THE TREATMENT OF EXPECTATIONS IN MACROECONOMIC MODELS

Get Complete Project Material File(s) Now! »

INTRODUCTION

Macroeconometric modelling is a fairly mature field in the discipline of economics. It has over the years continued to fascinate modellers and policy makers alike because of its dynamic nature and usefulness. Its importance in policy-making decisions need not be emphasized any further. Macroeconometric models have over the years and continue to provide effective analytical frameworks for policy-making decisions. The development of a sound and comprehensive macroeconometric model is of utmost importance if robust policy decisions are to be taken and followed. The entire exercise of profound macroeconomic management rests on the ability of the model to capture adequately and reasonably the key characteristic features of the economy under observation. As with other developing economies, sound economic analysis of the economy of Lesotho is to a large extent hindered by problems of inadequate and poor quality data. Furthermore, structural instability caused by persistent policy regime changes pose a major challenge in economic modelling. The fact that a robust database is a prerequisite for a model to produce reasonable forecasting and policy analysis qualities does not however preclude attempts to develop econometric models in such situations. It is however necessary that the limitations imposed by these constraints should be borne in mind.

Methodology

The study develops a model that captures both the supply and the demand side of the economy. The model is designed within the framework of the small-open economy ISLM-aggregate supply framework. It consists of four sectors, namely the real sector, the external sector, the monetary sector and the public sector. The real sector consists of the production sector and the employment sector which make up the supply side, and the aggregate demand and prices. As might be expected, many interrelationships exist between endogenous variables of the model. However, given the softness of the data and to minimize problems that may arise from this limitation, the model is designed to be as parsimonious as possible.
The model is made to mimic an economy that is small and open, supply constrained, has a surplus of labour, and in which markets do not clear. There are 65 endogenous variables of which 22 are determined by stochastic behavioural equations and 43 are determined by identities. The model also has 36 predetermined and exogenous variables. There are 39 dummy variables that are intended to capture structural and policy regime changes and other events that might have had significant impacts on the economy. Thus the model has 140 variables in total. The equations in the model are estimated using annual time series data spanning the period 1980 to 2000. The primary source of data is the Central Bank of Lesotho database and Annual Reports. Other sources include the Bureau of Statistics (BOS), the IMF’s International Financial Statistics (IFS) and South African Reserve Bank (SARB) Quarterly Bulletins.

Estimation techniques

The estimation technique used in this study is the Engle-Granger (1987) two-step procedure. The procedure involves testing for the order of integration of the individual data series to avoid the common problem of spurious regressions that incorrectly give the impression that relationships exist between two or more variables.1 Both the Augmented Dickey-Fuller (ADF) and the Phillips-Perron (PP) tests were used for this purpose. This is followed by testing for the presence of a long-run relationship between various sets of variables. This involves estimating the cointegration or long-run relationships by ordinary least squares (OLS) and subjecting the residuals derived from these relationships to unit root testing.
The ADF test was used in this case, although the critical values in this case are the McKinnon (1991) response surface values. The second stage involves constructing the error correction model (ECM), on condition that an equilibrium or longrun relationship between a set of non-stationary variables exists. This implies that the stochastic trends of these variables are linked and that the variables cannot move independently of each other. The ECM is then intended to estimate the short-run or dynamic adjustment process to long-run equilibrium. This approach has several merits; i) in the event that the concerned variables are cointegrated, the ECM captures both the short-run and long-run effects. The short-run component of the model becomes non-zero during periods of disequilibrium and imparts information about the distance of the system from equilibrium; ii) assuming cointegration exists and that estimates of the concerned parameters exist, all terms within the ECM are stationary.
This implies that standard OLS estimation techniques can be applied; iii) since the ECM is linked directly to the concept of cointegration, Granger’s representation theorem for dynamic modelling effectively implies that the presence of cointegration renders the ECM immune from the problem of spurious regressions; and iv) because it is possible to specify the ECM in a multivariate form, it is also practically possible to allow for a set of cointegrating vectors. By far, a major advantage of this approach rests in its better forecasting abilities over other methods and its ability to limit specification errors to those that arise because of inclusion of irrelevant variables rather than the omission of relevant variables.

Price developments

The 1980s were almost invariably characterised by a double-digit inflation rate. The inflation rate declined from 16.5 per cent in 1981 to 9.1 and 6.1 per cent in 1996 and 2000, respectively.6 This improvement was explained by the decline of the annual growth rates of individual indices from which the CPI is compiled. According to the CBL (1996), approximately 75 per cent of a combination of goods and services from which the CPI is computed comprise imports from South Africa (SA). This implies that imported inflation and exchange rate movements in SA that pass through to domestic prices explain most of the changes in prices. This, however, does not rule out the influence of pressure from aggregate demand despite efforts to curb inflation through wage restraints and tight monetary policy. Though the authorities have not had much leverage with respect to monetary policy, it is notable that the general price developments also match the stable growth of money supply during the 1990s.

READ Relationships between internal auditing and internal control system

SUMMARY OF MACROECONOMIC DEVELOPMENTS

The end of the 1980s and the beginning of the 1990s were characterised by major changes in both the structure of the economy and macroeconomic policy. Firstly, the inception of the LHWP in 1987/88 represented significant changes in the structure of the economy directly through elevated investment and capital inflows as well as imports. These have had great implications for the real sector of the economy as well as the balance of payments. Secondly, the adjustment programs in 1988/89 marked major policy changes for the economy. The ultimate outcomes of these have been observed and felt to a large extent in the public sector. Thirdly the decline in labour income, which has historically played a pivotal role in national income, was a major milestone during the review period. Having gone through a recession in the 1980s, an analysis of the performance of the economy in the 1990s reveals mixed outcomes.
On the fiscal front, a combination of fiscal restraints, tax reforms and increased imports and hence SACU receipts, seem to have jointly benefited the fiscal stance directly. Through stringent revenue raising and expenditure cutting policies, government has managed to secure a surplus position in its budget for a series of years. Increased LHWP capital expenditures and the improved fiscal discipline have tended to propel investment expenditures, while the fall in labour income has led to a decline in aggregate consumption expenditures. From the supply side, the former changes have jointly increased domestic output through the increased value added of the manufacturing and construction sub-sectors in particular. A view from the demand side reveals a rise in investment expenditures and a fall in consumption expenditures, though consumption remains significantly higher than investment in absolute terms.

TABLE OF CONTENTS :

LIST OF TABLES
LIST OF FIGURES
LIST OF ACRONYMS AND ABBREVIATIONS
CHAPTER BACKGROUND AND INTRODUCTION
- 1.1 INTRODUCTION
- 1.2 PROBLEM STATEMENT
- 1.3 OBJECTIVES OF THE STUDY
- 1.4 METHODOLOGY AND ESTIMATION PROCEDURES
  - 1.4.1 Methodology
  - 1.4.2 Estimation techniques
- 1.5 OUTLINE OF THE STUDY
CHAPTER AN OVERVIEW OF THE ECONOMY
- 2.1 INTRODUCTION
- 2.2 AN OVERVIEW OF THE ECONOMY OF LESOTHO
  - 2.2.1 Supply sectors
  - 2.2.2 Aggregate demand components
  - 2.2.3 Employment and wages
  - 2.2.4 Fiscal developments
  - 2.2.5 The monetary sector
  - 2.2.6 Price developments
  - 2.2.7 The balance of payments
- 2.3 SUMMARY OF MACROECONOMIC DEVELOPMENTS
- 2.4 THE CONDUCT OF MACROECONOMIC POLICY
  - 2.4.1 Monetary policy
  - 2.4.2 Fiscal policy
- 2.5 CONCLUSION
CHAPTER THEORETICAL BASE FOR MACROECONOMETRIC MODELLING
- 3.1 INTRODUCTION
- 3.2 BACKGROUND TO MACROECONOMETRIC MODELLING
- 3.3 PRINCIPLES OF STRUCTURAL MODELLING
- 3.4 MACROECONOMETRIC MODELLING IN THE CONTEXT OF DEVELOPING ECONOMIES
- 3.5 MACROECONOMIC MODELLING IN THE PRESENCE OF STRUCTURAL CHANGES
- 3.6 THEORETICAL GUIDELINES TO SECTORAL MODELLING IN THE CONTEXT OF MACROECONOMIC MODELS
  - 3.6.1 Theoretical models of aggregate supply
  - 3.6.2 Theoretical models of aggregate demand
  - 3.6.3 The government sector in a macroeconomic context
  - 3.6.4 Theoretical models of the external sector
  - 3.6.5 The labour market, prices and wage determination
  - 3.6.6 Theoretical models of the monetary sector
- 3.7 CONCLUSION
CHAPTER MACROECONOMETRIC MODELLING IN PRACTICE: A SURVEY OF EMPIRICAL LITERATURE
- 4.1 INTRODUCTION
- 4.2 A CRITICAL REVIEW OF LESOTHO MACROECONOMIC MODELS
  - 4.2.1 The Polak Model of the Central Bank of Lesotho
  - 4.2.2 The Central Bank of Lesotho Macroeconomic Model (CBLMM)
  - 4.2.3 The current macroeconomic model
- 4.3 OTHER RELATED LITERATURE
  - 4.3.1 The production sector
  - 4.3.2 The employment sector
  - 4.3.3 Modelling the level of economic activity in macroeconomic models
  - 4.3.4 The external sector in macroeconomic models
  - 4.3.5 The government sector
  - 4.3.6 Modelling the monetary sector in macroeconomic models
  - 4.3.7 Price and wage determination
- 4.4 THE TREATMENT OF EXPECTATIONS IN MACROECONOMIC MODELS
- 4.5 LOW QUALITY AND LIMITED DATA SETS AND STRUCTURAL CHANGES
- 4.6 CONCLUSION
  - 4.6.1 Overview of general literature
  - 4.6.2 Lesotho macroeconomic models

CHAPTER MODEL SPECIFICATION
- 5.1 INTRODUCTION
- 5.2 LIST OF VARIABLES
- 5.3 A SCHEMATIC VIEW OF THE MODEL
- 5.4 MODEL SPECIFICATION
  - 5.4.1 The real sector
  - 5.4.2 The external sector
  - 5.4.3 The government sector
  - 5.4.4 The monetary sector
- 5.5 ESTIMATION TECHNIQUES
- 5.5.1 Background to the methodology
  - 5.5.2 Tests for the order of integration of variables
  - 5.5.3 Cointegration analysis
  - 5.5.4 The Engle-Granger method
  - 5.5.5 The Johansen procedure
  - 5.5.6 Unit root and cointegration tests
  - 5.5.7 Diagnostic tests
  - 5.5.8 Forecast accuracy of endogenous variables
- 5.5.9 Issues relating to data availability and quality
- 5.6 CONCLUSION
CHAPTER ESTIMATION RESULTS OF INDIVIDUAL BEHAVIOURAL EQUATIONS
- 6.1 INTRODUCTION
- 6.2 LISTING OF BEHAVIOURAL EQUATIONS OF THE MODEL
  - 6.2.1 The real sector
  - 6.2.2 The external sector
  - 6.2.3 The government sector
  - 6.2.4 The monetary sector
- 6.3 STATISTICAL PROPERTIES OF INDIVIDUAL TIME SERIES
- 6.4 ESTIMATION RESULTS OF INDIVIDUAL EQUATIONS
  - 6.4.1 Estimation results of the real sector
  - 6.4.2 Estimation results of the external sector
  - 6.4.3 Estimation results of the government sector
  - 6.4.4 Estimation results of the monetary sector
- 6.5 CONCLUSION
CHAPTER MODEL SOLUTION AND EVALUATION
- 7.1 INTRODUCTION
- 7.2 TRACKING PERFORMANCE OF THE MODEL
- 7.2.1 In-sample tracking performance
- 7.2.2 Out-of-sample tracking performance of the model
- 7.3 EVALUATION OF THE FORECASTING PERFORMANCE OF THE MODEL
- 7.4 SUMMARY OF THE PERFORMANCE OF THE MODEL
CHAPTER POLICY SIMULATION EXPERIMENTS AND IMPACT ANALYSIS
- 8.1 INTRODUCTION
- 8.2 FISCAL POLICY SHOCKS
  - 8.2.1 An increase in government consumption expenditure by 10 per cent
  - 8.2.2 An increase in government investment expenditure by 10 per cent
- 8.3 MONETARY POLICY SHOCK
- 8.3.1 An increase in the nominal Treasury bill rate by 2 percentage points
- 8.4 EXTERNAL SHOCK
  - 8.4.1 A currency depreciation of 5 per cent
- 8.5 SUMMARY OF SIMULATION EXPERIMENTS
CHAPTER SUMMARY, CONCLUSIONS AND POLICY IMPLICATIONS
- 9.1 INTRODUCTION
- 9.2 SUMMARY OF THE STUDY
- 9.3 MAIN FINDINGS OF THE STUDY
- 9.4 CONCLUDING REMARKS
- 9.5 POLICY IMPLICATIONS
  - 9.6 AREAS FOR FURTHER RESEARCH
- REFERENCES
- APPENDIX A