THE COMPETITIVE SEMICONDUCTOR MANUFACTURING HUMAN RESOURCES PROJECT:

Second Interim Report
CSM-32
Clair Brown, Editor

4. Employee Experience, Pay, and Career Ladders

Vincent M. Valvano

4.1 Introduction
4.2 Employee Experience Levels
4.3 Wage Levels and Wage Paths for Operators, Technicians and Engineers
4.4 Incentive and Bonus Pay
4.5 Summary

4.1 Introduction

The value of an experienced and skilled workforce to a company is a key issue underlying current employment policy debates. Research on this question is limited and generally indirect but existing studies indicate that on-the-job training increases employee productivity. It is not clear, however, to what extent employees learn on the job solely as a function of experience, or in the absence of ongoing formal training, mentoring, or pay policies that encourage learning and skill acquisition. By convention, more experienced employees are assumed to have accumulated more on-the-job training and thereby to be more productive members of the workforce. If this is the case, we might expect companies to fashion pay policies and career ladders with the aim of retaining their more experienced employees, particularly in jobs where skill requirements are high and continuous learning is important, as in the semiconductor industry. We do not find a correlation between workforce experience levels and the existence of career pay policies among the fabs in our sample. This result is likely a function of both a small sample and the diversity of fabs in the sample. Our sample fabs are located in several countries with very different national economic institutions and labor market conditions, which exert significant influence on company pay policies. Our observations suggest that firms adapt their pay and training policies in response to such environmental forces and that fabs facing different labor market conditions can achieve similar performance levels with dissimilar workforce policies.

4.2 Employee Experience Levels

Eleven fabs in our sample provided information on the tenure of their workforce (Table 4-1). There is significant variation in employee experience levels in these fabs. Not surprisingly, older fabs with mature products and processes often had employees with long tenure. Operators with five or more years on the job comprised three-fourths or more of total operators employed at these fabs. Little or no hiring had occurred at these fabs in recent years.

Table 4-1. Selected Employee Characteristics for 11 Semiconductor Fabs

At the other extreme, a number of fabs operated with less-experienced work forces. In one fab, 70% of the operators had been employed for less than two years. At two other fabs, approximately 60% of operators had less than two years of experience. Two factors apparently accounted for the implicit high levels of turnover at these fabs. One fab operated in a local labor market that was exceedingly tight and where high levels of employee turnover were the norm. At two other fabs, the low experience levels of operators were largely driven by the characteristics of the national labor market, especially with respect to female labor supply. Young women operators lived and worked at these fabs for several years after graduation from high school but would leave the workforce when they married.

Seven of the ten fabs that provided data on employee demographics have majority-female operator work forces (Table 4-1). At three of these fabs, operators are completely or nearly all women. At the three fabs where men constitute a majority of the operator workforce, operator tenure is relatively high (50% or more of operators have at least five years of tenure). In contrast, only two of the seven fabs with majority-female operator work forces have comparably high levels of operator tenure.

In general, the engineering workforce in our fab sample had higher levels of fab tenure than operators (Table 4-1). The one fab reporting a low level of engineer tenure has been in operation for a relatively short period of time.

Acquisition of on-the-job training with experience is the primary way operators accumulate skills at the fabs in our sample. Nine out of ten responding fabs indicated that the minimum educational requirement for operators is a high school diploma or equivalent. One fab had recently instituted a technical AA degree requirement for newly hired operators. The great majority of operators in the work forces of our sample fabs possess high school education or the equivalent. Technicians generally hold an AA degree in electronics (or equivalent).

4.3 Wage Levels and Wage Paths for Operators, Technicians and Engineers

To make an approximate comparison of labor costs across fabs located in different regions, we collected information on base pay and incentive pay for each major occupation. Average, minimum, and maximum hourly pay by grade was requested for operators and technicians. Average, minimum, and maximum monthly pay by grade was collected for engineers. And average yearly bonus and incentive pay was collected for all three occupations. Wages and salaries that were reported in local currencies were converted to U.S. dollars based on the exchange rate for the year in which pay data were reported. Then wages and salaries were converted to constant 1994 dollars to control for changes in the price level and allow comparisons between fabs which reported pay data for different years.

The pay data can be used to estimate earnings progression over time for employees and thereby get an indication of the extent of career ladders in place for each occupation at the sample fabs. Extensive career ladders, in which significant pay increases are possible as an employee progresses through the grades associated with a particular job, can be a principle mechanism by which a fab tries to minimize costs associated with high levels of employee turnover. Their presence suggests that the fab may make significant investments in employee training and thereby have incentives to reward and retain its more productive experienced employees.

Pay ladders were constructed from the normalized wage and salary data. To construct ladders for operators and technicians (Figures 4-1, 4-2), the minimum pay level in the lowest grade was designated the entry wage, and the maximum pay level in the highest grade was designated the top wage in the ladder. Intermediate steps in the ladder correspond to the average wage reported for each grade. Accordingly, a fab that has two pay grades for its operators will be represented by a four-step ladder in Figure 4-1. Engineer pay ladders (Figure 4-3) were based only on the average salary level reported for each grade. The significant degree of earnings dispersion across fabs evident for all three occupations is not unexpected given that the sample fabs span six countries and three continents.

The pay ladders for operators vary in length and overall steepness among the sample fabs (Figure 4-1). Three fabs with long ladders (M6, A2, L16) have integrated ladders for operators and technicians. The higher grades in these ladders require technical skills and more complex job assignment, but operators are encouraged to acquire such skills and significant numbers of operators acquire technician-level skills over time. In fact, at two of these fabs, separate job categories for operators and technicians have been eliminated. In one of these fabs we observed a highly trained operator work force in place whereby many operators had the requisite skills that allowed them to perform a range of tasks that would be assigned to the equipment technician job category in other fabs. In the third fab with a combined job ladder, operators were encouraged by the company to get on-site and off-site training (a technical AA degree), and a significant number of operators were promoted to the technician job category. At the other fabs in the sample, the pay ladders for operators and technicians are distinct and the promotion of operators into technician jobs is infrequent.

The operator pay ladders for fabs M1 and M10 understate the actual pay levels in these fabs. A regular component of pay is a "bonus" paid monthly or bimonthly, which is a flat, non-varying percentage of monthly pay. Because this component of pay does not vary based on company or individual performance, it should be classified as base pay rather than incentive pay. Incomplete data prevents making this adjustment, although such pay may comprise up to 80% of regular pay at these fabs. This caveat applies to technician and engineer pay levels in these two fabs as well.

Career ladders for technicians are charted in Figure 4-2. Ladders of three or four grades are most common and most exhibit some steepness. One fab reported only one grade for its technicians and another fab reported six grades with a relatively flat salary progression.

The career ladders for engineers (Figure 4-3) tend to be longer and the salary paths are more steeply sloped relative to those for operators and technicians. Most fabs have at least four job grades in their career ladders for engineers, and one fab (L16) reported 16 salary grades (every other grade is plotted), with unusually steep wage progression in the highest grades of its wage path. A sizeable minority of fabs in this sample (5) had three or fewer grades with little wage progression for their engineers.

To measure theoretical wage progression in career ladders, the ratio of maximum pay in the top grade to minimum pay in the bottom grade is calculated for each occupation. These ratios are reported for each fab in Table 4-2. Theoretical wage progression is highest for engineers. However, the mean and median wage progression ratios for technicians are lower than those for operators. The fabs with the highest operator wage progression ratios are the three fabs with integrated operator-technician pay ladders. Fabs with relatively high wage progression ratios for operators tend to have high ratios for technicians (correlation of .56) and for engineers (correlation of .60).

Table 4-2. Top/Entry Wage Ratios, by Occupation, for 11 Semiconductor Fabs

Note: The wage ratio equals the maximum reported base pay for the top grade divided by the minimum base pay reported for the entry grade.

If a principal goal of career ladders is to retain experienced employees and encourage and reward skill development, we should observe positive correlations between workforce experience and the wage progression ratio. We do not observe such a relation for the fabs for which we can match employee experience with wage progression ratios. Correlations are near zero for operator wage ratios and operator experience, and for engineer wage ratios and experience. Because the small sample size makes it difficult to control for factors such as the age of a fab, such correlations are not easy to interpret.

4.4 Incentive and Bonus Pay

Comparisons of wage and salary levels in Figures 4-1, 4-2, and 4-3 can be misleading if the bonus and incentive component of compensation is a significant part of total compensation. This is the case for some of the fabs in our sample. Initial data on incentive compensation is restricted to three fabs. Several other fabs reported not having incentive pay. We did not collect information about stock or stock option grants and hence the value of such compensation is not included in our measure of incentive and bonus pay. Table 4-3 reports total annual incentive pay as a percentage of annual base wages or salary. One fab has a sizeable incentive component of pay, ranging from 15-25% of base pay. The other two fabs have much smaller levels of bonus/incentive pay, not exceeding 10% of base pay.

Table 4-3. Incentive and Bonus Pay as a Percentage of Base Pay for Four Semiconductor Fabs

Note: Ranges apply to the lowest and highest grades for a particular occupation.

4.5 Summary

1. There is significant variation in employee experience levels in our sample of semiconductor fabs. One-third of our sample fabs had experienced operator work forces such that at least half or more of operators had five or more years of tenure. At several fabs, however, a majority of operators had less than two years of tenure. The engineering work forces in our sample had higher levels of fab tenure than operators. Most fabs reported that a majority of their engineers had at least five years of tenure.

2. Majority-female operator work forces are common in semiconductor fabs and predominate in our sample.

3. We find a significant degree of wage-level dispersion across our sample fabs. This is expected given that the sample fabs span six countries and three continents.

4. Most fabs in our sample reported operator pay ladders with four grades and fairly flat wage paths across grades. Ladders for technicians exhibit more steepness than those for operators. Engineers have the most developed career ladders with longer and more steeply sloped pay paths, relative to those for operators and technicians.

End of Chapter 4

Go to Chapter 5
Go to Table of Contents for this Chapter
Go to Table of Contents for the CSM-HR Interim Report