Nationality diversity and leader–Member exchange at multiple levels of analysis. Test of a cross-level model

• Pages: 20-39
• Date: 11 February 2019
• DOI: https://doi.org/10.1108/EDI-03-2018-0054
• Published date: 11 February 2019
• Author: Shaun Pichler,Beth Livingston,Andrew Yu,Arup Varma,Pawan Budhwar,Arti Shukla
• Subject Matter: Hr & organizational behaviour

Nationality diversity and

leader–Member exchange at

multiple levels of analysis

Test of a cross-level model

Shaun Pichler

Department of Management, California State University,

Fullerton, California, USA

Beth Livingston

Department of Management, Henry B. Tippie College of Business,

University of Iowa, Iowa City, Iowa, USA

Andrew Yu

Department of Management, Eli Broad College of Business,

Michigan State University, East Lansing, Michigan, USA

Arup Varma

Loyola University Chicago, Chicago, Illinois, USA

Pawan Budhwar

Aston Business School, Birmingham, UK, and

Arti Shukla

Imdaad LLC, Dubai, United Arab Emirates

Abstract

Purpose –The diversity literature has yet to investigate relationships between diversity and leader–member

exchanges (LMX) at multiple levels of analysis. The purpose of this paper is to test a multilevel model of

nationality diversity and LMX. In doing so, the authors investigate the role of surface- and deep-level

diversity as related to leader–member exchange differentiation (LMXD) and relative LMX (RLMX), and hence

to subordinate job performance.

Design/methodology/approach –The authors test a multilevel model of diversity and LMX using

multisource survey data from subordinates nesting within supervisors. The authors do so in a context where

diversity in nationality is pervasive and plays a key role in LMXs, i.e., a multinational organizationin Dubai.

The authors tested the cross-level moderated model using MPlus.

Findings –The results suggestsurface-level similarityis more important to RLMX than deep-levelsimilarity.

The relationshipbetween surface-level similarityand RLMX is moderated by workgroupnationality diversity.

When workgroupsare more diverse, there is a positive relationshipbetween dyadic nationality similarityand

RLMX; when workgroups are less diverse, similarity in nationality matters less. Moreover, LMXD at the

workgroup level moderates the relationship between RLMXand performance at the individual level.

Originality/value –This study is one of very few to examine both diversity and LMX at multiple levels of

analysis. This is the first study to test the workgroup diversity as a cross-level moderator of the relationship

between deep-level similarity and LMX. The results challenge the prevailing notion that that deep-level

similarity is more strongly related to LMX than surface-level diversity.

Keywords Diversity, Nationalcultures, Relative leader–member exchange,

Leader–member exchange differentiation, Ethnic minorities, Social identity theory, Leader–member exchange

Paper type Research paper

We inhabit a universe that is characterized by diversity. (Desmond Tutu)

When it comes to diversity, one thing is clear that there are inconsistent findings

as to the effects of diversity in workgroups (van Knippenberg and Schippers, 2007;

Equality, Diversity and Inclusion:

An International Journal

Vol. 38 No. 1, 2019

pp. 20-39

© Emerald PublishingLimited

2040-7149

DOI 10.1108/EDI-03-2018-0054

Received 16 March 2018

Revised 7 June 2018

Accepted 20 July 2018

The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/2040-7149.htm

20

EDI

38,1

Williams and O’Reilly, 1998). Scholars have called for more complex models of the

processes through which diversity can influence various outcomes (van Knippenberg and

Schippers, 2007), such as leader–member exchanges (LMXs) (Goldberg and McKay, 2015).

Workgroup diversity may be centrally important to understanding how LMX

relationships influence workgroups (e.g. Nishii and Mayer, 2009), perhaps especially in

international and multinational settings (e.g. Varma et al., 2002; Allen et al.,2018).Thereis

a dearth of research considering both diversity and LMXs at multiple levels of analysis –

despite the fact that LMX theory is multilevel in nature. The over-arching purpose of the

current study is to develop a multilevel model of the processes by which diversity is

related to LMXs and, hence, to employee performance in an international setting.

Research based on social identity theory has suggested that deep-level diversity is more

important to LMXs than is surface-level diversity (e.g. Harrison et al., 1998); however, this

research was not developed with an international context in mind. We study diversity and LMXs

in a setting where managers and workgroup members are highly diverse in terms of country of

origin, a multinational organization in Dubai. In this context, where differences in nationality at

the surface level are prominent and may be used as a proxy for deep-level differences, nationality

diversity may be more important to LMXs than deep-level differences, namely values similarity.

This may be especially likely in countries such as the United Arab Emirates (UAE) where

nationalization efforts, which are designed to promote employment of nationals over expatriates,

have become more common (e.g. Budhwar et al., 2018). This would challenge the existing theory

and research on the role of diversity as related to leader–follower relationships (e.g. Nahrgang

and Seo, 2015). We also propose that nationality diversity at the workgroup level may be related

to leader–member exchange differentiation (LMXD), which may moderate the relationship

betweenLMXandemployeeperformanceattheindividual level. Our study integrates social

identity theory (Tajfel and Turner, 1986) and LMX theory (e.g. Liden and Maslyn, 1998) to test

thesepropositions.Weseethisasimportantbecause organizations are increasingly diverse in

nationality due to globalization, technology, international labor demand and migration, which

represents important talent management opportunities (Kossek and Pichler, 2006).

In doing so, we address calls from previous researchers to examine LMX in the context of

the workgroups (Anand et al., 2011). In the context of international settings, we see diversity

as the key to understanding LMX within workgroups. Research on diversity and LMX has

focused almost exclusively on follower LMX quality (Gerstner and Day, 1997). Drawing

from social identity theory, we assert that differences between subordinates in LMX quality,

and thus relative LMX, are an important consideration in a highly diverse setting. That is,

supervisor–subordinate dyads may differ in their LMX quality based on differences in

nationality, and relative dissimilarities between subordinates in LMX quality can have

important influences on employee performance. Our findings offer novel insights into the

relationships between diversity and LMX, challenge previous LMX research and offer

important implications for relative s in highly diverse international settings.

Theory and hypotheses

The concept of diversity, “a characteristic of groups of two or more people [which] typically

refers to demographic differences of one sort or another[…]”(Ely and Thomas, 2001, p. 230)is

at the heart of key theoriesand literatures relevant to thepresent study. Research on LMX, for

instance, suggests that dyadic similarities between supervisors and subordinates are

positively related to higher levels of exchange quality (e.g. Brouer et al., 2009). Social identity

theory helps to explain why it proposes that individuals classify others into social categories

based on identity-salient characteristics (e.g. demography) and assign stereotypes (both

positive and negative) to members of those social categories (Turner, 1982). Individuals

evaluateothers in comparison to themselves,and a person’s self-concept is enh anced when the

individual feels part of groups with positive social characteristics (Tajfel and Turner, 1979).

21

Nationality

diversity and

LMX