A stable demand for money function is a necessary condition for the supply of money to be utilized as an instrument of monetary policy (Serletis, 2001). Poole (1970) showed that the rate of interest (supply of money) should be used as an instrument of monetary policy when the money demand function is unstable (stable). Due to financial reforms since the 1980s, it is widely recognized that the condition for a stable demand for money relation has broken down (Orden and Fisher, 1993). Evidence suggests that the switch from reliance on the money supply to the interest rate as the primary monetary policy instrument has taken place in advanced OECD countries because of an instability in money demand functions. Following this change many developing countries have also moved towards using interest rates without significant evidence that their money demand functions have become unstable.

The purpose of this study is to re-assess the stability of M1 (narrow money) demand for selected advanced OECD and developing countries using more up-to-date econometric techniques and data. This allows us to re-evaluate the contemporary stance on monetary policy because it appears that most central banks of developing countries could still utilise the supply of money as an instrument of monetary policy. According to Poole (1970), selecting the wrong instrument may result in large fluctuations in output. He suggested that money demand stability information helps to serve the central banks to formulate an optimal monetary policy. This study emphasizes the importance of monitoring demand for money in monetary policy decisions.

Money demand stability is an empirical issue and given the importance of the topic, robust empirical results are required. Contradictory evidence exists on the income elasticity estimates of money demand and their stability. Regrettably, the existing literature does not offer guidelines for identifying strong empirical evidence given the multitude of econometric methods that are on offer. Given the emphasis placed on the interest rate it appears that most central banks now pay less attention to the stability of the money demand functions. This study establishes that money demand stability is useful for policy but it requires strong empirical investigations. To this end, utilising a wide range of new empirical tests, this study presents estimates of the canonical and extended specifications of the demand for M1 for selected advanced OECD and developing countries.

This thesis offers a structured, logical and detailed way of navigating through the empirical issues and challenges that have emerged in the literature. The first major contribution to the literature is that a wide array of recently developed time series and panel data techniques are used to examine the M1 demand relationships for a large sample of countries (10 advanced OECD, 5 Pacific Islands, 18 Asian, 10 African and 10 Latin American). Results show that the income elasticity is less (around or slightly higher) than unity in advanced OECD (developing) countries; contrary to what is achieved by most studies in the literature. To this end, financial markets seem to be relatively well developed in the advanced OECD countries.

The second major contribution of this study is that a comprehensive set of unit root and cointegration tests are organised in the form of flowcharts to illustrate the possible instances in which they could be utilised. This will serve as a useful guide to applied economists who are working with non-stationary time series and panel data. This study also applies the flowcharts to provide a comprehensive body of examples and cross-checks of the results for specific countries and regions using the appropriate methodologies. This thesis therefore serve as an examination of other’s studies while also offering new results and establishing the degree of importance of following the correct technique.

The third major contribution of this study is associated with the process of monetary policy procedures. The stability tests on M1 demand imply that the developing countries should re-consider their choice of using the rate of interest as an instrument of monetary policy; perhaps using the supply of money is a feasible alternative. It appears that most developing countries are imitating the monetary policy procedures of the advanced OECD countries. It is pragmatic for the advanced OECD countries, especially Switzerland and the USA, to utilise the rate of interest policy because their money demand functions are found to be temporally unstable.