This study sets out to investigate the relationships that exist among key macroeconomic variables in Fiji. The purpose of the investigation is to ascertain whether, given the interrelationships that do exist among the key macroeconomic variables, the existing monetary policy setting is appropriate to this small semiopen economy in its endeavours to stabilise fluctuations in output and the price level. A semiopen economy is defined as one where restrictions exist on the capital account of the balance of payments. These restrictions imply that changes in domestic interest rates can differ in the short run from changes in international interest rates.
Fiji maintains an anchor on its nominal effective exchange rate which is tied to a basket of currencies comprising the currencies of the major trading partners. These trading partners have credible anti-inflationary strategies. Apart from maintaining a nominal exchange rate anchor, authorities also pursue an active monetary policy and respond to inflationary pressures that arise from within the Fijian economy. As such, the Reserve Bank of Fiji is actively engaged in following simultaneously an exchange rate policy and monetary policy. The use of monetary policy necessitates activities by the Reserve Bank of Fiji to manage the level of liquidity in the financial sector.
This thesis addresses a number of specific questions. Is the policy of setting an anchor on the nominal exchange rate appropriate for Fiji? Are there any arguments in favour of making the nominal exchange rate more flexible? Is the use of monetary policy appropriate? Should the Reserve Bank sterilise the monetary effects of the balance of payments? What factors determine the supply of and demand for money in Fiji? What is the role of interest rates in Fiji?
To answer these questions, a general equilibrium macroeconomic model that captures the characteristics of the Fijian economy is formulated and then estimated. The model is basically a Mundell-Fleming type of model but includes the supply side and models policy response explicitly. It is assumed that prices and interest rates are sticky in the short run, foreign assets are imperfect substitutes for domestic assets, and capital is imperfectly mobile.
The model is estimated using the Cowles Commission approach using 3SLS and FIML estimates. The time series modelling strategy has not been used. In recent times, it has been demonstrated that in the context of structural equation modelling, nonstationarity does not necessarily call for a modelling strategy different from the Cowles Commission approach. It has been demonstrated that prior information, both theoretical and empirical, are more important in a nonstationary framework than a stationary one for identification on the one hand and for ensuring that the parameters possess desirable statistical properties on the other.
The estimates of the model for Fiji reveal that there is a divide between the real and monetary sectors: the rate of interest has no effects on real demand or supply. Real demand and supply are determined almost exclusively by foreign prices and the exchange rate anchor. This raises doubts about the effectiveness of monetary policy and about the rationale behind liquidity management. With regard to sterilisation policies aimed at managing liquidity, the estimates reveal that the monetary effect of the balance of payments does not conform to the traditional assumption that the monetary effects are realised only when there is intervention in the foreign exchange market. The estimates demonstrate that an inflow of foreign exchange sees the money multiplier work through the economy twice: once when the foreign currency initially enters the country and once when the Reserve Bank of Fiji intervenes. This has implications for the sterilisation policies.