AbstractThe relationship between government expenditure and gross domestic product (GDP) has been subject to extensive research both in the field of public finance and macroeconomic modelling. More than one hundred years ago Adolph Wagner proposed a positive correlation between the level of gross domestic product growth and public spending. In this study six versions of Wagner's law were empirically tested employing aggregate and disaggregated annual time series data for the Libyan economy covering the period 1962-2005.
This thesis investigated the relationship between government expenditure and gross domestic product growth, in terms of total GDP and non-oil GDP. Engle and Granger's two-step cointegration analysis has been used to test the long-run relationship between government expenditure and total real GDP for Libya, whereas the short-run relationship is estimated using the error correction model (ECM). The causation between government expenditure and GDP growth is examined using the Granger causality test.
It was found that public expenditure and GDP variables in all six versions of Wagner's law are non-stationary in levels, but stationary in first differences, that is, they are integrated of order one 1(1), in terms of total GDP and non-oil GDP and the six categories also. The cointegration tests indicated that there is mixed evidence of a long-run relationship between government expenditure and gross domestic product in terms of total GDP and non-oil GDP using aggregate data. Furthermore, a long-run equilibrium relationship with disaggregate data is also established. The results suggest mixed evidence in support of Wagner’s law for the period under review.
The results from the ECM equations reconfirmed the validity of Wagner’s law in the short-run for total real GDP and government expenditure, as well as the short-run relationship with disaggregate data. Also, the results indicate that the short-run relationship between government spending and total real non-oil GDP does not exist for the period under review, with the exceptions of versions two and three where dummies were used.
Finally, the study used Granger causality testing procedure to determine the direction of causality. The results provide some evidence of a unidirectional causation running from gross domestic product to government expenditure in total real GDP, and mixed results with total real non-oil GDP. Also, this study has made contribution to knowledge. Specifically, it fills the gap in the public finance area of Libyan growth studies by testing Wagner's law on the Libyan economy. Also, this study has used the long-run and short-run relationship between government expenditure and total gross domestic product with GDP and non-oil GDP, as well as undertaking a causality analysis between the relevant variables.
|Date of Award||Mar 2010|
|Supervisor||Peter Romilly (Supervisor)|