Since the early 1990s, discussions on total factor productivity (TFP) have become increasingly important for Singapore, and there has been substantial amounts of empirical work done in this area using both cross-country and inter-temporal analysis. This article reviews the conceptual and empirical aspects of some of the TFP growth studies on Singapore and provides suggestions for future research. Introduction Total factor productivity serves as an important measure of the productive performance of an economy for the following reasons. First, unlike the partial productive measure, it considers the contribution of more than one input to output. Second, it is important to the growth process in the long run, as there are constraints imposed by population growth, together with diminishing returns that set in as capital intensity is increased. One way to secure economic growth beyond these limits is to secure ongoing increases in TFP. This is a very relevant issue for Singapore, which the OECD has upgraded to the status of an advanced newly developing country. Although Singapore has enjoyed impressive levels of economic growth, in terms of sustainable long-term growth given by TFP, it was Young (1992) who first argued that Singapore was nowhere near its twin city, Hong Kong. This was then emphasized by Krugman (1994) who singled out Singapore as the only newly-industrialized economy (NIE) which experienced no TFP growth. But Lim (1986, p. 5) once commented that, if a country can raise its standard of living so spectacularly with a very low TFP growth, does it then matter whether TFP growth is low?' Peebles and Wilson (1996, p. 205) as a reply to Lim, argued that an important point is missed in this comment which only looks at the benefits resulting from high growth and ignores the cost of achieving such growth. This means that TFP analysis in Singapore is not to be taken lightly. It was noted that most developed countries at the same development stage as Singapore were having TFP growth rates between 2 per cent and 4 per cent while Singapore registered an insignificant 0.4 per cent between 1980 and 1992. The target is now set to reach at least 2 per cent in order to sustain a (labour) productivity growth of 4 per cent and a GDP growth of 7 per cent.2 The importance of TFP in Singapore is further reflected in the numerous studies undertaken to examine the issue. Review of TFP Studies Table 1 provides a summary of various studies' estimates of TFP growth for Singapore's aggregate economy, manufacturing (Manu) and services (Serv). Most of the studies above show that TFP growth in Singapore has been insignificant, particularly in comparison with the NIEs and other countries. After the pioneering work of Tsao (1982) on Singapore, various studies have reexamined this issue. While Wong and Tok (1994) and Rao and Lee (1995) showed that TFP growth increased in the latter half of the 1980s, the National Productivity Board (1994) and Sarel (1997) showed improvements in TFP growth for the early 1990s. However, the Department of Statistics (1997) and Renuka M. (1998) show otherwise and Leung's (1997) results are inconclusive for the early 1990s. It is obvious that TFP results can vary significantly from one another and this is due to different methodologies used and different time periods of study. The objective of this article is to critically analyse some of the often-cited cross-country and inter-temporal TFP studies in order to understand the problems that still exist in the analysis of TFP growth. In particular, the conceptual and empirical aspects of the studies are reviewed and suggestions for future research are put forth. Unfortunately, the space constraint and the disparate and complex nature of the various studies reviewed are not amenable to a more integrated approach to the article.3 For instance, a detailed comparison of various methodologies in TFP measure and their caveats as well as data problems in Singapore could not be discussed extensively but an attempt was made to incorporate these issues within the studies whenever possible. …
11
