co-integrating equations were found to be integrated of order one, I(1), and the residual formed from the co-integrating vector was I(0) at the 5% significance levels of ADF and PP tests. The third step consisted of testing whether or not the relevant I(1) time series were co-integrated. This was achieved by performing the Johansen Cointegration test. The final step involved specifying the form for the short run relationship. In the short run, changes in the endogenous variable on the left-hand-side of the equation adjust to the long run via the error correction term and also in some cases adjust directly to changes in the factors included in the long run equation. The annual data set used in the model estimation spans the period 1967–1996. In general, data availability is relatively good. The national account statistics and government finance have been continuously available since independence in 1964. Data on monetary aggregates have been available since 1965. The export and import prices are available since 1967. There are, however, a number of gaps, which affected the overall modeling strategy and choice of specification of individual equations. In particular, data on employment and wages are relatively weak and are not up-to-date at sector level. The data used to estimate the model were obtained from various secondary sources such as the International Monetary Fund's International Financial Statistics, Reserve Bank of Malawi's Financial and Economic Review, and Malawi Statistical Year Book. The remainder of this paper is organized as follows. Section 2 highlights the key features and developments in the Malawian economy. Section 3 describes the specification and structure of the model and includes a discussion of the theoretical underpinnings of the equations and the empirical results. In Section 4, the results of simulation experiments to evaluate the forecasting accuracy of the long run version of the model and the impact of alternative policies on main macroeconomic variables are reported. Finally the conclusions and policy implications which have emerged from the study are summarized.

This paper has described the structure and properties of a medium-scale macroeconometric model of the Malawian economy. The objectives of the paper were twofold. First, to construct a small-open economy model for Malawi with the aim of analyzing stabilization and other short-run problems. Second, to use the estimated parameters of the model to perform dynamic simulation experiments and thereby determine the effects of changes in exogenous variables on the key macroeconomic variables. The model has 37 endogenous variables of which 23 are explained by stochastic equations. In estimating the short run version of the model, the co-integration technique was applied to annual data spanning the period from 1967 to 1996. The estimated slope coefficients always occurred with theoretically consistent signs and most of them were significant at the 5% level. Although the paper provides a detailed analysis of the demand side, the supply side of the economy also received adequate treatment thereby maintaining a fairly balanced synthesis between the demand and supply sides of the Malawian economy. Labour and capital stock were found to be important factors of production explaining changes in output. This was evident from the statistical significance of both labour and capital variables. On the demand side, disposable income is the main determinant of private consumption. The major determinants of fixed investments are domestic saving, output, and imports of plant and machinery equipment. The results of this study show that there is a strong positive complementary relation between the import of plant and machinery equipment and both the private and government fixed investment expenditures. Real GNP of industrial countries and the relative price are the most important determinants of demand for Malawi's merchandise exports, while Malawi's real GNP, relative price, and availability of foreign exchange are the main determinants of her import demand. The economic reforms of 1981, 1988 and 1994 have also played a significant role in influencing the levels of exports and imports. In the monetary sector, the main determinants of demand for monetary aggregates are real GNP and, for some aggregates, nominal interest rates. The oil price of 1972–1974 also had a significant and positive impact on demand for narrow money. Real GNP, nominal interest rate, and dummy variables D81 and D88 determine the demand for time and savings deposits. According to the estimates of the model, the main determinants of employment are demands for output both by domestic and foreign consumers. The determinants of nominal wage rates are the tightness of the labour market and the cost of living as measured by the CPI. Lagged wage bill per output and price of imports determine both the GDP deflator and CPI. The GDP deflator and CPI indexes were estimated in the long run version of the model subject to the linear constraint that coefficients of the wage bill per output and price of imports add up to unity. Having estimated the structural equations of the model, the second objective was achieved by conducting dynamic simulation experiments to determine the responsiveness of various endogenous variables to sustained discretionary changes in selected exogenous variables. The long run version of the model was used in this regard. Three experiments were conducted. The first experiment involved a 10% devaluation of the Kwacha. The second involved a 10% increase in government consumption. And, the third entailed a 10% increase in foreign GNP. The results for the current study confirm that, for a small open economy under a managed exchange rate regime, policy changes can have highly significant and distinct effects on the macroeconomic situation. The results show that devaluation in the long run produces an improved balance of trade position but is contractionary in the sense that it reduces real GDP. The results also show that an increase bond-financed government spending has positive effect on real GDP but deteriorates the real trade balance position. The results are also indicative of lesser harmful effects on prices of a bond-financed government consumption spending. An increase in foreign GNP increases real GDP via a rise in real exports. The trade balance also improves in the second half of the simulation period. As noted earlier, continuing work with the model includes simulation experiments to evaluate the impact of changes in exogenous variables on endogenous variables using the estimated parameters of the short run version of the model.