By the 1980s, Taiwan had gradually transformed itself from an economy in which labor was plentiful, while capital was in short supply, into an economy in which a surplus of capital was accompanied by a shortage of labor. For labor economists interested in rapidly evolving labor markets, there are many intriguing questions arising out of this transition. This study explores how conventional labor market and efficiency wage theories apply during this transformation. Based on monthly data over the period 1982–2007, we examine the linear causality between pay and productivity using Geweke's [Geweke, J. (1982). Measurement of linear dependence and feedback between multiple time series. Journal of the American Statistical Association, 77, 304–313] linear feedback technique. The results indicate that both pay as reward and pay as incentive behaviors are significant in the Taiwanese economy, with pay as incentive especially strong in the manufacturing sector.
Standard economic theory predicts productivity growth should lead to higher wages. However, several variations of the efficiency wage theme have been put forward explaining why higher wages may lead to increases in productivity. Shapiro and Stiglitz (1984) suggest, by creating unemployment, above equilibrium wages will increase worker effort. Salop (1979) indicates that worker turnover is related to pay, and thus wages will rise above the market clearing level as firms try to reduce turnover costs. Akerlof (1984) provides several additional mechanisms, sociological in nature, where wage increases could improve worker performance. Additionally, higher wages may allow the firm to attract better employees (Stiglitz, 1976 and Weiss, 1980), or may directly impact worker productivity in poor countries by increasing nutrition (Bose, 1997).
A variety of empirical methods have been used to test for efficiency wages, with the resulting evidence being mixed. In this paper we use the Geweke (1982) method to test the association between wages and productivity in Taiwan, and find strong support for both directions of causality; productivity increases lead to wage increases, and wage increases lead to productivity increases. However, most of the increases in productivity occur immediately following the wage increase, suggesting that the effect may diminish rather quickly. The results also indicate that the connection between wages and productivity has become more pronounced in Taiwan during the last decade.
The remainder of the paper is organized as follows: Section 2 provides a summary of previous studies on efficiency wages. Section 3 provides a brief description of the evolution of labor market conditions in Taiwan. In Section 4, we present the empirical methodology and model specification. Section 5 describes the data and presents the results of the analysis. Summary and concluding remarks follow in the final section.
As wage setting behavior differs across countries, it seems likely that the interaction between pay and productivity will differ as well. For instance, Mahlau and Lindenberg (2003) suggest that U.S. firms use wages as a signaling device, whereas Japanese firms do not. Franz and Pfieffer (2006) conclude that, relative to U.S. firms, German firms attach more importance to specific human capital and labor union contracts. Furthermore, cross-country differences in unionization rates could impact the relative importance of efficiency wage behavior, as efficiency wage motivations should be strongest in non-union firms (Straka, 1989).
These differences, combined with statistical issues associated with testing efficiency wages, would likely lead to significantly different results depending on the country studied or method applied. Indeed, using a variety of techniques, wage and productivity changes consistent with efficiency wage behavior have been suggested for countries such as Brazil (Arbache, 2001), Canada (Gera & Grenier, 1994), China (Fleisher & Wang, 2001), and Zimbabwe (Velenchik, 1997).
This paper studies the relationship between pay and productivity in Taiwan using a technique that disentangles the potential bi-directional causality between pay and performance. Our results strongly support the hypothesis that pay increases productivity in Taiwan, evidence suggestive of efficiency wage behavior. Fuess and Millea (2002) and Millea and Fuess (2005) find limited support for this behavior when they apply the same technique to U.S. and Japanese data. The differing results may be due to the nature of the data sets, as monthly data, which is used in our study, may be more likely to capture the motivational effects of a wage increase. Indeed, the evidence presented in this paper suggests that most of the productivity impact occurs immediately following the increase in pay.
Alternatively, perhaps wage-setting differences explain why pay has a relatively stronger, and more consistent, effect on productivity in Taiwan as compared to the U.S. or Japan. Since the late 1980s, Taiwan has risen to be one of the newly industrialized economies, and has subsequently faced a structural transformation in its labor market. This transformation has led to a shortage of skilled labor, while at the same time the number of unemployed workers increased. The evidence from the measures of directional feedback indicates that the Taiwanese manufacturing sector, which has been significantly impacted by this transformation, may have altered its pay structure to draw sufficient workers and maintain productivity growth. Nevertheless, in all sectors of the Taiwanese economy, higher pay appears to motivate workers to increase productivity.
