Academic journal article Journal of Southeast Asian Economies

Trends in Output and Income Disparities across the States of Malaysia, 1972-2009

Academic journal article Journal of Southeast Asian Economies

Trends in Output and Income Disparities across the States of Malaysia, 1972-2009

Article excerpt

I. Introduction

Since at least 1966, successive Malaysian governments have committed to balanced regional development across the country's various states and federal territories, and have taken actions as outlined in a series of five-year economic plans towards that objective. Yet, as recently as 2006, the Ninth Malaysia Plan assessed that "despite all states recording economic growth, the development gaps between regions, states and rural-urban areas remained wide". The Plan also promised measures to "accelerate the development of less developed states" and "attain regional balance and reduce development gaps" (EPU 2006, p. 371).

Patterns and trends in income distribution, growth, and development in Malaysia have long been studied by authors such as Snodgrass (1980), Young, Bussink and Hasan (1980), Anand (1983), as well as Ishak and Ragayah (1990), to name a few. More recently, Ragayah (2008) reported that overall income inequality (based on monthly household income) increased after 1990, reversing the downward trend exhibited during the 1976-90 period. Similarly, using location quotient analysis of GDP data at the state level, Ali and Ahmad (2009) found that interstate development gaps remained wide during the 1970-2006 period, and attributed these mainly to the fact that economic activity tended to be more concentrated in agriculture in the lagging states than in the leading states.

By contrast, Habibullah and Sivabalasingam (2008) reported finding that interstate disparities were reduced during the 1960-2003 period, with their conclusion being based on panel unit root tests of relative GDP per capita. A similar approach was adopted by Habibullah, Smith and Dayang-Affizzah (2008) to analyse the growth performance of Kelantan relative to the other states; these authors reached conclusions that were consistent with those of Habibullah and Sivabalasingam (2008).

Given the previous contradictory findings regarding interstate disparities, as well as the differences between the methods used and the main variables of interest, this paper seeks to utilize several sets of data and to apply to them several methods of analysis, with a view to reconcile (where possible) and update previous findings. Our main research question is: have the states of Malaysia converged or diverged in terms of average output and income?

We employ the log-t test developed by Phillips and Sul (hereafter, PS) (2007, 2009) to test for convergence across all states as a group, as well as for the possibility of sub-groups ("clubs") which may diverge from each other but within each of which members tend to converge. Results obtained via this method are compared and combined with those relating to univariate stationarity (KPSS) tests, and to a simple measure of dispersion (the coefficient of variation) which is often used to detect [sigma]-convergence. Where available, we analyse data for household income as well as GDP per capita. The aim is to derive conclusions that are reasonably robust and consistent with one another.

The paper is organized as follows. Section II will describe the methods of analysis and the data sources. Section III presents our analysis of the trends and patterns in interstate disparities based on the GDP per capita data available. Section IV provides the analysis based on household income data. Although a thorough examination of the underlying causes of such trends and patterns is beyond the scope of this paper, section V offers a number of conjectures regarding possible determinants. Finally, section VI concludes with a summary of the main findings and some reflections regarding future research directions.

II. Methods and Data

To test for [sigma]-convergence, we use the scale-free coefficient of variation (CV):

[CV.sub.t] = [[sigma].sub.t]/[m.sub.t] (1)

where [m.sub.t] is the sample mean across all relevant states, and [[sigma].sub.t] is the sample standard deviation at time t. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.