PUBLICATIONS & REPORTS

Contents Introduction

This paper investigates the effects of foreign worker shares and MNE ownership on wages after controlling for worker sex and occupation in Malaysian manufacturing plants during 1994-1996, an important period during which use of foreign workers began to increase substantially. In a previous paper, I estimated similar wage equations separately for five occupation groups of both sexes in large heterogeneous samples of plants in many industries and more homogeneous samples of plants in seven industries. Results indicated that use of foreign workers generally had insignificant effects on plant wages for most occupation-sex-(and industry) combinations and that that MNE-local differentials were almost always insignificant in three industries and consistently significant in only one. Although separate estimation by sex and occupation has the strong advantage of accounting for worker characteristics relatively well, it has the disadvantages of complexity (10 results per sample) and being difficult to compare to more common approaches, which use sex and occupation as controls. The primary purpose of this paper is thus to see if using sex and occupation as independent variables generates results that differ from estimating wage equations separately for each sex-occupation cohort. Results suggest that the effects of foreign worker shares differ substantially among foreign worker occupations and among industries. Plants that have relatively large foreign manager shares tend to pay relatively high wages in most industries, but the effects of other foreign worker occupations are usually insignificant or inconsistent. Results that assume all foreign workers impart the same effects thus appear misleading, as do results assuming identical slope coefficients among industries. Similar to previous estimates, MNE-local wage differentials were consistently positive and significant in only two relatively small industries, chemicals and food, in marked contrast to previous results for 2000-2004, which did not account for the effects of foreign worker shares.