– Turnoverratesandorganizationalperformance11.pdf

Article

Turnover rates andorganizational performance:Review, critique, andresearch agenda

Jason D. ShawUniversity of Minnesota

AbstractThe author of this article reviews the burgeoning literature on turnover rates and dimensions oforganizational performance, and concludes that substantial evidence indicates that turnover rateshave negative implications for several dimensions of organizational performance (e.g., safety,productivity, and monetary), that the content of turnover rates plays a role in the magnitude andform of the relationship between turnover rates and organizational performance, and that turn-over rates affect distal measures (e.g., profitability, financial performance) through decreasedproductivity and losses in human and social capital. A roadmap is provided for futuretheory-building and empirical work in this area.

Keywordsinvoluntary turnover, organizational performance, productivity, voluntary turnover

Paper received 21 January 2010; revised version accepted 29 July 2010.

Researchers are continually fascinated with

understanding individual turnover decisions in

organizations. Major reviews and a surfeit of

literature on individual-level turnover issues

appear regularly (e.g., Holtom, Mitchell, Lee,

& Eberly, 2008), but the literature is so volumi-

nous that even ‘‘reviews of the literature

reviews’’ join the mix (e.g., Price, 1989). The

turnover literature at the organizational level

is much less well developed, but has increased

dramatically in recent years with new theories

(e.g., Dess & Shaw, 2001) and a wave of

empirical testing of key relationships with turn-

over rates (e.g., Alexander, Bloom, & Nuchols,

Corresponding author:

Jason D. Shaw, Carlson School of Management, University of Minnesota, 321 19th Avenue South, Minneapolis, MN 55455,

USA.

Email: [email protected]

Organizational Psychology Review1(3) 187–213

ª The Author(s) 2011Reprints and permission:

sagepub.co.uk/journalsPermissions.navDOI: 10.1177/2041386610382152

opr.sagepub.com

OrganizationalPsychologyReview

1994; Arthur, 1994; Glebbeek & Bax, 2004;

Guthrie, 2001; Hauskneckt, Trevor, & Howard,

2009; Kacmar, Andrews, van Rooy, Steilberg,

& Cerrone, 2006; McElroy, Morrow, & Rude,

2001; Shaw, Delery, Jenkins, & Gupta, 1998;

Shaw, Duffy, Johnson, & Lockhart, 2005;

Shaw, Gupta, & Delery, 2005; Shaw, Kim, &

Park, 2009; Siebert & Zubanov, 2009; Takeuchi,

Lepak, Wang, Shaw, & Takeuchi, 2009; Ton &

Huckman, 2008; Way, 2002).

Perhaps the most rapidly growing and argu-

ably the most important area of knowledge

development concerns the relationship between

turnover rates and dimensions of organizational

performance. Many roots of this literature are

found in organizational psychology, but studies

on this relationship are also found in economics,

sociology, medical fields, and human resource

management. The literature is clearly divided

among alternative views of the turnover–

organizational performance relationship, includ-

ing the human-capital and organizational-

disruption-based linear negative and attenuated

negative perspectives, and the commonly

accepted inverted-U-based formulation. Each

view has established a foothold in the literature.

As I report below, however, some of these views

are better supported than others. These areas of

theorizing have existed in parallel, and few

researchers have attempted to understand the

underlying level of support for each approach,

to understand the boundary conditions of each

theory, or to integrate them (but see Shaw, Gupta,

& Delery, 2005, for a rare comparative analysis).

I suggest here that it is time to take stock of this

literature, to evaluate the alternative views of the

turnover–organizational performance relation-

ship, and to assess the level of support for each

view. In addition, I examine the moderators of the

relationship between turnover rates and organiza-

tional performance and outline an agenda for

future research in this area.

This paper is designed to be a representative

but not exhaustive literature review. I generally

exclude studies that include turnover rates

as a proximal dimension of organizational

performance, but make no predictions or

attempts to understand the focal relationship

(e.g., Chow, Huang, & Liu, 2008; Detert,

Trevino, Burris, & Andiappan, 2007) and focus

on the relationship between voluntary turnover

and organizational performance rather than

discharge or fire rates (Shaw, Gupta, & Delery,

2005). As will be apparent, however, the liter-

ature is beset by measures of total turnover

rates (voluntary and involuntary) and thus it is

impossible in this review to completely sepa-

rate the effects. I broadly view organizational

performance to include proximal measures such

as productivity, safety, and customer service

and distal measures of financial or accounting

performance.

This paper is organized as follows. I (a) look

at the history of the turnover rates and organi-

zational performance relationship, (b) describe

the three prevailing direct effect views

and evaluate the empirical evidence for each,

(c) examine empirical evidence concerning

moderators of the relationship, and (d) set an

agenda for future research by outlining causal

sequences and highlighting methodological

shortcomings that hinder our current under-

standing (see Figure 1 which depicts the

structure of this review).

History: The costs of turnover

The first forays into examining consequences

of turnover tended to focus on detailing the

costs of turnover (see Hom and Griffeth,

1995, for a thorough review of this literature)

or related perspectives based in utility analysis

of turnover rates (e.g., Boudreau & Berger,

1985). Under this approach, early researchers

attempted to isolate the precise costs of

separation including those associated with

exit interviews, advertising, recruitment, new-

hire training, and general administrative bur-

dens (e.g., Hall, 1981; Smith & Watkins,

1978). Not only have private and public busi-

ness organizations accepted these approaches

well, but they have also played a role in

188 Organizational Psychology Review 1(3)

Tur

nove

rra

tes

Vie

ws

and

Issu

es(l

iste

d in

box

es a

bove

)

Lin

ear

nega

tive

Att

enua

ted

nega

tive

Inve

rted

U

Mea

sure

men

t:V

olun

tary

tur

nove

rT

otal

tur

nove

r

Em

ploy

ee g

roup

:K

ey/c

ore

wor

kers

All

em

ploy

ees

Exe

cuti

ves

HR

M a

ndem

ploy

men

tsy

stem

s

Tur

nove

r ra

te c

onte

ntIn

-rol

e-pe

rfor

man

ce l

osse

sS

ocia

l ca

pita

l lo

sses

Ave

rage

ten

ure

loss

es

Org

aniz

atio

nal

cont

ext

and

char

acte

rist

ics

Uni

t si

zeP

roce

ss c

onfo

rman

ceN

ewco

mer

conc

entr

atio

n

Mod

erat

ors

Pro

xim

al p

erfo

rman

cedi

men

sion

s (e

xam

ples

)P

rodu

ctiv

ity

Saf

ety

Val

ue a

dded

Cus

tom

er s

ervi

ce q

uali

tyC

usto

mer

sat

isfa

ctio

n

Dis

tal

perf

orm

ance

dim

ensi

ons

(exa

mpl

es)

Pro

fit

Ret

urn

on a

sset

sR

etur

n on

equ

ity

Mar

ket

perf

orm

ance

Fig

ure

1.

Sum

mar

yof

the

causa

lse

quence

,vi

ew

s,is

sues,

and

hyp

oth

esi

zed

modera

tors

inth

etu

rnove

rra

tes

and

org

aniz

atio

nal

perf

orm

ance

litera

ture

.

Shaw 189

assessments of turnover costs in government

(e.g., Cascio, 1981; Lewis, 1991). This

approach to costing turnover is perhaps most

widely found in the literature on nurse turn-

over (e.g., Jones, 1990a, 1990b, 2004, 2005,

2008; O’Brien-Pallas et al., 2006), where ram-

pant turnover rates have plagued health care

organizations in the United States, Europe,

and elsewhere. In a series of papers, Jones

developed the nursing turnover cost calcula-

tion methodology (NTCCM) that includes a

variety of pre and post hire costs.

While informative, these studies and others

under this line of research (e.g., Waldman,

Kelly, Sanjeev, & Smith, 2004; Wise, 1990)

are hampered by small samples and the

somewhat idiosyncratic nature of costs across

regions, countries, and industries. That is,

although the logic of cost-based perspectives

is straightforward, and it is difficult to argue

that turnover confers no costs, an open ques-

tion remains as to whether turnover rates

driving these costs reduce productivity

(costs being only a part of the calculation)

or lower organizational financial perfor-

mance. Tellingly, to my knowledge the only

large-scale, cross-organization study of the

relationship between turnover rates and solely

cost-related and administrative outcomes is

Kasarda’s (1973) nearly 40-year-old study

of schools in Colorado. He found that teacher

turnover rates related positively to adminis-

trative intensity, defined as the proportion

of school employees assigned to administra-

tive duties, administrative overhead, and the

proportion of operating expenditures allo-

cated to general regulation, coordination, and

control functions. Thus, we can reasonably

conclude that costs increase with turnover

rates, based on cost-based logic, accounting-

based case studies (e.g., Cascio, 1981; Jones,

1990b, 2005), and Kasarda’s (1973) study,

but this literature stream fails to ‘‘empirically

demonstrate a relationship between turnover,

productivity, and effectiveness’’ (Price,

1977, p. 115).

Views of the turnover rates–organizationalperformance relationship linear negative:The human capital loss view

Perhaps dominating the economics-based per-

spective is the theoretical view that a linear and

negative relationship exists between turnover

rates and organizational performance. Human

capital theory perceives that the workforce’s

accumulated, firm-specific human capital

determines performance (Strober, 1990). Under

this view, new employees bear initial costs

because they accept wages below their mar-

ginal revenue product hoping to recoup their

losses with higher future wages, but lose that

possibility with voluntary turnover (Osterman,

1987). From the organization’s perspective,

turnover depletes these human capital stores;

replacement employees cannot perform as well

as departing job incumbents. As turnover rates

rise, organizational performance declines.

The linear negative view can also be supported

by arguments from organizational psychology

and sociology concerning organizational disrup-

tion, interference and distraction, and pool of

human capital depletion effects. Higher levels of

voluntary turnover rates are disrupting and may

interfere with a workforce’s performance—

arguments reflected in Katz and Kahn’s (1978)

and Staw’s (1980) early writings on turnover

consequences. High turnover levels disrupt

organizational systems designed to be stable, and

this interference ‘‘causes organizations to expend

potentially more energy in maintaining the input/

throughput/output process than they take in from

the environment’’ (Alexander et al., 1994,

p. 507). Related sociological arguments suggest

that high turnover levels signal that the organi-

zation is out of control (Price, 1977). Under such

circumstances, organizations must choose where

to direct limited attention and resources. They

may focus on regaining control, resulting ‘‘in a

diversion of resources from basic production into

controlling the workforce, which is likely to

lower performance’’ (Alexander et al., 1994). Or

they may focus all their energies on maintaining

190 Organizational Psychology Review 1(3)

production or service schedules, directing

attention away from safety and maintenance

concerns and ultimately lowering performance

by increasing accidents, injuries, and other

failures (Staw, 1980).

Empirical evidence. In a review of the literaturecurrent at that time, Osterman (1987) concluded

the literature had an ‘‘uncomfortably equivocal

quality’’ (p. 314), a conclusion based largely on

individual-level models from economics and to

a lesser degree on organizational-level studies

such as Medoff et al.’s (e.g., Freeman & Medoff,

1984) work on unions and productivity. The

review pointed to Brown and Medoff’s (1978)

finding that a 10% quit rate reduction wasassociated with a 1% increase in productivityas perhaps the most compelling evidence that

turnover and organizational performance were

negatively related. As summarized in Table 1,

the literature has grown dramatically in the past

20 years, especially in tests of the relationship

between turnover rates and organizational per-

formance, with analyses of dimensions of work-

force productivity growing the most. The table

shows investigations that include a test of the

relationship between turnover rates, a descrip-

tion of the samples and the levels of analysis, the

predicted relationship between the key vari-

ables, and a summary of the findings. Evidence

accumulated to date indicates that the literature

mostly supports the linear negative view. The

last column in the table, however, includes infor-

mation about whether the study reported tests of

the curvilinear relationship between turnover

and performance and, if so, what the findings

revealed about nonlinear tests. Importantly,

many studies find support for a linear negative

relationship, but conclusions about support for

other possible relationships should include a

caveat when they fail to address nonlinearity.

Several of these studies addressed the

relationship between turnover rates and

productivity-related dimensions, and some also

addressed the mediating role of productivity

and efficiency between turnover rates and more

distal measures of profitability or financial

performance. Kacmar et al. (2006) provided a

strong example showing that crew and manage-

ment turnover rates in units of a popular fast-

food chain related not only to key dimensions

of workforce productivity (e.g., customer wait

times and food waste), but also indirectly

affected unit profitability. Using total turnover

rates (quits and discharges) and organizational

performance data, they found that (a) both

forms of turnover rates (crew and management)

were associated with longer wait times,

(b) crew turnover was associated with more

food waste, and (c) turnover through increased

wait times significantly and indirectly affected

store sales and profits.

In a similar study of mortgage banking units,

Morrow and McElroy (2007) argued that vol-

untary turnover rates were associated with

lower productivity and efficiency, which in turn

indirectly led to distal measures of organiza-

tional performance (customer satisfaction and

profit). Their findings, with a cleaner measure

of voluntary turnover than the Kacmar et al.’s

(2006) study, revealed that voluntary turnover

rates related negatively to customer satisfaction

measures and profit, and further, that two pro-

ductivity measures mediated the distal effects.

Several other papers report partial tests of

this general model, with the largest concentra-

tion of studies and perhaps the strongest evi-

dence residing in the retail and customer service

contexts (see also Koslowksy & Locke, 1989;

Koys, 2001; McElroy et al., 2001, for weaker

findings). Van Iddekinge et al. (2009) estimated

a larger model of the effects of selection and

training practices on retention and profits

among a large sample of fast-food restaurants,

and found that retention rates (an approxima-

tion for the inverse of total turnover rates) sig-

nificantly and positively affected the change

in unit profitability over time. Importantly, their

analyses over six time periods allowed stronger

conclusions about the causal direction of the

turnover rates–organizational performance

relationship than we find in typical studies.

Shaw 191

Tab

le1.

Sum

mar

yofem

pir

ical

studie

sexam

inin

gth

ere

lationsh

ipbetw

een

turn

ove

rra

tes

and

org

aniz

atio

nal

perf

orm

ance

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Ale

xan

der,

Blo

om

,an

dN

uch

ols

(1994)

Com

munity

hosp

ital

sN

urs

es

Org

aniz

atio

nT

ota

lLin

ear

nega

tive

,in

vert

ed

UPers

onnelco

sts/

pat

ient

day

Lin

ear

nega

tive

(sig

nre

vers

ed)

Yes

(att

enuat

ed

nega

tive

,p

<.1

0)

Nonpers

onnel

opera

ting

cost

s/pat

ient

day

Lin

ear

nega

tive

(sig

nre

vers

ed)

Yes

(not

sign

ific

ant)

Art

hur

(1994)

Ste

elm

inim

ills

Pro

duct

ion

and

mai

nte

nan

ceem

plo

yees

Fac

ility

Tota

lLin

ear

nega

tive

,m

odera

ted

by

HR

M

Lab

or

hours

per

ton

Not

sign

ific

ant

No

Scr

apra

teN

ot

sign

ific

ant

No

Bar

on,H

annan

,an

dBurt

on

(2001)

Hig

hte

chnolo

gyst

art-

ups

All

em

plo

yees

Org

aniz

atio

nal

Tota

lLin

ear

nega

tive

Chan

gein

reve

nue

Lin

ear

nega

tive

2N

o

Bat

t(2

002)

Cal

lce

nte

rsC

ust

om

er

serv

ice

and

sale

sre

pre

senta

tive

s

Fac

ility

Volu

nta

ryLin

ear

nega

tive

Sal

es

grow

thLin

ear

nega

tive

No

Bead

les,

Low

ery

,Pett

y,an

dEze

ll(2

000)

Reta

ilst

ore

sSal

es

em

plo

yees

Unit

Tota

lfu

nct

ional

Lin

ear

nega

tive

Sal

es

grow

thLin

ear

nega

tive

No

Lin

ear

posi

tive

Sal

es

grow

thLin

ear

posi

tive

No

Can

nella

and

Ham

bri

ck(1

993)

Rece

ntly

acquir

ed

firm

s

Execu

tive

sO

rgan

izat

ion

Tota

lLin

ear

nega

tive

Retu

rnon

equity

Lin

ear

nega

tive

No

Senio

rposi

tion

(e.g

.,C

EO

,pre

sident)

Lin

ear

nega

tive

Retu

rnon

equity

Lin

ear

nega

tive

No

Less

senio

rposi

tion

Lin

ear

nega

tive

(but

weak

er

than

senio

r-posi

tion

turn

ove

r)

Retu

rnon

equity

Not

sign

ific

ant

No

Dolton

and

New

son

(2003)

Pri

mar

ysc

hools

Teac

hers

Fac

ility

Tota

lN

ot

speci

fied

Stu

dent

SA

Tsc

ore

sLin

ear

nega

tive

No

(continued

)

192 Organizational Psychology Review 1(3)

Tab

le1.

(continued)

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Gle

bbeek

and

Bax

(2004)

Tem

pora

ryem

plo

yment

agency

Pro

fess

ional

staf

fU

nit

Tota

lIn

vert

ed

USal

es

min

us

wag

eco

sts

Inve

rted

UY

es

(inve

rted

U)

Chan

gein

sale

sm

inus

wag

eco

sts

Inve

rted

UY

es

(inve

rted

U)

Guth

rie

(2001)

Cro

ss-indust

ryA

llem

plo

yees

Org

aniz

atio

nal

Tota

lLin

ear

,m

odera

ted

by

HR

MSal

es

per

em

plo

yee

Not

sign

ific

ant

No

Hau

sknech

t,T

revo

r,an

dH

ow

ard

(2009)

Leis

ure

and

hosp

ital

ity

All

em

plo

yees

Unit

Volu

nta

ryLin

ear

nega

tive

,m

odera

ted

by

unit

size

,co

hesi

veness

,an

dnew

com

er

conce

ntr

atio

n

Cust

om

er

serv

ice

qual

ity

Lin

ear

nega

tive

No

Huse

lid(1

995)

Cro

ss-indust

ryA

llem

plo

yees

Org

aniz

atio

nal

Tota

lLin

ear

Sal

es

per

em

plo

yee

Lin

ear

nega

tive

No

Tobin

q(m

arket

valu

ediv

ided

by

repla

cem

ent

cost

s)

Lin

ear

nega

tive

No

Gro

ssra

teofre

turn

on

capital

Not

sign

ific

ant

No

Ilm

akunnas

,M

alir

anta

,an

dV

ainio

mäk

i(2

005)

Man

ufa

cturi

ng

All

em

plo

yees

Fac

ility

Tota

lIn

vert

ed

UPro

duct

ivity

grow

thIn

vert

ed

UY

es

(inve

rted

U)

Kac

mar

,A

ndre

ws,

van

Rooy,

Ste

ilberg

,an

dC

err

one

(2006)

Rest

aura

nts

Cre

wU

nit

Tota

lLin

ear

nega

tive

Cust

om

er

wai

ttim

eLin

ear

nega

tive

(sig

nre

vers

ed)

No

Food

was

teLin

ear

nega

tive

(sig

nre

vers

ed)

No

Sal

es

Lin

ear

nega

tive

(biv

aria

te)

No

Pro

fit

Lin

ear

nega

tive

(biv

aria

te)

No

(continued

)

Shaw 193

Tab

le1.

(continued)

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Man

agem

ent

Cust

om

er

wai

ttim

eLin

ear

nega

tive

(sig

nre

vers

ed)

No

Food

was

teN

ot

sign

ific

ant

No

Sal

es

Not

sign

ific

ant

(biv

aria

te)

No

Pro

fit

Lin

ear

nega

tive

(biv

aria

te)

No

Keck

(1997)

Cem

ent

com

pan

ies

Execu

tive

sO

rgan

izat

ional

Tota

lLin

ear

nega

tive

,m

odera

ted

by

envi

ronm

enta

lst

abili

ty

Retu

rnon

asse

tsgr

ow

thLin

ear

posi

tive

intu

rbule

nt

year

s,lin

ear

nega

tive

inst

able

year

s

No

Min

icom

pute

rin

dust

ryR

etu

rnon

asse

tsgr

ow

thN

ot

sign

ific

ant

No

Kesn

er

and

Dal

ton

(1994)

Cro

ssin

dust

ryT

op

man

agem

ent

team

Org

aniz

atio

nT

ota

lIn

vert

ed

UR

etu

rnon

asse

tstr

end

Not

sign

ific

ant

Yes

(not

sign

ific

ant)

Kosl

ow

ksy

and

Lock

e(1

989)

Reta

ilst

ore

sSal

es

em

plo

yees

Unit

Tota

lLin

ear

nega

tive

Pro

fit

%N

ot

sign

ific

ant

No

Sal

es

per

squar

efo

ot

Not

sign

ific

ant

No

Merc

han

dis

eth

eft

and

loss

Not

sign

ific

ant

No

Koys

(2001)

Rest

aura

nts

All

em

plo

yees

Unit

Tota

lLin

ear

nega

tive

Pro

fit

Not

sign

ific

ant

No

Pro

fit

div

ided

by

tota

lsa

les

Not

sign

ific

ant

No

Cust

om

er

satisf

action

Not

sign

ific

ant

No

McE

lroy,

Morr

ow

,an

dR

ude

(2001)

Fin

anci

alse

rvic

es

All

em

plo

yees

Unit

Volu

nta

ryN

ore

lationsh

ip(n

ull)

Pro

fit

(Year

1)

Lin

ear

nega

tive

No

Pro

fit

(Year

2)

Not

sign

ific

ant

No

Cust

om

er

satisf

action

(Year

s1

and

2)

Not

sign

ific

ant

No

Pro

duct

ivity

(loan

sfu

nded

div

ided

by

tota

lsa

les

em

plo

yees)

(Year

1)

Not

sign

ific

ant

No

(continued

)

194 Organizational Psychology Review 1(3)

Tab

le1.

(continued)

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Pro

duct

ivity

(loan

sfu

nded

div

ided

by

tota

lsa

les

em

plo

yees)

(Year

2)

Lin

ear

nega

tive

No

Cost

per

loan

(Year

2)

Lin

ear

nega

tive

(sig

nre

vers

ed)

No

Meie

ran

dH

icklin

(2007)

Sch

ooldis

tric

tsT

eac

hers

Org

aniz

atio

nal

(sch

ool

dis

tric

t)

Tota

lIn

vert

ed

UStu

dent

stat

est

andar

diz

ed

test

s

Lin

ear

nega

tive

Yes

(sig

nific

ant

beca

use

of

larg

esa

mple

size

but

very

weak

)Stu

dent

SA

Tsc

ore

sIn

vert

ed

UY

es

(inve

rted

U)

Mess

ers

mith,

Guth

rie,an

dJi

(2009)

Cro

ss-indust

ryT

op

man

agem

ent

team

sO

rgan

izat

ional

Tota

lLin

ear

nega

tive

,m

odera

ted

by

indust

rydis

cretion

and

TM

Tte

nure

Retu

rnon

asse

tsLin

ear

nega

tive

No

Morr

ow

and

McE

lroy

(2007)

Mort

gage

ban

kin

gA

llem

plo

yees

Units

Volu

nta

ryLin

ear

nega

tive

Cost

per

loan

Lin

ear

nega

tive

(sig

nre

vers

ed)

No

Loan

sfu

nded

per

month

per

em

plo

yee

Lin

ear

nega

tive

No

Pro

fits

Lin

ear

nega

tive

No

Cust

om

er

satisf

action

Lin

ear

nega

tive

No

Pau

lan

dA

nan

thar

aman

(2003)

Soft

war

eA

llem

plo

yees

Org

aniz

atio

nal

Tota

lLin

ear

nega

tive

Fin

anci

alperf

orm

ance

(key

info

rman

tsu

bje

ctiv

ere

port

)

Lin

ear

nega

tive

No

Plo

mondon

et

al.

(2007)

Man

aged

care

Pri

mar

yca

repro

viders

Unit

(heal

thpla

nle

vel)

Tota

lLin

ear

nega

tive

Heal

thpla

nm

em

ber

satisf

action

Lin

ear

nega

tive

No

(continued

)

Shaw 195

Tab

le1.

(continued)

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Pre

venta

tive

care

(e.g

.,im

muniz

atio

nra

tes,

heal

thsc

reenin

gs)

Lin

ear

nega

tive

No

Sels

et

al.(2

006)

Cro

ss-indust

ryA

llem

plo

yees

Org

aniz

atio

nal

Volu

nta

ryLin

ear

nega

tive

Val

ue

added

per

work

ing

hour

Lin

ear

nega

tive

No

Shaw

,D

uffy,

Johnso

n,an

dLock

har

t(2

005)

Rest

aura

nts

All

em

plo

yees

Unit

Tota

lLin

ear

nega

tive

,m

odera

ted

by

soci

alca

pital

loss

es

and

netw

ork

densi

ty

Sal

es

per

em

plo

yee

Not

sign

ific

ant

No

Shaw

,G

upta

,an

dD

ele

ry(2

005)

Concr

ete

pip

em

anufa

cturi

ng

Pro

duct

ion

work

ers

Fac

ility

Volu

nta

ryLin

ear

nega

tive

,at

tenuat

ed

nega

tive

,in

vert

ed

U,m

odera

ted

by

HR

M

Lab

or

hours

per

ton

Att

enuat

ed

nega

tive

(sig

nre

vers

ed)

Yes

(att

enuat

ed

nega

tive

)

Acc

ident

rate

Att

enuat

ed

nega

tive

(sig

nre

vers

ed)

Yes

(att

enuat

ed

nega

tive

)

Tru

ckin

gD

rive

rsO

rgan

izat

ional

Volu

nta

ryLin

ear

nega

tive

,at

te-

nuat

ed

nega

tive

,in

vert

ed

U,

modera

ted

by

HR

M

Reve

nue

per

dri

ver

Att

enuat

ed

nega

tive

Yes

(att

enuat

ed

nega

tive

)

Acc

ident

frequency

ratio

Att

enuat

ed

nega

tive

(sig

nre

vers

ed)

Yes

(att

enuat

ed

nega

tive

)

Out-

of-

serv

ice

Perc

enta

geA

ttenuat

ed

nega

tive

(sig

nre

vers

ed)

Yes

(att

enuat

ed

nega

tive

)

Opera

ting

ratio

Att

enuat

ed

nega

tive

(sig

nre

vers

ed)

Yes

(att

enuat

ed

nega

tive

)

Retu

rnon

equity

Not

sign

ific

ant

Yes

(not

sign

ific

ant)

(continued

)

196 Organizational Psychology Review 1(3)

Tab

le1.

(continued)

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Shaw

,K

im,an

dPar

k(2

009)

Cro

ss-indust

ryA

llfu

ll-tim

eem

plo

yees

Org

aniz

atio

nal

Volu

nta

ryA

ttenuat

ed

nega

tive

,m

odera

ted

by

HR

M

Sal

es

per

em

plo

yee

Att

enuat

ed

nega

tive

Yes

(att

enuat

ed

nega

tive

)

Independent

superm

arkets

Sal

es

per

em

plo

yee

Att

enuat

ed

nega

tive

Yes

(att

enuat

ed

nega

tive

)A

ccid

ent

rate

Not

sign

ific

ant

Yes

(not

sign

ific

ant)

Shen

and

Can

nella

(2002)

Cro

ss-indust

rySenio

rexecu

tive

sO

rgan

izat

ional

Tota

lM

odera

ted

by

conte

nder

or

out-

sider

CEO

succ

ess

ion

Retu

rnon

asse

tsLin

ear

nega

tive

No

Shevc

huk,Lean

a,an

dM

itta

l(2

007)

Ele

menta

rysc

hools

Teac

hers

Unit

(sch

ool

within

dis

tric

t)

Tota

lLin

ear

nega

tive

Stu

dent

achie

vem

ent

Lin

ear

nega

tive

Yes

(not

sign

ific

ant)

Sie

bert

and

Zuban

ov

(2009)

Reta

ilst

ore

sFull-

tim

eem

plo

yees

Unit

Tota

lLin

ear

nega

tive

,in

tera

ctio

nw

ith

par

t-tim

etu

rnove

rra

tes

Sal

es

per

hour

work

ed

Lin

ear

nega

tive

Yes

(not

sign

ific

ant)

Par

t-tim

eem

plo

yees

Inve

rted

UIn

vert

ed

UY

es

(inve

rted

U)

Ton

and

Huck

man

(2008)

Reta

ilst

ore

sA

llem

plo

yees

Unit

Tota

lLin

ear

nega

tive

,m

odera

ted

by

pro

cess

confo

rman

ce

Cust

om

er

serv

ice

Att

enuat

ed

nega

tive

Yes

(att

enuat

ed

nega

tive

)

Pro

fit

mar

gin

Att

enuat

ed

nega

tive

Yes

(att

enuat

ed

nega

tive

)Full-

tim

eem

plo

yees

Cust

om

er

serv

ice

Lin

ear

nega

tive

No

Pro

fit

mar

gin

Lin

ear

nega

tive

No

Par

t-tim

eem

plo

yees

Cust

om

er

serv

ice

Lin

ear

nega

tive

No

Pro

fit

mar

gin

Lin

ear

nega

tive

No

Van

Iddekin

geet

al.

(2009)

Fas

t-fo

od

rest

aura

nts

All

em

plo

yees

Unit

Tota

lLin

ear

nega

tive

Pro

fit

mar

gin

Lin

ear

nega

tive

No

(continued

)

Shaw 197

Tab

le1.

(continued)

Pap

er

Sett

ing

Em

plo

yee

group

Leve

lof

anal

ysis

Turn

ove

rty

pe

Theory

bas

isPerf

orm

ance

dim

ensi

on

Dir

ect

rela

tionsh

ipw

ith

perf

orm

ance

1

Curv

ilinear

ity

test

ed?

Vir

any,

Tush

man

,an

dR

om

anelli

(1992)

Mic

roco

mpute

rfirm

sExecu

tive

team

sO

rgan

izat

ional

Tota

lM

odera

ted

by

CEO

succ

ess

ion

and

reori

enta

tion

Retu

rnon

asse

tsLin

ear

posi

tive

No

Wag

ner,

Pfe

ffer,

and

O’R

eill

y(1

984)

Man

ufa

cturi

ng

Top-m

anag

em

ent

team

sO

rgan

izat

ional

Tota

lU

-shap

eR

etu

rnon

inve

stm

ent

Lin

ear

nega

tive

Yes

(not

sign

ific

ant)

Wie

rsem

aan

dBan

tel(1

993)

Man

ufa

cturi

ng

Top-m

anag

em

ent

team

sO

rgan

izat

ional

Tota

lLin

ear

nega

tive

Retu

rnon

asse

tsN

ot

sign

ific

ant

No

Yan

adori

and

Kat

o(2

007)

Cro

ss-indust

ryA

llem

plo

yees

Org

aniz

atio

nal

Volu

nta

ryLin

ear

nega

tive

Sal

es

per

em

plo

yee

Lin

ear

nega

tive

Zim

merm

an,

Gru

ber-

Bal

din

i,H

ebel,

Slo

ane,

and

Mag

azin

er

(2002)

Nurs

ing

hom

es

Nurs

es

Fac

ility

Tota

lLin

ear

nega

tive

Infe

ctio

nra

tes

Lin

ear

nega

tive

(sig

nre

vers

ed)

No

Hosp

ital

izat

ion

rate

sLin

ear

nega

tive

(sig

nre

vers

ed)

No

Zim

merm

anet

al.

(2005)

Ass

iste

dca

reA

llem

plo

yees

Fac

ility

Tota

lN

ot

speci

fied

Pat

ient

funct

ional

decl

ine

(dai

lyliv

ing,

cogn

itio

n,

behav

ior,

etc

.)

Not

sign

ific

ant

for

5of6

dim

ensi

ons

No

Nurs

eai

de

Not

sign

ific

ant

for

5of6

dim

ensi

ons

No

Note

:The

dir

ect

ion

ofth

etu

rnove

r–perf

orm

ance

rela

tionsh

ipis

reve

rsed

from

the

ori

ginal

inca

ses

where

hig

her

score

sre

flect

poor

perf

orm

ance

(e.g

.,ac

cidentra

tes,

labor

hours

per

ton,an

dsa

fety

viola

tions)

.1

Rela

tionsh

ips

are

from

multiv

aria

teequat

ions

unle

sssp

eci

fied.

2T

he

turn

ove

rra

teva

riab

lew

astr

ansf

orm

ed

asth

esq

uar

ero

ot

tore

move

pote

ntial

nonlin

ear

ity.

198 Organizational Psychology Review 1(3)

Hausknecht et al. (2009) examined the relation-

ship between voluntary turnover rates and

aggregated customer service quality percep-

tions among units of a large hotel and casino.

Using an operational disruption-based frame-

work, they found that increases in voluntary

turnover rates resulted in a decrease in positive

customer service perceptions, which the authors

argued was a leading indicator of customer

retention and profitability in the gaming

industry (see also, Batt, 2002).

A spate of evidence for the linear negative

approach comes from samples of schools as

well. For example, Dolton and Newson (2003)

examined the relationship between total teacher

turnover rates and school performance among

primary schools in London, and found that a

10% increase in teacher turnover was associ-ated with declines of 2% and 2.5% for Englishand math Scholastic Aptitude Test (SAT)

scores, respectively. Beyond controls for area,

class sizes, special needs, and socioeconomics,

schools with the highest turnover rates had test

scores 10% to 11% lower than other schools.Extending these findings, some ambitious

studies by Shevchuk, Leana, and Mittal (2007)

examined the relationship between teacher

retention rates and school performance in a

large sample of U.S. elementary schools, and

found significant positive effects of teacher

retention (the inverse of total turnover rates)

and, supporting a human capital depletion

argument, further found that human capital-

based variables mediated this effect.

Plomondon et al. (2007) took the study of

turnover rates and quality to a managed-care

setting. In line with other studies of customer

satisfaction and service quality, they observed

a negative relationship between primary-

care-provider turnover rates and plan-member

satisfaction, but they also showed that these

relationships extended beyond attitudes to

actual plan-member behaviors. Increases in

primary-care-provider turnover rates related to

lower rates of childhood immunizations,

screenings for cholesterol and cervical cancer,

and childhood wellness visits before 15 months

of age. In a large-scale study of nursing homes,

Zimmerman, Gruber-Baldini, Hebel, Sloane,

and Magaziner (2002) found that nursing home

residents suffered higher infection and hospitali-

zation rates when staff had higher turnover rates.

But in a similar study of assistant care facilities,

the authors (2002) found equivocal turnover rate

relationships with medical outcomes.

Although the vast majority of studies fol-

lowing human capital loss and organizational

disruption frameworks have been conducted

within industries, two recent studies examined

these issues in cross-industry settings.

Huselid’s (1995) influential study of a random

sample of publicly traded U.S. organizations

did not examine theoretical links between turn-

over rates and productivity, but it found that

turnover rates related negatively to both pro-

ductivity and profitability. Similarly, Yanadori

and Kato (2007) surveyed a random sample of

publicly traded organizations in Japan, and

found that turnover rates related negatively to

productivity and that average employee tenure

(an operationalization of human capital accu-

mulations; e.g., Shevchuk et al., 2007)

mediated these effects. In a study of small orga-

nizations (<100 employees), Sels et al. (2006)

found that voluntary turnover rates impacted

distal measures of organizational performance

such as liquidity, solvency, and profitability

through productivity (value added per

employee).

Finally, quite a number of studies have

examined the relationship between turnover

rates among executives and dimensions of

organizational performance. Although many of

the studies were cleverly conducted, it is diffi-

cult to draw firm conclusions to either support or

disconfirm a human capital-based view. The

equivocal findings occur because of the research

questions and the nature of the samples. Many

studies (e.g., Kesner & Dalton, 1994; Shen &

Cannella, 2002; Virany, Tushman, & Romanelli,

1992) focused on the consequences of CEO

succession and drew samples requiring CEO

Shaw 199

change. This approach is akin to sampling on the

dependent variable and may have contributed to

the large variation in relationships from signifi-

cant and positive (e.g., Keck, 1997; Virany

et al., 1992), significant and negative (Cannella

& Hambrick, 1993; Shen & Cannella, 2002;

Wagner, Pfeffer, & O’Reilly, 1984), to nonsigni-

ficant (Kesner & Dalton, 1994; Wiersema &

Bantel, 1993). In a recent study, however,

Messersmith, Guthrie, and Ji (2009) focused

specifically on top management team turnover

and organizational performance. Using a large

cross-industry sample, these authors found that

a unit increase in turnover rates was associated

with reductions in the three-year rolling average

in return on assets. As these authors noted, we

need additional studies to focus on top manage-

ment team turnover rates outside of the CEO

succession paradigm.

Attenuated negative: The sociological view

The attenuated negative view has some roots in

organizational psychology, but has been largely

promulgated by Price’s (1977) sociological take

on turnover. As this influential author states:

‘‘successively higher amounts of turnover will

be found ultimately to produce, more often than

not, successively lower amounts of effectiveness

at a decreasing rate’’ (p. 119). Other researchers

have found that conceptual argument to be

sound, but Osterman (1987), primarily through

an economics lens, rather critically countered:

the ‘‘factual basis for this conclusion is shaky

and, . . . the conclusion itself is so highly con-tingent as not to be very helpful’’ (p. 299).

The underlying reason for Price’s (1977)

argument can be viewed as a variation on the

human capital depletion and organizational

control arguments in the linear negative view.

When organizations have low voluntary quit

rates, their employee groups have accumulated,

on average, high levels of human capital. Then

additional quits significantly damage human

capital accumulations and, in turn, should

weaken organizational performance. As quit

rates increase from low to moderate levels,

average accumulations of human capital are

lowered so that incumbent employees perform

less well on average. At this point, any

additional quits should be less damaging to

organizational performance. Through the lens of

organizational control and operational disruption,

at high voluntary quit rates the organization is

constantly replacing departing employees, so

incremental quits are only marginally disruptive.

Although voluntary turnover interferes with

input–throughput–output processes, and energy

and resources are redirected from safety concerns

to operation maintenance, increases in voluntary

turnover beyond a point are minimally more

disruptive.

In their comparative analysis of alternative

theories of the voluntary turnover–organiza-

tional performance relationship, Shaw, Gupta,

& Delery (2005) also used learning curve

theory to ground the attenuated negative

prediction. A learning-curve-theory approach

concerns skill and ability levels as they relate

to job performance (e.g., Logan, 1992;

Ohlsson, 1996). When quit rates are low, a typical

departing employee has a high level of human

capital, and a replacement takes quite long to

acquire that level. When turnover rates are high,

however, average firm-specific human capital

accumulations are low, and replacements can

quickly reach the performance levels of replaced

employees. At high levels, new hires typically

replace short-tenured employees and detrimental

performance effects are minimal. Shaw, Gupta,

& Delery (2005) wrote:

when the work force is being constantly

replaced (e.g., 100% turnover rate), marginal

increases in voluntary turnover (e.g., to

110%) are proportionally less problematic in

terms of productivity and safety than increases

at lower average turnover rates (e.g., from

10% to 20%). (p. 52)

Empirical evidence. Curiously, despite Price’s(1977) influence on the turnover literature,

200 Organizational Psychology Review 1(3)

nearly 20 years passed before further devel-

opment and specific empirical tests of this

formulation appeared in the literature. One

easily accessible explanation for this absence

is that the inverted-U formulation, reviewed

in the next section, grew in popularity and

general acceptance. The first evidence sup-

porting Price’s (1977) prediction appeared

in Alexander et al.’s (1994) national study

of nursing turnover rates and organizational

performance in hospitals. These authors

outlined two theories of the turnover–

performance relationship (linear negative and

inverted U), but uncovered a marginally sig-

nificant curvilinear pattern supporting Price’s

(1977) view. Had the authors used Price’s

(1977) logic as their foundation, we might

reasonably conclude that their finding (i.e.,

a relationship that was strongly negative ini-

tially but weaker at higher turnover rates)

supported his theory.

Three recent studies also supported Price’s

(1977) prediction. In their comparative analysis

of different theories, Shaw, Gupta, & Delery

(2005) found support for this formulation in

two intraindustry studies such that the voluntary

turnover rates–organizational performance

relationship was strongly negative initially, but

was later attenuated. Among a sample of con-

crete pipe manufacturers, they found the atte-

nuated U-shaped form predicting a common

productivity measure. Similar results were

found for accident rates as a second measure

of performance in the concrete pipe sample.

In a follow-up study among trucking compa-

nies, Shaw, Gupta, & Delery. (2005) replicated

the attenuated negative findings for productiv-

ity measures such as revenue generated per

driver and accident rates. Furthermore, they

showed that productivity measures partially

mediated the attenuated U-shaped relationship

between voluntary turnover and a distal mea-

sure of financial performance. Similarly, Ton

and Huckman (2008) found that the negative

effects of increasing total turnover rates on

bookstore performance were much more severe

for stores with low overall turnover levels.

Although these authors used a measure of total

turnover rather than quit rates because of archi-

val data constraints, managers’ reports sug-

gested that involuntary turnover rates were

minimal in the setting. Shaw, Kim, & Park

(2009) attempted constructive replications and

extensions (discussed in more detail below) of

the attenuated negative effect among a cross-

industry sample in Korea, and also in a sample

of single-unit U.S. supermarkets. In line with

Shaw, Gupta, & Delery (2005) and Ton and

Huckman (2008), they found a sharp negative

relationship as voluntary turnover rates rose

from low to moderate levels, but a weaker slope

as quit rates increased from moderate to high

levels.

Inverted U: The organizationalbehavior view

The inverted-U formulation of the turnover

rates–organizational performance relationship

is perhaps the most well-known, having made

its way into the lexicon and the realm of con-

ventional wisdom. Indeed, Glebbeek and Bax

(2004) stated that the optimal turnover model

from Abelson and Baysinger (1984) ‘‘can still

be regarded as the standard theoretical model

for inferring the consequences of turnover’’

(Glebbeek & Bax, 2004, p. 278). Beginning

with the pioneering papers of Dalton and Todor

(1979), Staw (1980), and Abelson and

Baysinger (1984), scholars began to delineate

the conceptual differences between zero and

optimal turnover rates and to appropriately, in

my view, criticize the existing literature for an

overemphasis on ‘‘understanding the turnover

‘problem’ rather than evaluating it as being

excessively high or low’’ (Abelson &

Baysinger, 1984, p. 334). But the literature on

the inverted-U relationship has suffered some-

what under the weight of its acceptance, largely

because of the lack of compelling and suppor-

tive findings. It is ironic that Dalton and Todor

(1979), in their pioneering essay on the positive

Shaw 201

functions of turnover, stated that the conclusion

had become axiomatic that turnover’s effects

were generally negative. But just three years

later, Bluedorn (1982) concluded that the bal-

ance of evidence tended to support an inverted-

U-shaped relationship, although as Osterman

(1987) pointed out, the conclusion was based

on the results of only three studies—a kibbutz,

a basketball team, and a single group of scien-

tists. To be fair, clearly the evidence for a

negative relationship at that time was equivocal

(e.g., see Osterman’s, 1987, review), but it seems

that in Bluedorn’s (1982) review, one axiomatic

conclusion replaced another.

Hypothesizing an inverted-U-shaped rela-

tionship has a straightforward foundation. At

low levels of voluntary turnover, the workforce

can become stagnated and closed-minded

(Dalton & Todor, 1979; Dubin, 1970). At low

to moderate levels, however, turnover can be

revitalizing by increasing workforce innova-

tion, flexibility, and adaptability (Abelson &

Baysinger, 1984; Dalton & Todor, 1979).

Moderate levels of voluntary turnover may

have other benefits. Alexander et al. (1994)

argued that newly arriving employees may

be highly motivated to perform well and may

even have more updated or current technologi-

cal skills. A modicum of turnover may also

have positive effects in terms of lowering pay-

roll and fringe-benefit costs, a key component

of certain productivity, efficiency, and ulti-

mately profitability metrics. At very high levels,

however, scholars agree that the negatives out-

weigh the positives; after a moderate amount,

voluntary turnover rates and organizational per-

formance are likely to be negatively related.

Thus, this view predicts that turnover rates and

performance are positively related between zero

and moderate turnover rates, reach a zero-slope

point, and become negatively related between

moderate and high turnover.

Empirical evidence. Recent studies have exam-ined the inverted-U formulation and have pro-

vided some of the first evidence supporting

this view. Perhaps organizational literature’s

most direct test of Abelson and Baysinger’s

(1984) hypothesized curve is Glebbeek and

Bax’s (2004) investigation among staff

employee turnover rates and performance in a

temporary agency. Using total turnover as their

key predictor, their regressions found a signifi-

cant nonlinear relationship between turnover

rates and performance. Their evaluation of the

shape of the curve is somewhat confusing, how-

ever. Although the turning or zero-slope point of

the curve was between 6.3% and 9.9% turnoverrates depending on the equation, few organiza-

tions (between 5% and 14%) had turnover ratesbelow these levels; thus, within their data range,

turnover rates generally negatively affected per-

formance. Moreover, their report was unclear as

to whether a significant positive slope occurred

at the left of the apex. Thus, they concluded that

they could not completely rule out a linear

negative relationship.

Stronger evidence is found in Meier and

Hicklin’s (2007) study of the performance of

Texas school districts. Using performance on

state-level standardized tests and college-bound

district SAT and ACT scores as dimensions of

organizational performance, these authors

found a significant nonlinear effect consistent

with Abselson and Baysinger’s (1984) hypoth-

esis—an optimal turnover rate of about 16%,which was slightly higher than the mean level

turnover rate for the districts (14%). Thus,unlike the Glebbeek and Bax (2004) study, a

substantial percentage of organizations had

turnover rates to the left of the apex; indeed the

slope of the turnover–performance line was pos-

itive at mean levels. Siebert and Zubanov (2009)

argued that the inverted-U formulation would

hold for part-time employees in their sample of

units of a retail organization in the United

Kingdom. Although their theorizing implies a

contingency that will be discussed further in the

next section, they found the hypothesized

inverted U with an optimal total turnover rate

of 15% for part-timers on a measure of laborproductivity.

202 Organizational Psychology Review 1(3)

Summary, integration of the perspectives,and issues

As the above review demonstrates, the literature

is replete with views about the shape of the

turnover rates and organizational performance

relationship. As noted by Shaw, Gupta, & Delery

(2005), these alternative perspectives are not

necessarily competing views. It is straightfor-

ward to speculate that all three views can be

integrated into a common form. One possibility

for integration is a cubic curvilinear shape where

organizational performance increases initially as

turnover rates rise, reaches an apex, and takes a

negative slope; but this negative slope is

attenuated at high turnover levels. Under an

integrative view, the prevailing slope of the

relationship between turnover rates and organi-

zational performance would be negative,

reflecting losses in human capital, social capital,

and the generally negative effects associated with

organizational disruption, as several authors have

argued. But, as turnover rates increase from low

to moderate levels, some organizational perfor-

mance improvement resulting from reduction in

stagnation and influx of new ideas increases may

be found. At high levels of turnover, the atte-

nuated negative effects may prevail as human and

social capital is depleted and performance

declines are not as incrementally damaging.

Although the integration of the perspectives

is straightforward conceptually, as more

empirical tests of curvilinear forms accumulate

in the literature, a distinct pattern of findings

which casts doubt on this possibility is begin-

ning to emerge. As noted above, most of the

evidence now favors a linear negative view,

although often curvilinear tests are not reported.

A wave of recent empirical tests supports an

attenuated negative relationship in cross-

organization samples, and these studies also

tend to examine voluntary, rather than total, turn-

over rates. The inverted-U perspective has much

less supportive evidence behind it than its popu-

larity would suggest, but three recent studies pro-

vide some support for this view formulation.

Interestingly, however, these studies have been

conducted in what amount to cross-unit, rather

than cross-organization, samples and in each case

total turnover rates, rather than voluntary turn-

over rates, have been examined.

Thus, while I encourage researchers to explore

the potential integration in empirical research

there are currently two evidence-based reasons

that cast doubt on whether this is a substantive

explanation. First, the distinctions between sam-

ples that are cross-organization (with different

policies, practices, and organizational forms) ver-

sus cross-unit (with similar policies, practices,

and organizational forms) are key issues for

researchers to address. In addition, future investi-

gations are needed to determine if choice of sam-

ple or choice of turnover rate is driving the

inverted-U effects. An alternative to the concep-

tual arguments proposed by Abelson and

Baysinger (1984) concerning stagnation is simply

that high fire rates in these settings create some

positive effects.

Resolving these issues will take time. In the

following sections, I argue that the key to a

resolution may come from examinations of the

moderators of the relationship and from over-

coming methodological problems that hamper

our understanding. Instead of attempting to

integrate perspectives into a single curvilinear

form, which also may be difficult to detect

because of unreliability and statistical power

issues, I suggest that it would be more fruitful to

isolate the conditions that might support each

formulation. I turn to these issues below.

Moderators of the turnoverrates–organizationalperformance relationship

In this section, I detail the existing evidence

regarding important moderators of the voluntary

turnover rates and organizational performance

relationship. I categorize these contingency

factors under three labels—human resources

management (HRM) and employment systems,

Shaw 203

content of turnover rates, and organizational and

work environment factors.

HRM and employment systems

Arthur’s (1994) broadly cited work in the steel

minimill industry was the first to propose that

an organization’s investments in HRM systems

play a role in gauging how severely voluntary

turnover rates damage performance. This con-

textual view holds that when investments in

HRM practices are substantial, losses through

voluntary turnover strongly and negatively

affect workforce performance and, ultimately,

organizational performance as a whole, but the

negative relationship is attenuated when HRM

investments are low. Arthur (1994) grounded

this prediction by suggesting that among high-

investment organizations—organizations with

commitment systems in his parlance—jobs

require high skill and training levels. In these

circumstances, employees take significant time

to reach adequate performance levels; turnover

greatly disrupts performance because employ-

ees ‘‘take on more managerial-level decision-

making tasks, their organizational centrality,

and hence the potential for their departure to

disrupt organizational functioning’’ (p. 674).

Guthrie (2001) refined these arguments by

arguing that high levels of HRM investments

create workforces that are rare, valuable, and

difficult for organizations to recreate and their

competitors to imitate. Organizations are also

more likely to use such practices when they

deem employees to be critical to their success.

Losses through turnover are therefore substan-

tially more detrimental. These studies provided

impressive evidence of the moderation of a

linear relationship by HRM systems. After

clustering mills into commitment (high invest-

ments) and control (low investments) cate-

gories, Arthur (1994) found very strong total

turnover rates–productivity correlations among

commitment organizations, but nonsignificant

correlations among control organizations.

Guthrie (2001) later replicated these findings

in a cross-industry sample of organizations in

New Zealand, finding that as total turnover

rates increased from mean levels to one stan-

dard deviation above the mean, per-employee

productivity decreased by nearly $34,000, but

not for organizations with little invested in

HRM.

Two recent studies have also attempted to

advance the HRM-moderated arguments.

Shaw, Kim, & Park (2009) argued that a better

specification for the HRM-moderated approach

would include a consideration of the potential

for nonlinearity in the direct relationship

between voluntary turnover rate and perfor-

mance. A concurrent consideration of curvili-

nearity and HRM moderation would rule out

the possibility that Arthur’s (1994) and Guthrie’s

(2001) findings happened only because they did

not test a curvilinear relationship between turn-

over and performance (Cohen, Cohen, West, &

Aiken, 2003). In a cross-industry study of Korean

organizations and an intraindustry study of

U.S. supermarkets, Shaw, Kim, & Park (2009)

found evidence for this curvilinear interaction;

they observed the attenuated negative pattern

only among high HRM investment organizations.

Siebert and Zubanov’s (2009) study also has

implications for employment relationships and

HRM investments. Like Arthur (1994), they

argued that under commitment HRM systems,

jobs require considerable formal training and

tacit knowledge, so firms must select employ-

ees carefully. Under these systems, which were

operationalized as full-time employees in a

retail chain, the authors argued that total turn-

over rates (quits and discharges) should

negatively affect performance. In contrast, the

authors reasoned that total turnover rates

should have an inverted-U-shaped relationship

with performance in secondary employment

relationships, which they operationalized as

part-time employees in the chain. Interest-

ingly, these arguments were not grounded in

the typical inverted-U reasoning outlined

earlier, but rather primarily in discharge-

rate arguments. That is, careful selection is

204 Organizational Psychology Review 1(3)

typically not used for hiring part-time work-

ers, so more turnover including discharges is

needed to eliminate poor performers.

Siebert and Zubanov (2009) provided strong

evidence for the curvilinear relationship among

part-time employees. The turnover–performance

results for full-time employees were much

murkier. Neither the linear term nor the squared

term for full-time employee turnover rates was

significant in performance equations, but the

authors concluded that a significant interaction

of full- and part-time turnover rates ‘‘give the

conventional negative turnover–performance

link for full-timers’’ (p. 305). At best, this inter-

pretation is unconventional for a main effect.

Looking closely at their results, their interac-

tion plot (Figure 4a, p. 309) includes part-time

values only above mean levels, and shows

a strongly negative full-time turnover rate–

organizational performance relationship only

when part-time turnover rates are more than þ1standard deviation above the mean. In addition,

they failed to consider the underlying main

effects when calculating interaction simple

slopes (Siebert & Zubanov, 2009, p. 310).

Back-of-the-envelope calculations using their

coefficients and standard þ1 and �1 standarddeviation values show that the full-time turnover

rate slopes are only negative above mean levels of

part-time turnover, but are positive (albeit not sig-

nificant) below mean part-time levels. Thus, their

conclusion regarding a prevailing negative effect

for commitment systems seems overstated.

In addition, the authors make a larger point:

optimal turnover rates may occur for low HRM-

investment employee groups, but those optimal

rates may differ as a function of the turnover

rates for other, perhaps more central, employee

groups. Conceptually, this is an important step

forward, but while Siebert and Zubanov (2009)

estimated an interaction between the two turn-

over rates, their model lacks a key higher order

term (the interaction between part-time turnover

rates squared and the linear full-time turnover

rates term) that would be necessary to provide

empirical evidence of this relationship.

Content of turnover rates

The literature on the content of turnover rates can

be broadly grouped into two categories—losses

relative to in-role performance or human capital

and social capital losses. The idea of functional

versus dysfunctional turnover having different

implications for organizational performance has

a long history in organizational psychology.

Individual-level researchers have long been con-

cerned with whether good performers stay or

leave (e.g., Dalton, Krackhardt, & Porter, 1981;

Hollenbeck & Williams, 1986; Trevor, Gerhart,

& Boudreau, 1997) and organizational-level

research has also begun to explore these issues

more fastidiously (e.g., Park, Ofori-Dankwa, &

Bishop, 1994; Shaw, Dineen, Fang, & Vellella,

2009; Shaw & Gupta, 2007).

Several studies have made strides in

determining the impact of functional versus

dysfunctional forms of voluntary turnover on

organizational performance. Beadles, Lowery,

Petty, and Ezell (2000) collected data on turn-

over and in-role performance from 1,750 indi-

viduals in 26 retail stores. Calculating in-role

performance losses from performance records

using meta-analytic techniques across the stores,

these authors found that turnover frequency

rates were negatively related to sales growth

(�.15), but that turnover functionality—a com-posite index of good performer retention and

poor performer withdrawal—was positively

related (.18). They calculated that losing an

employee in the highest performing category

was five times more detrimental to organiza-

tional performance than losing a less well-

performing but still acceptable employee.

Two recent studies have directly addressed

the content of turnover rates by attempting to

capture the losses organizations experience

through in-role performance or human capital

losses and social capital losses, or the damage

to interpersonal relationships and communica-

tion networks when employees leave. Building

on the elements of a social-capital theory of

turnover and performance from Dess and

Shaw 205

Shaw (2001), Shaw, Duffy et al. (2005) argued

that the relationship between turnover rates and

organizational performance would be stronger

when key individuals in the organizational net-

work were lost. These authors operationalized

social-capital losses as the extent to which

employees in key bridging or ‘‘structural hole’’

positions departed. Results among a sample of

units of a restaurant chain indicated that social-

capital losses substantially and negatively

related to store performance (productivity and

change in productivity) when overall store

turnover rates were low. Social-capital losses

were most damaging when the first network

communication holes were created, but less

damaging in high turnover stores where many

gaps were already apparent. Interestingly, in-

role performance losses (calculated from super-

visor reports of employee performance) were

not significantly related to store performance.

Shevchuk et al. (2007) advanced these

results and argued that human- and social-

capital losses would have multiplicative

effects on organizational performance. In their

sample of schools, they operationalized

human-capital losses as those associated with

tenure and social-capital losses as those asso-

ciated with the closeness of connections with

other teachers and administrators. They found

substantial support for their predictions—

beyond the main effects of turnover rates,

human- and social-capital losses interacted such

that the relationship between human-capital

losses and school performance was significantly

stronger (negative) when social-capital losses

were also high.

Organizational context and characteristics

Studies of organizational context factors that may

exacerbate or attenuate the effects of turnover

rates on organizational performance are rare, but

two recent studies have provided promising

evidence concerning these effects. Hausknecht

et al. (2009) argued that the concentration of

newcomers in a unit would exacerbate the effects

of turnover rates on performance because of a

lack of resource availability for socialization and

training. Similarly, Ton and Huckman (2008)

argued that process conformance, or the degree to

which managers aim to reduce variation in

operations in accordance with prescribed stan-

dards, would mitigate the effects of turnover rates

on performance. This line of reasoning shares

some common ground with Hausknecht et al.’s

(2009) hypotheses concerning resources avail-

able for socialization and knowledge transfer.

As Ton and Huckman (2008) explained, high

levels of process conformance allow knowledge

concerning task performance and other critical

issues to be transferred more easily to new

employees, while in low-conformance situations

where deviations from the norm are accepted,

passing along new information is more difficult.

Both studies reported support for their hypoth-

eses—turnover was more strongly and negatively

related to performance when newcomer concen-

tration was high (Hausknecht et al., 2009) and

process conformance was low (Ton & Huckman,

2008).

In their organizational-disruption frame-

work, Hausknecht et al. (2009) also argued that

higher turnover rates would be more damaging

to organizational performance in larger units, in

part because it would exacerbate coordination,

communication, and existing inefficiencies

associated with larger groups. They found sup-

port for this proposition as well—that is, the

negative relationship between turnover rates

and customer service quality (in gaming units)

was more strongly negative for larger units.

Toward the future: Research agendaand methodological assessments

My suggestions for future research in this area

and for methodological improvements overlap

considerably. Indeed, some advances supporting

existing and new theory can come only if we can

improve measurements of key variables (turn-

over rates primarily) as well as discover research

designs and analysis approaches that allow us to

206 Organizational Psychology Review 1(3)

rule out alternative explanations. I address a

variety of these issues below.

Overcoming measurement issues. Individual-levelresearch that isolates high- and low-

performance leave is quite well developed.

Much progress in this area can be traced to the

ambitious work of Trevor et al. (e.g., Trevor

et al., 1997). At the organizational level, how-

ever, we know comparatively little about the

impact of functional and dysfunctional turnover

on organizational performance (see McElroy

et al., 2001, for an exception). Some of this,

as noted above, has to do with measurement

problems in the literature; primarily the reliance

on measures of total turnover that include quit

and discharge rates. In my judgment, a

literature-level pattern is emerging in terms of

turnover–organizational performance relation-

ships when we use different operationalizations

of turnover rates. When researchers operationa-

lize turnover rates using a combination of quits

and discharges (e.g., Glebbeek & Bax, 2004;

Meier & Hicklin, 2007; Siebert & Zubanov,

2009), inverted-U or optimal turnover level

effects are more commonly found. When

researchers examine voluntary turnover among

full-time employees, or when they employ total

turnover rates in settings where discharges are

minimal (e.g., Alexander et al., 1994; Shaw,

Gupta, & Delery, 2005; Shaw, Kim, & Park,

2009; Ton & Huckman, 2008), the evidence

increasingly supports Price’s (1977) attenuated

negative theory. Although minimizing the

impact of this conflation on the cumulative

body of knowledge is tempting, unless and until

we can trace the sources of turnover and address

the content of turnover rates, we are unlikely to

resolve the theoretical confusion. In particular,

although total turnover measures may yield

inverted-U relationships with performance

dimensions, it is impossible to conclude

whether support exists for underlying theoreti-

cal arguments about reductions in stagnation

(e.g., the basis for Abelson & Baysinger’s,

1984, arguments) or Siebert and Zubanov’s

(2009) sorting arguments.

Beyond the terms of voluntary and invo-

luntary distinctions in turnover rates, recent

works by Beadles et al. (2000) and Shaw, Duffy

et al. (2005) assessed performance losses from

turnover and took steps forward for developing

evidence about functional and dysfunctional

rates. Researchers could advance the literature

substantially by testing alternative forms of the

turnover–performance relationship across dif-

ferent types of turnover rates. Such studies

could answer such questions as ‘‘Do good-

performer and poor-performer quit rates affect

organizational performance differently?’’

‘‘What shape do we find in the turnover–

performance relationship for quit rates among

good performers and poor performers?’’

Developing richer conceptualizations ofemployment relationships. At the organizationallevel, researchers have begun to investigate the

antecedents of separate quit rates by perfor-

mance level (e.g., Shaw, Dineen et al., 2009;

Shaw & Gupta, 2007), heeding calls for further

understanding workplace sorting effects (e.g.,

Gerhart & Rynes, 2003). For example, Shaw

and Gupta (2007) found that performance- and

seniority-based pay dispersion would result in

different quit patterns across high, average, and

poor performers. Shaw, Dineen et al. (2009)

further argued and found that different employ-

ment relationships—drawing on Tsui, Pearce,

Porter, and Tripoli’s (1997) model—resulted

in different quit patterns by performance level.

These direct tests of employment relationships

and functional and dysfunctional turnover

rates, combined with insinuations that these

sorting effects occur and have implications

for organizational performance (Siebert &

Zubanov, 2009), could extend our understand-

ing of the turnover rates–organizational per-

formance relationship. In particular, the

HRM-moderated approach of Arthur (1994)

and Guthrie (2001), and implied by Siebert and

Zubanov (2009), could be enhanced by

Shaw 207

employing richer conceptualizations of HRM-

based employment relationships. Following

Shaw, Dineen et al. (2009), one approach would

be to adopt the employee–organization-

relationship approach outlined by Tsui et al.

(1997; see also Hom, Tsui, Wu, & Lee, 2009),

showing that from an employer perspective,

some HRM practices represent company indu-

cements and investments in employees, while

other practices represent employers’ require-

ments or expectations of their workforce

(expectation-enhancing practices). Practices

such as base pay and benefits levels represent

inducements and investments, while

performance-based pay and emphasis

on performance appraisal suggest higher levels

of expected contributions. By crossing these two

axes, we form a typology of four employee–

organizational relationships—spot contract,

underinvestment, overinvestment, and mutual

investment. Shaw, Dineen et al. (2009) showed

that while overall quit rates tend to be low under

mutual-investment systems, sorting effects

indicate that expectation-enhancing practices

attenuate the negative relationship between

inducements and investments and good-

performer quit rates, and exacerbate the negative

relationship with poor-performer quit rates.

Good-performer quit rates tended to be highest

in spot-contract situations, but otherwise low,

including underinvestment situations where

employers offered no long-term commitment

but still expected much from employees. The

authors reasoned that the likelihood of relative

advantage may have outweighed the stability

of long-term investments for good performers

in such systems. In contrast, poor-performer quit

rates tended to be highest in underinvestment

situations but generally low otherwise.

While Shaw and Gupta’s (2007) and Shaw,

Dineen et al.’s (2009) studies showed that

employee–organization exchange relation-

ships can predict differential quit rates, a

fruitful path would be to explore how turnover

under these different models affects organiza-

tional performance. When evaluating the current

HRM-moderated literature, the operationaliza-

tion of employment systems in Guthrie (2001)

and Shaw, Kim, & Park (2009) runs on a single

continuum from spot contract (or low road) to

mutual investment, while the operationaliza-

tion is a dichotomy in Arthur (1994) and Siebert

and Zubanov (2009). These approaches fail to

consider that many organizations may have

imbalanced under- or overinvestment systems.

Applying the Shaw, Dineen et al.’s (2009) find-

ings on workforce sorting, the focus on a single

continuum of HRM practices would reveal

little information about how the workforce was

being sorted as, indeed, quit rates among good

and poor performers were low in a mutual-

investment (or commitment) system. Because

the breadth and depth of employee contributions

differ across employment systems (Siebert &

Zubanov, 2009), a fruitful endeavor would be

to determine the implications of different turn-

over patterns for organizational performance

under a richer conceptualization of employment

relationship.

Turnover and the social fabric of organizations.A most promising direction for future turn-

over research, I suggest, is the examination of

how turnover changes, damages, or perhaps

improves the organization’s social fabric

(through functional quit patterns). Dess and

Shaw’s (2001) foray into the realm of social

capital has brought some progress in terms of

detailing how losses relate to communication

patterns and accumulated trust and confidence

(e.g., Shaw, Duffy et al., 2005; Shevchuk et al.,

2007). This research remains at an early stage,

but dovetails well with individual-level

research on social networks and individual

turnover decisions. Krackhardt and Porter’s

(1986) early work showed snowball effects;

that is, restaurant turnover patterns were linked

to employees’ social networks and often

occurred in clusters. Recent contributions have

shown convincingly that social networks and

social relationships of individuals (e.g.,

Mossholder, Settoon, & Henagan, 2005) and

208 Organizational Psychology Review 1(3)

their coworkers (Felps et al., 2009) substan-

tially affect individual turnover decisions.

When key individuals embedded in social net-

works leave, the effects are highly damaging

to proximal workforce performance outcomes,

and these effects are apparent beyond tradi-

tional in-role performance and human capital-

based losses from departures.

In addition to answering the need for addi-

tional tests of social-capital loss effects, future

research could address several important ques-

tions. First, little is known about the damage

to social networks and patterns of relationships

when turnover occurs in isolation or in clusters.

When key actors with bridging or linking posi-

tion in the network decide to quit, a communi-

cation gap is left that ultimately damages

organizational performance (Shaw, Duffy

et al., 2005) because social networks provide

conduits for sharing, expanding, and transform-

ing knowledge (Nahapiet & Ghoshal, 1998;

Shevchuk et al., 2007). But we do not know

whether these gaps persist for long or are

quickly filled by current employees or replace-

ments. In addition, existing research on social

capital and human capital losses have estimated

the performance effects fairly statically, but future

investigations that include network changes

would be a step forward (e.g., Subramony &

Holtom, 2010; van Iddekinge et al., 2009).

Conclusions

I encourage future researchers to obtain mea-

sures of turnover rates that include the type of

turnover—quits, discharges, and, if available,

other sources such as reduction in force and

retirement—or, if not possible, to obtain esti-

mates on the relative percentages of each.

Truly, measures can be critiqued for containing

errors; granted, in certain instances the line

between a quit and a discharge is blurred. But,

largely distinctions are clear, and this type will

help rule out alternative explanations. I also

encourage tests of nonlinearity in all future

studies.

My goal in this review is to provide a plat-

form from which future researchers could

advance this literature, which in the last decade

has made outstanding progress with many

unique and insightful contributions. I concur

with prior researchers who have called for

competitive tests of alternative and competing

hypotheses (e.g., Holtom et al., 2008; Platt,

1964; Shaw, Gupta, & Delery, 2005); it is time

to step forward by designing studies that allow

for fair tests of alternative perspectives and/or

by developing more precise predictions that

will reveal through empirical testing the con-

ditions supporting each view.

Funding

This research received no specific grant from any

funding agency in the public, commercial, or not-

for-profit sectors.

References

Abelson, M. A., & Baysinger, B. D. (1984). Optimal

and dysfunctional turnover: Toward an organiza-

tional level model. Academy of Management

Review, 9, 331–341.

Alexander, J. A., Bloom, J. R., & Nuchols, B. A.

(1994). Nursing turnover and hospital effici-

ency: An organizational-level analysis. Industrial

Relations, 33, 505–520.

Arthur, J. B. (1994). Effects of human resource man-

agement systems on manufacturing performance

and turnover. Academy of Management Journal,

37, 670–687.

Baron, J. N., Hannan, M. T., & Burton, M. D. (2001).

Labor pains: Change in organizational models

and employee turnover in young high-tech firms.

American Journal of Sociology, 106, 960–1012.

Batt, R. (2002). Managing customer services:

Human resource practices, quit rates, and sales

growth. Academy of Management Journal, 45,

587–597.

Beadles, N. A., II., Lowery, C. M., Petty, M. M., &

Ezell, H. (2000). An examination of the relation-

ships between turnover functionality, turnover

Shaw 209

frequency, and organizational performance. Jour-

nal of Business and Psychology, 15, 331–338.

Bluedorn, A. (1982). The theories of turnover:

Causes, effects, meanings. In S. Bacharach

(Ed.), Research in the sociology of organizations

(pp. 75–128). Greenwich, CT: JAI Press.

Boudreau, J. W., & Berger, C. J. (1985). Decision-

theoretic utility analysis applied to employee

separations and acquisitions. Journal of Applied

Psychology, 70, 581–612.

Brown, C., & Medoff, J. L. (1978). Trade unions in

the production process. Journal of Political

Economy, 86, 355–378.

Cannella, A. A., Jr., & Hambrick, D. C. (1993).

Effects of executive departures on the perfor-

mance of acquired firms. Strategic Management

Journal, 14, 137–152.

Cascio, W. F. (1981). Costing human resources: The

financial impact of behaviors in organizations.

Cincinnati, OH: South-Western Publishing.

Chow, I. H., Huang, J. C., & Liu, S. (2008). Strategic

HRM in China: Configurations and competitive

advantage. Human Resource Management, 47,

687–706.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S.

(2003). Applied multiple regression/correlation

analysis for the behavioral sciences (3rd ed.).

Mahwah, NJ: Erlbaum.

Dalton, D. R., Krackhardt, D. M., & Porter, L. W.

(1981). Functional turnover: An empirical assess-

ment. Journal of Applied Psychology, 66, 716–721.

Dalton, D. R., & Todor, W. D. (1979). Turnover

turned over: An expanded and positive perspec-

tive. Academy of Management Review, 4, 225–235.

Dess, G. G., & Shaw, J. D. (2001). Voluntary

turnover, social capital, and organizational per-

formance. Academy of Management Review, 26,

446–456.

Detert, J. R., Trevino, L. K., Burris, E. R., &

Andiappan, M. (2007). Managerial modes of influ-

ence and counterproductivity in organizations:

A longitudinal business-unit-level investigation.

Journal of Applied Psychology, 92, 993–1005.

Dolton, P., & Newson, D. (2003). The relationship

between teacher turnover and school perfor-

mance. London Review of Education, 1, 131–140.

Dubin, R. (1970). Management in Britain—

Impression of a visiting professor. Journal of

Management Studies, 7, 183–198.

Felps, W., Mitchell, T. R., Hekman, D. R., Lee, T. W.,

Holtom, B. C., & Harman, W. S. (2009). Turnover

contagion: How coworkers’ job embeddedness

and job search behavior influence quitting.

Academy of Management Journal, 52, 545–561.

Freeman, R. B., & Medoff, J. L. (1984). What do

unions do? New York, NY: Basic Books.

Gerhart, B., & Rynes, S. L. (2003). Compensation:

Theory, evidence, and strategic implications.

Thousand Oaks, CA: Sage.

Glebbeek, A. C., & Bax, E. H. (2004). Is high

employee turnover really harmful? An empiri-

cal test using company records. Academy of

Management Journal, 47, 277–286.

Guthrie, J. P. (2001). High-involvement work prac-

tices, turnover, and productivity: Evidence from

New Zealand. Academy of Management Journal,

44, 180–190.

Hall, T. E. (1981). How to estimate employee

turnover costs. Personnel, 58, 43–52.

Hausknecht, J. P., Trevor, C. O., & Howard, M. J.

(2009). Unit-level turnover rates and customer

service quality: Implications for group cohesive-

ness, newcomer concentration, and size. Journal

of Applied Psychology, 94, 1068–1075.

Hollenbeck, J. R., & Williams, C. R. (1986). Turn-

over functionality versus turnover frequency:

A note on work attitudes and organizational

effectiveness. Journal of Applied Psychology,

71, 606–611.

Holtom, B. C., Mitchell, T. R., Lee, T. W., & Eberly,

M. B. (2008). Turnover and retention research:

A glance at the past, a closer review of the pres-

ent, and a venture into the future. Academy of

Management Annals, 2, 231–274.

Hom, P. W., & Griffeth, R. W. (1995). Employee turn-

over. Cincinnati, OH: South-Western Publishing.

Hom, P. W., Tsui, A. S., Wu, J. B., & Lee, T. W.

(2009). Explaining employment relationships

with social exchange and job embeddedness.

Journal of Applied Psychology, 94, 277–297.

Huselid, M. A. (1995). The impact of human

resource management practices on turnover,

210 Organizational Psychology Review 1(3)

productivity, and corporate financial performance.

Academy of Management Journal, 38, 635–642.

Ilmakunnas, P., Maliranta, M., & Vainiomäki, J.

(2005). Worker productivity and productivity

growth. Applied Economic Letters, 12, 395–398.

Jones, C. B. (1990a). Staff nurse turnover costs,

Part 1: A conceptual model. Journal of Nursing

Administration, 20, 18–22.

Jones, C. B. (1990b). Staff nurse turnover costs,

Part 2: Measurements and results. Journal of

Nursing Administration, 20, 27–32.

Jones, C. B. (2004). The costs of nurse turnover,

Part 1: An economic perspective. Journal of

Nursing Administration, 34, 562–570.

Jones, C. B. (2005). The costs of turnover, Part 2:

Application of the nursing turnover cost

calculation methodology. Journal of Nursing

Administration, 35, 41–49.

Jones, C. B. (2008). Revisiting nurse turnover costs:

Adjusting for inflation. Journal of Nursing

Administration, 38, 11–18.

Kacmar, K. M., Andrews, M. C., van Rooy, D.,

Steilberg, C., & Cerrone, S. (2006). Sure every-

one can be replaced . . . but at what cost?

Turnover as a predictor of unit-level perfor-

mance. Academy of Management Journal, 49,

133–144.

Kasarda, J. D. (1973). Effects of personnel turnover,

employee qualifications, and professional staff

ratios on administrative intensity and overhead.

Sociological Quarterly, 14, 350–358.

Katz, D., & Kahn, R. L. (1978). The social psychol-

ogy of organizations (2nd ed.). New York, NY:

Wiley.

Keck, S. L. (1997). Top management team structure:

Differential effects by environmental context.

Organization Science, 8, 143–156.

Kesner, I. F., & Dalton, D. R. (1994). Top manage-

ment turnover and CEO succession: An

investigation of the effects of turnover on per-

formance. Journal of Management Studies, 31,

701–713.

Koslowksy, M., & Locke, G. (1989). Turnover and

aggregate organizational performance. Applied

Psychology: An International Review, 38,

121–129.

Koys, D. J. (2001). The effects of employee satisfac-

tion, organizational citizenship behavior, and

turnover on organizational effectiveness: A unit-

level, longitudinal study. Personnel Psychology,

54, 101–114.

Krackhardt, D., & Porter, L. W. (1986). The snowball

effect: Turnover in embedded social networks.

Journal of Applied Psychology, 72, 382–386.

Lewis, G. B. (1991). Turnover and quiet crisis in the

federal civil service. Public Administration

Review, 51, 145–155.

Logan, G. D. (1992). Shapes of reaction-time distri-

butions and shapes of learning curves: A test of

the instance theory of automaticity. Journal of

Experimental Psychology, 18, 883–914.

McElroy, J. C., Morrow, P. C., & Rude, S. C. (2001).

Turnover and organizational performance:

A comparative analysis of voluntary, involuntary,

and reduction-in-force turnover. Journal of

Applied Psychology, 86, 1294–1299.

Meier, K. J., & Hicklin, A. (2007). Employee turn-

over and organizational performance: Testing a

hypothesis from classical public administration.

Journal of Public Administration Research and

Theory, 18, 573–590.

Messersmith, J. G., Guthrie, J. P., & Ji, Y. Y. (2009).

Turnover at the top: Executive team departures

and firm performance. Working paper, University

of Kansas.

Morrow, P. C., & McElroy, J. C. (2007). Efficiency

as a mediator in turnover–organizational perfor-

mance relations. Human Relations, 60, 827–849.

Mossholder, K. W., Settoon, R. P., & Henagan, A. P.

(2005). A relationship perspective on turnover:

Examining structural, behavior, and attitudinal

predictors. Academy of Management Journal,

48, 607–618.

Nahapiet, J., & Ghoshal, S. (1998). Social capital, intel-

lectual capital, and the organizational advantage.

Academy of Management Review, 23, 242–266.

O’Brien-Pallas, L., Griffin, P., Shamian, J.,

Buchan, J., Duffield, C., Hughes, F., . . . Stone,

P. W. (2006). The impact of nurse turnover on

patient, nurse, and system outcomes: A pilot study

and focus for a multicenter international study.

Policy, Politics, and Nursing Practice, 7, 169–179.

Shaw 211

Ohlsson, S. (1996). Learning from performance

errors. Psychological Review, 103, 241–262.

Osterman, P. (1987). Turnover, employment

security, and the performance of the firm. In

M. Kleiner (Ed.), Human resources and the

performance of the firm (pp. 275–317).

Madison, WI: Industrial Relations Research

Association.

Park, H. Y., Ofori-Dankwa, J., & Bishop, D. R.

(1994). Organizational and environmental deter-

minants of functional and dysfunctional turnover:

Practical and research implications. Human

Relations, 47, 353–366.

Paul, A. K., & Anantharaman, R. N. (2003). Impact

of people management practices on organiza-

tional performance: Analysis of a causal model.

International Journal of Human Resource

Management, 14, 1246–1266.

Platt, J. R. (1964). Strong inference. Science, 146,

347–353.

Plomondon, M. E., Magid, D. J., Steiner, J. F.,

MaWhinney, S., Gifford, B. D., Shih, S. C., . . .

Rumsfeld, J. S. (2007). Primary care provider

turnover and quality in managed care organiza-

tions. The American Journal of Managed Care,

13, 465–472.

Price, J. L. (1977). The study of turnover. Ames, IA:

Iowa State University Press.

Price, J. L. (1989). The impact of turnover on the orga-

nization. Work and Occupations, 16, 461–473.

Sels, L., De Winne, S., Maes, J., Delmotte, J.,

Faems, D., & Forrier, A. (2006). Unravelling the

HRM-performance link: Value-creating and

cost-increasing effects of small business HRM.

Journal of Management Studies, 43, 319–342.

Shaw, J. D., Delery, J. E., Jenkins, G. D., & Gupta, N.

(1998). An organizational-level analysis of vol-

untary and involuntary turnover. Academy of

Management Journal, 41, 511–525.

Shaw, J. D., Dineen, B. R., Fang, R., & Vellella, R. V.

(2009). Employee–organization exchange rela-

tionships, HRM practices, and quit rates of good

and poor performers. Academy of Management

Journal, 52, 1016–1033.

Shaw, J. D., Duffy, M. K., Johnson, J. L., & Lockhart,

D. E. (2005). Turnover, social capital losses, and

performance. Academy of Management Journal,

48, 594–606.

Shaw, J. D., & Gupta, N. (2007). Pay system character-

istics and quit patterns of good, average, and poor

performers. Personnel Psychology, 60, 903–928.

Shaw, J. D., Gupta, N., & Delery, J. E. (2005). Alter-

native conceptualizations of the relationship

between voluntary turnover and organizational

performance. Academy of Management Journal,

48, 50–68.

Shaw, J. D., Kim, E., & Park, T. Y. (2009). Voluntary

turnover, HRM strategy, and organizational per-

formance: Theoretical extensions, tests, and con-

structive replication. Working paper, University

of Minnesota.

Shen, W., & Cannella, A. A., Jr. (2002). Revisiting

the performance consequences of CEO succes-

sion: The impacts of successor types, postsucces-

sion senior executive turnover, and departing

CEO tenure. Academy of Management Journal,

45, 717–733.

Shevchuk, I., Leana, C., & Mittal, V. (2007).

Employee retention and organizational perfor-

mance: The mediating role of organization- and

task-specific forms of human and social capital.

Working paper, University of Pittsburgh.

Siebert, W. S., & Zubanov, N. (2009). Searching for

the optimal level of employee turnover: A study

of a large UK retail organization. Academy of

Management Journal, 52, 294–313.

Smith, H. L., & Watkins, W. E. (1978). Managing

manpower turnover costs. Personnel Administra-

tor, 23, 46–50.

Staw, B. M. (1980). The consequences of turnover.

Journal of Occupational Behavior, 1, 253–273.

Strober, M. H. (1990). Human capital theory: Impli-

cations for HR managers. Industrial Relations,

29, 214–239.

Subramony, M., & Holtom, B. (2010). Effects of

turnover levels and trajectories on customer out-

comes: A longitudinal investigation. Working

paper, Northern Illinois University.

Takeuchi, R., Lepak, D. P., Wang, H., Shaw, J. D., &

Takeuchi, K. (2009). Ties that bind: Examining

the impact of human and social capital on

employee turnover and performance. Working

212 Organizational Psychology Review 1(3)

paper, Hong Kong University of Science and

Technology.

Ton, Z., & Huckman, R. S. (2008). Managing the

impact of employee turnover on performance:

The role of process conformance. Organization

Science, 19, 56–68.

Trevor, C. O., Gerhart, B., & Boudreau, J. W. (1997).

Voluntary turnover and job performance: Curvili-

nearity and the moderating influences of salary

growth and promotions. Journal of Applied

Psychology, 82, 44–61.

Tsui, A. S., Pearce, J. L., Porter, L. W., & Tripoli,

A. M. (1997). Alternative approaches to the

employee organization relationship: Does

investment in employees pay off? Academy of

Management Journal, 40, 1089–1121.

Van Iddekinge, C. H., Ferris, G. R., Perrewé, P. L.,

Perryman, A. A., Blass, F. R., & Heetderks, T. D.

(2009). Effects of selection and training on

unit-level performance over time: A latent

growth modeling approach. Journal of Applied

Psychology, 94, 829–843.

Virany, B., Tushman, M. L., & Romanelli, E.

(1992). Executive succession and organization

outcomes in turbulent environments: An organi-

zation learning approach. Organization Science,

3, 72–91.

Wagner, W. G., Pfeffer, J., & O’Reilly, C. A., III.

(1984). Organizational demography and turnover

in top-management groups. Administrative

Science Quarterly, 29, 74–92.

Waldman, J. D., Kelly, F., Sanjeev, A., & Smith, H. L.

(2004). The shocking cost of turnover in health

care. Health Care Management Review, 29, 2–7.

Way, S. A. (2002). High performance work systems

and intermediate indicators of firm performance

within the U.S. small business sector. Journal of

Management, 28, 765–785.

Wiersema, M. F., & Bantel, K. A. (1993). Top

management team turnover as an adaptation

mechanism: The role of the environment.

Strategic Management Journal, 14, 485–504.

Wise, L. C. (1990). Tracking turnover. Nursing

Economics, 8, 45–51.

Yanadori, Y., & Kato, T. (2007). Average employee

tenure, voluntary turnover ratio, and labour pro-

ductivity: Evidence from Japanese firms. Interna-

tional Journal of Human Resource Management,

18, 1841–1857.

Zimmerman, S., Gruber-Baldini, A. L., Hebel, J. R.,

Sloane, P. D., & Magaziner, J. (2002). Nursing

home facility risk factors for infection and

hospitalization: Importance of registered nurse

turnover, administration, and social factors.

Journal of the American Geriatrics Society, 50,

1987–1995.

Zimmerman, S., Sloane, P. D., Eckert, J. K., Gruber-

Baldini, A. L., Morgan, L. A., Hebel, J. R., . . .

Chen, C. K. (2005). How good is assisted living?

Findings and implications from an outcome

study. Journal of Gerontology, 60b, s195–s204.

Author biography

Jason D. Shaw is a professor and the Curtis L.

Carlson School-wide Professor in the Carlson

School of Management at the University of

Minnesota. He received his PhD from the

University of Arkansas in 1997. His research

interests include the psychology of pay,

turnover, and person–environment congruence

issues. His research has appeared in publications

such as the Academy of Management Journal,

Academy of Management Review, Journal of

Applied Psychology, Personnel Psychology,

Organizational Behavior and Human Decision

Processes, and Strategic Management Journal,

among other outlets. He is currently an associate

editor of the Academy of Management Journal.

Shaw 213