The demand for money, asset substitution and the inflation tax in a liberalizing economy : an econometric analysis for Kenya
This thesis develops empirical econometric models of the private sector aggregate demand for real and financial assets in Kenya over the period 1973 to 1990. Single-equation error-correction models of the demand for money are estimated using systems cointegration methods developed by Johansen (1988). The models are found to be statistically stable functions throughout the period, and are capable of encompassing existing studies. Across a range of monetary aggregates, including a Divisia index aggregate for broad money, the models describe demand for money functions in which inflation and illegal foreign currency substitution are significant determinants of money holdings, and where the private sector adjusts rapidly to deviations from its stable longrun equilibrium real money demand. The demand for money is then integrated within a neo-classical model of asset demands, which examines the behaviour of the aggregate private sector asset portfolio in response to changes in relative prices between assets and to external shocks to the economy, principally the 1976-77 coffee boom. A variant of the Almost Ideal Demand System model developed by Deaton and Muellbauer (1980) is estimated for a class of six assets: base money, banking system deposits, government securities, tradable capital, nontradable capital and inventories. The asset substitution model, which also takes an errorcorrection form, and which allows for credit rationing, generates results which are consistent with the earlier demand for money models, where private agents are also denied access to foreign-denominated assets. Using this model, the maintenance of policies of financial repression are shown to cause the private sector to offset inflationary shocks through the accumulation of real assets, principally in the form of non-tradable capital in the construction and property sectors. The evidence from the two models is used to analyze the fiscal effects of the inflation tax and financial repression measures. Policies of financial liberalization are shown to reduce the revenue maximizing rate of inflation (estimated to be 14% per annum) and the implicit tax on domestic holders of government liabilities. This dampens asset substitution in response to inflationary shocks and offsets the adverse effects of "construction-boom" investment on non-tradable capital prices.