The influence of organizational context on the managerial turnover–performance relationship

• Author: Cristina Simón,Beatriz Rodriguez‐Prado,Elena Revilla
• Published date: 01 September 2020
• DOI: http://doi.org/10.1002/hrm.22003
• Date: 01 September 2020

ORIGINAL ARTICLE

The influence of organizational context on the managerial

turnover–performance relationship

Elena Revilla

1

| Beatriz Rodriguez-Prado

2

| Cristina Simón

1

1

IE Business School-IE university, Madrid,

Spain

2

Universidad de Valladolid, Valladolid, Spain

Correspondence

Elena Revilla, IE Business School-IE University,

Maria de Molina 13, 28006 Madrid, Spain.

Email: elena.revilla@ie.edu

Abstract

This study performs a comprehensive examination of organizational context in the

relationship between managerial turnover and organizational performance. Using

theoretical frameworks of human and social capital, we focus on the moderating roles

of entity size, employment system, industry brand, and location. To test our hypothe-

ses, we worked with the company records of a multinational fashion retail group with

more than 4,000 stores grouped into eight different brands and 100,000 employees

in more than 31 countries. To estimate the causal contextual effects of the relation-

ship between voluntary managerial turnover and organizational performance, we

designed a quasi-experiment using propensity score matching analysis. Our results

show that the dysfunctional side of managerial turnover is significant for stores that

are large, for stores managed under a primary employment system, for brands operat-

ing with higher levels of service orientation, and for countries with more restrictive

employment protection legislation. We discuss the implications of these findings for

practice and for future research.

KEYWORDS

contextual moderators, managerial turnover, propensity score matching, unit performance

1|INTRODUCTION

The relationship between turnover and organizational performance

has been important for management scholars and practitioners for

decades (Hausknecht, 2017; Heavey, Holwerda, & Hausknecht, 2013;

Hom, Lee, Shaw, & Hausknecht, 2017). Meta-analyses have shown

that turnover generally has a significant negative impact on various

productivity-related outcomes and financial performance indicators

(Hancock, Allen, Bosco, McDaniels, & Pierce, 2013; Hancock, Allen, &

Soelberg, 2017; Park & Shaw, 2013). In theory, turnover entails a loss

of human and social capital that disrupts operations and communica-

tions networks and destabilizes organizational systems (Price, 1977;

Shaw, Duffy, Johnson, & Lockhart, 2005; Staw, 1980). Additionally,

the literature also reports evidence with much less empirical support

for an inverted-U shape, suggesting that turnover could impact per-

formance positively or negatively according to the level of turnover

experienced by units (Hancock et al., 2013; Park & Shaw, 2013).

While the vast majority of studies have focused on employee

turnover (Hancock et al., 2013; Hom et al., 2017) regardless of the

position held by the person who departs, we advocate models tailored

to the characteristics of a particular job position because the conse-

quences that occur at the level of nonmanagerial employees may not

be generalizable to managers (Eckardt, Skaggs, & Youndt, 2014; Hale

Jr., Ployhart, & Shepherd, 2016; Hausknecht & Holwerda, 2013;

Kacmar, Andrews, Van Rooy, Steilberg, & Cerrone, 2006; Shaw, Duffy,

et al., 2005). A managerial role entails the coordination of individuals

and activities within an organization, the determination of staffing

levels, and the reassignment of roles and redistribution of tasks within

the unit (Staw, 1980). As such, the work of managers often involves a

number of complex and unstructured tasks grounded upon significant

tacit knowledge compared with nonmanagerial employees (Eckardt

et al., 2014). Losing managers can be particularly damaging, then,

because the company loses tacit knowledge as well as a person

responsible for complex social connections between firm functions

DOI: 10.1002/hrm.22003

Hum Resour Manage. 2020;59:423–443. wileyonlinelibrary.com/journal/hrm © 2019 Wiley Periodicals, Inc. 423

and routines (Brymer & Sirmon, 2018). Hausknecht and Holwerda

(2013) argue that turnover will be more disrupting to performance

when leavers depart from relatively valuable groups that possess

greater firm-specific human and social capital.

At the same time, previous research has consistently documented

that organizational context plays a significant role in the impact of turn-

over on firm effectiveness (e.g., Hausknecht & Holwerda, 2013;

Hausknecht, Trevor, & Howard, 2009; Nyberg & Ployhart, 2013). How-

ever, current meta-analyses show that the literature has yet to reach a

state of maturity because these effects are not clearly understood,

requiring expanded turnover studies to better capture the organizational

context (Hancock et al., 2017; Hausknecht, 2017; Hom et al., 2017). To

address this situation, Lee and colleagues suggest studying the moderat-

ing effect of a particular job level because the efficacy of moderators

may vary across levels (Lee, Hom, Eberly, Li, & Mitchell, 2017). Along this

line, Eckardt et al. (2014) found that the impact of organizational capital,

as a moderator of the turnover–performance relationship, is greater for

employee turnover than for manager turnover in service firms. Thus, as

the moderating role of context may differ in nonmanagerial versus mana-

gerial turnovers, our approach to interpreting context-specific relation-

ships of turnover and organizational performance is to perform a

comprehensive examination of the organizational context of the mana-

gerial turnover–performance link. Specifically, we build upon the contex-

tual factors identified by Park and Shaw (2013) in their meta-analysis

(size, employment system, industry, and location).

Our research directly answers two questions: Does voluntary

managerial turnover consistently produce a negative significant net

effect on organizational performance? Are the impacts of voluntary

managerial turnover related to the organizational context? In

addressing these research questions, we build on individual level turn-

over theories and mainly discuss the consequences of managerial

departure on a retail store environment. By working with managers

this research provides an opportunity to investigate the losses associ-

ated with this relatively valuable group, isolate the effects of those

losses, and offer new insights into the factors that shape this

turnover–organizational performance link. Also, we focus on voluntary

turnover exclusively, not on involuntary turnover, since researchers

often have suggested that the etiology and consequences of these

two phenomena are different (e.g., Hausknecht & Trevor, 2011; Oste-

rman, 1987; Shaw, 2011). Voluntary turnover is unexpected, and gen-

erally occurs among high-performing employees with high levels of

employability. It is therefore more damaging than involuntary turnover

like discharges and layoffs planned by the company in advance (Shaw,

2011). Involuntary turnover is a decision made by the company to cor-

rect for selection mismatches and eliminate poor performers (Batt &

Colvin, 2011).

To test our hypotheses, we relied on a unique database collected

from a multinational fashion retail group with more than 4,000 stores

grouped into eight different brands and more than100,000 employees

in 31 countries. In our cross-unit study, we controlled for sources of

extraneous variance affecting the relationship between turnover and

performance, improving our capacity to isolate the examined effects

(Glebbeek & Bax, 2004; Kacmaret al., 2006). By holding certain threats

to internal validity constant (Shadish, Cook, & Campbell, 2001), cross-

unit studies (vs. cross-organization) “can be better for addressing cau-

sality issues”and ensure “consistent definition and measurement of

turnover”(Park & Shaw,2013, p. 272). Also, we concentrated on store

managers, not considering managers from headquarters who are tech-

nically more qualified and perform different job responsibilities and

tasks. Given the characteristics of our database, we believe that our

results will apply to most multinationals requiring management over

dispersed worldwide sites and facing different labor conditions,

employment systems,and degrees of competition.

This study makes three main contributions. First, responding to

calls to better explain turnover consequences of core employees such

as managers (Hausknecht & Holwerda, 2013; Nyberg & Ployhart,

2013), we blend human and social capital theories to build and test

theoretical arguments regarding the impact of losing both firm-

specific tacit knowledge and coordination between firm functions and

routines. Past studies mainly have been conducted on total turnover

rates regardless of the job level. In general, managers are compara-

tively less studied in turnover research, usually because of data avail-

ability or issues with small sample sizes.

Second, our study answers ongoing calls for contextualization of

the turnover–performance relationship (Hausknecht, 2017; Hom et al.,

2017). In a review of the contextualization models of turnover, Lee

et al. (2017) recommended tailoring the turnover–performance rela-

tionship to an organizational position. Our research offers a compre-

hensive assessmentof contextual moderators by studying a worldwide,

multiorganization, and multiunit data set of managers who voluntarily

leave their position. Since voluntary turnover is often surprising and

unmanageable (Shaw, Delery, Jenkins Jr., & Gupta, 1998), this study

predicts when themost adverse consequences could emerge.

Third, in agreement with a recent study that encourages

researchers to introduce new analytical tools in the turnover literature

(Lee et al., 2017), we address the foregoing questions using propensity

score matching. This methodology, introduced by Rosenbaum and

Rubin (1983) and widely applied in other social science fields

(e.g., labor economics, accounting, and finance) (Li, 2013; Soublière &

Gehman, 2019), allowed us to design this study as a quasi-experiment.

The use of quasi-experimental designs instead of correlational studies

has been recommended to overcome the potential dominant analyti-

cal mindset in turnover research and to better infer causality (Allen,

Hancock, Vardaman, & Mckee, 2013), since it rethinks usual causal

relations in a counterfactual stance and corrects the bias of traditional

regression models to address causal inference in observational studies

(Li, 2013; Nichols, 2007).

2|CONCEPTUAL BACKGROUND AND

HYPOTHESES

2.1 |Managerial turnover

Human capital theory suggests that organizational performance is

determined by the stock of knowledge or characteristics of employees

424 REVILLA ET AL.