Do Higher Wages Reduce Knowledge Worker's Job Mobility? Evidence for Swedish Inventors

• Author: Olof Ejermo,Torben Schubert
• DOI: http://doi.org/10.1111/joms.12317
• Published date: 01 January 2018
• Date: 01 January 2018

Do Higher Wages Reduce Knowledge Worker’s Job

Mobility? Evidence for Swedish Inventors

Olof Ejermo and Torben Schubert

Lund University; Fraunhofer Institute for Systems and Innovation Research ISI

ABSTRACT Based on linked employer-employee panel data on all Swedish inventors, this paper

analyses how wages affect inventors’ job mobility. It is commonly assumed that higher wages

reduce mobility because they reduce the value of outside opportunities. We argue that higher

wages also send performance signals to potential employers, who raise their wage offers in

response. By disentangling the effects of higher wages, we show evidence of a utility and an

opportunity cost effect, which reduce mobility, and a performance-signalling effect, which

increases mobility. In our data, the effects cancel each other out, with no effects of wages on

mobility rates on average. We find, however, that for star inventors, who have sufficiently

strong alternative performance signals (e.g., strong patent records), the performance signal sent

by wages is crowded out by the alternative signals. Accordingly, for star inventors we find that

higher wages decrease mobility.

Keywords: inventors, mobility, signalling, wages

INTRODUCTION

Losing key employees to competitors is an important threat to firms because the firm’s

capabilities and competencies are weakened (see Adner and Helfat, 2003; Shaw et al.,

2013; Trevor, 2001). Analysing the mechanisms that can help prevent the loss of key

employees has therefore become a central topic in strategic human capital management.

An important strand of the literature has discussed how wages affect the incentives for

mobility (Bloom and Michel, 2002; Blyler and Coff, 2003; Carnahan et al., 2012). This

literature seems to take for granted that higher wages make employees less inclined to

switch jobs because employees derive utility from wages. Although this assumption is

intuitive, empirical accounts of the wage-mobility relationship are less clear cut. Some

authors indeed find that higher wages have the expected mobility-reducing effect (Falch,

2011; Hammida, 2004). Yet, others find a mobility-increasing effect (Kim, 1999; Le

Address for reprints: Olof Ejermo, CIRCLE, Lund University, P.O. Box 117, SE-22100 Lund, Sweden

(olof.ejermo@circle.lu.se).

V

C2017 John Wiley & Sons Ltd and Society for the Advancement of Management Studies

Journal of Management Studies 55:1 January 2018

doi: 10.1111/joms.12317

Grand and Tåhlin, 2002; Pfeifer and Schneck, 2012) – a finding which is inconsistent

with the standard argument that higher wages reduce incentives for mobility.

We trace the failure of the standard argument to account for mobility-increasing

effects of higher wages to the implicit assumption that a rise in current wages has no

effect on the wages associated with outside job opportunities. Although the assumption

that current wages do not affect outside wages may seem innocuous at first, it can fail if

outside employees interpret higher wages as signals of higher performance (see Becchetti

et al., 2013; Golan, 2009). In response to presumed higher performance, prospective

employers may raise their wages making mobility more attractive. Thus higher wages

do not necessarily contribute to retaining workers in the firm (compare Allen et al.,

2010). As a consequence, paying higher wages may not only be costly but also ineffective

as a retention mechanism in strategic human capital management (among others Carna-

han et al., 2012; Gardner, 2005).

We test our arguments based on a linked employer-employee dataset for the popula-

tion of Swedish inventors between 1990 and 2007. We use inventors in firms as a spe-

cific case because inventors are employees of considerable strategic value to the firm.

Furthermore, inventors differ widely in terms of individual capability (Marx et al.,

2009), which makes issues of performance signalling particularly relevant. To identify

the effects, we employ the Swedish accession to the European Union (EU) in 1995 as a

natural experiment. Our results show the existence of three structural effects of higher

wages. On the one hand, higher wages lead to lower mobility rates because of the higher

value of the current occupation (utility effect) and the relative decrease in the value of

the outside alternatives (opportunity cost effect). On the other hand, we find that wages

increase mobility through a performance signalling effect. The results show that the

three effects cancel one another out for the average inventor in our sample. We con-

clude that higher wages do usually not reduce inventor mobility. Wages reduce mobility

only for the sub-group of high performing ‘star’ inventors. We provide evidence that

high performers signal their performance by alternative measures (e.g., valuable patents

in their track record). Strong alternative signals can completely crowd out the perform-

ance signal of higher wages.

BACKGROUND

Theoretical Framework

Many authors take for granted that higher wages reduce mobility. For example, Carna-

han et al. (2012), Gardner (2005), and Harris and Helfat (1998) argue that firms com-

pete by attracting and retaining key personnel, which is also determined by the firm’s

ability to pay higher wages. Underlying the expectation that higher wages lead to lower

mobility is the implicit assumption that at any point in time employees face a fixed set of

job alternatives, which they rank according to their preferences. If the current alterna-

tive is considered best, employees will stay. Otherwise, they will switch jobs. Because

employees are expected to prefer higher wages over lower wages, receiving a high cur-

rent wage will disincentivize mobility.

109Do Higher Wages Reduce Job Mobility?

V

C2017 John Wiley & Sons Ltd and Society for the Advancement of Management Studies

Some empirical works lend support to the reasoning that higher current wages reduce

mobility by showing that high earners are less mobile than low-income groups (Clotfelter

et al., 2008; Falch, 2011; Hammida, 2004). However, other studies find the opposite rela-

tionship (Kim, 1999; Le Grand and Tåhlin, 2002; Pfeifer and Schneck, 2012), which high-

lights that some assumptions of the framework, in which employees choose the alternative

that maximizes their utility from a static set, may not be correct. We argue that the expec-

tation that wages reduce mobility critically rests on the assumption that the set of available

choices is exogenously determined and, in particular, independent of current wages.

Although this wage-exogeneity assumption may sometimes be reasonable, we discuss con-

ditions that can lead to its failure. We argue that failure in the exogeneity assumption can

imply that higher wages are associated with higher rather than lower mobility rates.

Effects of Wages on Mobility under Exogeneity of the Set of Alternatives

If outside wages are exogenously given, standard arguments about rational behaviour

imply that higher current wages reduce mobility, because employees prefer higher over

lower wages. Two types of arguments have been put forward. First, scholars have

argued that wages push the employees’ budget constraint outward. Thus, higher utility

results from increased consumption possibilities (see Akerlof, 1982; Pfeifer and Schneck,

2012; Salop, 1979; Salop and Salop, 1976). Furthermore, authors have argued that

higher wages do not only increase consumption possibilities but also create higher wage

gaps to the employee’s peer, implying a status effect (Clark and Oswald, 1996; Clark

and Senik, 2010; Wolbring et al., 2013). Irrespective of whether employees prefer higher

over lower wages because of higher consumption possibilities or because of status gains,

wages are expected to decrease mobility because they increase the employee’s utility.

We posit our first hypothesis:

Hypothesis 1 (utility effect): The higher the wage, the less likely the inventor is to leave

the firm.

Hypothesis 1 expresses the idea that employees stay in their jobs when the current job

alternative ranks highest. Ranking the current job alternative highest is more likely if the

current firm pays better than the average potential outside employers. If the wages asso-

ciated with outside opportunities are low compared with the current wage, the employee

faces low opportunity costs associated with staying and will thus be less likely to switch

jobs (Christensen et al., 2005). In practice, the opportunity costs can be sizable. A prime

source of opportunity costs are large and persistent wage differentials between firms

(Helpman et al., 2012) implying that even for similar tasks/occupations wages can differ

substantially between firms. Accordingly, wage differentials offer room for employees to

increase their wage by moving to better paying firms (Frederiksen et al., 2016; Keith

and McWilliams, 1999; Yankow, 2003). Hence we propose our second hypothesis:

Hypothesis 2 (Inverse opportunity cost effect): The higher the average wage level at the

current employer relative to the average wage level in the industry, the less likely

the inventor is to leave the firm.

110 O. Ejermo and T. Schubert

V

C2017 John Wiley & Sons Ltd and Society for the Advancement of Management Studies