林明仁 Ming-Jen Lin​

This paper adds in a new assumption, ‘product price’, into the tournament model and shows the convexity of pay structure should become stronger when the produce price increases. Empirical results from a Taiwanese textile company are consistent with the implication.

Empirical evidence for a relationship between democracy and crime has not been investigated. This paper combines several international data sets to show that, compared to non-democratic governments, democratic governments punish major (minor) crime more (less) severely and hence this crime rate is lower (higher). Namely, the effect of democracy on crime is negative for serious crime such as murder and positive for minor crime such as theft. This empirical result is robust across different data sets and estimation methods. A 2SLS method is also used to solve the endogeneity problem embedded in OLS. In addition, GDP has a negative effect on all crime rates. Furthermore, inequality and unemployment increase, while clearance rate decreases, crime. Evidence also suggests that democracy per se rather than transitional or political instability causes the relationship. Using prison sentence length, clearance rates, and prisoners per crime as measures of deterrence, it was found that deterrence can explain approximately 40% of the effect of democracy on crime.

This research note combines two national Taiwanese data sets to investigate the relationships among low birth weight (LBW) babies, their parents' educational levels, and their future academic outcomes. We find that LBW is negatively correlated with the probability of such children attending college at age 18; however, when both parents are college or high school graduates, such negative effects may be partially offset. We also show that discrimination against daughters occurs, but only for daughters who were LBW babies. Moreover, high parental education can buffer the LBW shock only among moderately LBW children (as compared with very LBW children) and full-term LBW children (as compared with preterm LBW children).

This paper analyzes the personnel records of a Taiwanese auto dealer employing three distinct internal labor markets (ILMs), adding new evidence that builds upon recent empirical and theoretical works on ILMs. We find that the career mobility is different amongst different workers. The positive effects of levels, on both salary and bonus equations, are smaller under a fixed effects model than under an OLS (combined) model. However, part of the wage variations is contributed by individual heterogenity rather than the hierarchy itself. Evidence shows that education plays a much more important role in the determination of levels for staff workers (amongst whom the effective determination of output is difficult), than for salespersons and technicians (where output is much more easily measured). We hence argue that education serves as a mechanism to sort workers by ability. We also show that the public learning model proposed by Farber and Gibbons (1996) is not supported in general by our data. In addition, we find strong support of a serial correlation for salary, but not for a bonus. The overall picture is that ILM structures exist in different cultures, and their practice is different amongst workers who have different characteristics.

This paper sets out to analyze an internal data set on a Taiwanese auto dealer employing three distinct types of workers. The effects of jobs and levels are positive on both the salary and bonus equations, albeit smaller under a fixed effects than under OLS; however, when factoring in individual fixed effects, the reductions in the bonus equations are greater than those in the salary equations. With changing economic conditions, any consequent variations are greater in bonuses than in salaries, with the most extreme variations being felt by higher ranking employees than lower‐level workers. Promotion premiums between levels are smaller than the average differences in pay, and although wage variations do exist within and between levels, the greater effect is on bonuses rather than salaries. The variations in both salaries and bonuses, defined by the coeffficient variations, are also greater in those years when demand is high, as opposed to years of low demand. Entry and exit behavior is observed at all levels, although it is more likely to occur among the lower levels of the hierarchy. Finally, we present strong evidence in support of the cohort effect. Overall, our findings confirm the prevalence of internal labor market (ILM) theories.