Literature review on labour turnover

and retention strategies

Method
MINTRAC commissioned acirrt to undertake a literature review examining the causes
of high labour turnover in a range of industries in Australia and overseas. The review
examined a wide range of factors that could contribute to high labour turnover
including characteristics of emploees! recruitment practices! wages and conditions
and career development opportunities.
"actors of particular interest to MINTRAC! such as the impact of target income
workers! the role of training! industr seasonalit and the conse#uences of rural
isolation! were considered. In addition! the results of a recent stud on mobilit in the
Australian labour market are presented.
The second part of the literature review presents some of the strategies used b
emploers to increase retention of staff and discusses the different was in which
turnover can be measured and costed.
Causes of labour turnover
Introduction
$mploee turnover is a much studied phenomenon. There is a vast literature on the
causes of voluntar emploee turnover dating back to the %&'(s. ) developing
multivariate models that combine a number of factors contributing to turnover and
empiricall testing the models researchers have sought to predict wh individuals
leave organisations. Man studies are based on onl a small number of variables
which often onl explain a small amount of variabilit in turnover. Another criticism
of turnover studies is that the do not ade#uatel capture the complex pschological
processes involved in individual turnover decisions. A recent stud of turnover b
)oxall et al *+((,- in New .ealand confirmed the view that motivation for /ob change
is multidimensional and that no one factor will explain it.
0owever! over time there have been a number of factors that appear to be consistentl
linked to turnover. An earl review article of studies on turnover b Moble et al
*%&1&- revealed that age! tenure! overall satisfaction! /ob content! intentions to remain
on the /ob! and commitment were all negativel related to turnover *i.e. the higher the
variable! the lower the turnover-. In %&&'! a meta2analsis of some 3(( turnover
studies was conducted b 0om and 4riffeth! which was recentl updated *4riffeth et
al! +(((-. Their analsis confirmed some well2established findings on the causes of
turnover. These include5 /ob satisfaction! organisational commitment! comparison of
alternatives and intention to #uit. These variables are examined in more detail below!
as are a number of other factors where the evidence on the link to turnover is less
conclusive.
Comparison of alternatives
Aggregate level economic studies provide
consistent and significant evidence of the
impact of labour market conditions on
turnover rates at an aggregate level. As
Moble
et al *%&1&- pointed out! at an aggregate level the relationship between economic
factors such as emploment levels or /ob vacancies and turnover has been well
established. At an individual level! the labour market approach emphasises expected
utilit and rational economic choice among emploees and the perceived availabilit
of alternative /ob opportunities.
The relationship between alternatives and turnover on an individual level has been
researched widel since March 6 7imon8s %&'3 seminal work on ease of movement.
Much of the subse#uent research focused on the link between /ob satisfaction!
perceived alternative opportunities and turnover. 9ater! researchers began to focus on
the role of both actual and perceived opportunities in explaining individual turnover
decisions.
7ubse#uent research has indicated that actual alternatives are a better predictor of
individual turnover than perceived opportunities. Research on the impact of
unemploment rates as a prox for actual opportunities in emploee turnover revealed
that unemploment rates affected the /ob2satisfaction:turnover intent relationship but
not actual turnover *;irschenbaum 6 Mano2Negrin! %&&&-. The concluded that
macro level analsis predicted turnover patterns but perceptions of opportunities did
not. This point was reinforced in their stud on medical centres in various locations
used measures of perceived and ob/ective opportunities in internal and external labour
markets. The authors concluded that ob/ectives opportunities were a better set of
explanations of actual turnover behaviour than either perceived internal or external
labour market opportunities.
Nevertheless! while actual alternatives appear to be a better predictor of turnover!
there is also well2established evidence of the link between perceived alternatives and
actual turnover. In their most recent meta2analsis! 4riffeth et al *+(((- confirmed
that perceived alternatives modestl predict turnover.
Intentions to quit
Much of the empirical research on turnover is based on actual turnover! although
some studies are based on intentions to #uit. Apart from the practical difficult in
conducting turnover research among people who have left an organisation! some
researchers suggest that there is a strong link between intentions to #uit and actual
turnover.
Moble et al *%&1&- noted that the relationship between intentions and turnover is
consistent and generall stronger than the satisfaction2turnover relationship! although
it still accounted for less than a #uarter of the variabilit in turnover. Much of the
research on perceived opportunities has been found to be associated with intentions to
leave but not actual turnover *;irschenbaum 6 Mano2Negrin! %&&&-. <ne of the
possible reasons is that intentions do not account for impulsive behaviour and also
that turnover intentions are not necessaril followed through to lead to actual
turnover.
Organisational commitment
Man studies have reported a significant association between organisational
commitment and turnover intentions *9um et al! %&&3-. Tang et al8s *+(((- stud
confirmed the link between commitment and actual turnover and 4riffeth et al8s
*+(((- analsis showed that organisational commitment was a better predictor of
turnover than overall /ob satisfaction.
Researchers have established that there are different tpes of organisational
commitment. Allen 6 Meer *%&&(- investigated the nature of the link between
turnover and the three components of attitudinal commitment5 affective commitment
refers to emploees8 emotional attachment to! identification with and involvement in
the organisation= continuance commitment refers to commitment base on costs that
emploees associate with leaving the organisation= and normative commitment refers
to emploees8 feelings of obligation to remain with the organisation. >ut simpl!
emploees with strong affective commitment sta with an organisation because the
want! those with strong continuance commitment sta because the need to! and those
with strong normative commitment sta because the feel the ought to. Allen and
Meer8s stud indicated that all three components of commitment were a negative
indicator of turnover. In general! most research has found affective commitment to be
the most decisive variable linked to turnover.
Job satisfaction
The relationship between satisfaction and turnover has been consistentl found in
man turnover studies *9um et al! %&&3-. Moble et al %&1& indicated that overall /ob
satisfaction is negativel linked to turnover but explained little of the variabilit in
turnover. 4riffeth et al *+(((- found that overall /ob satisfaction modestl predicted
turnover. In a recent New .ealand stud! )oxall et al *+((,- found the main reason b
far for people leaving their emploer was for more interesting work elsewhere. It is
generall accepted that the effect of /ob satisfaction on turnover is less than that of
organisational commitment.
The link between satisfaction and commitment
7ome researchers have established a relationship between satisfaction! organisational
commitment and turnover. 9um et al8s *%&&3- stud of paediatric nurses suggested
that organisational commitment has the strongest and most direct impact on the
intention to #uit whereas /ob satisfaction has onl an indirect influence. The
suggested that satisfaction indirectl influences turnover in that it influences
commitment and hence turnover intentions. *Mueller 6 >rice! %&&( cited in 9um-.
$langovan *+((%- noted that the notion of /ob satisfaction and organisational
commitment being causall related has not been incorporated in most turnover
models. 0is stud indicated there were strong causal links between stress and
satisfaction *higher stress leads to lower satisfaction- and between satisfaction and
commitment *lower satisfaction leads to lower commitment-. 0e further noted a
reciprocal relationship between commitment and turnover intentions *lower
commitment leads to greater intentions to #uit! which in turn further lowers
commitment-. In summar! onl commitment directl affected turnover intentions.
Characteristics of employees
?espite a wealth of research! there appear to be few characteristics that meaningfull
predict turnover! the exceptions being age and tenure. Age is found to be negativel
related to turnover *i.e. the older a person! the less likel the are to leave an
organisation-. 0owever! age alone explains little of the variabilit in turnover and as
age is linked to man other factors! alone it contributes little to the understanding of
turnover behaviour.
Tenure is also negativel related to turnover *the longer a person is with an
organisation! the more likel the are to sta-. Mangione in Moble et al concluded
that length of service is one of the best single predictors of turnover. 4riffeth et al also
found that age and tenure have a negative relationship to turnover.
There is little evidence of a person8s sex being linked to turnover. 4riffeth et al8s +(((
meta2analsis re2examined various personal characteristics that ma be linked to
turnover. The concluded that there were no differences between the #uit rates of men
and women. The also cited evidence that gender moderates the age2turnover
relationship *i.e. women are more likel to remain in their /ob the older the get! than
do men-. The also found no link between intelligence and turnover! and none
between race and turnover.
Wages and conditions
The research conducted on the link between dissatisfaction with pa and voluntar
turnover appears to be inconclusive. Moble et al *%&1&- concluded that results from
studies on the role of pa in turnover were mixed but that often there was no
relationship between pa and turnover. <ther studies found no significant
relationship.
<n the other hand Campion *%&&%- cited in Tang suggests that the most important
reason for voluntar turnover is higher wages:career opportunit. Martin *+((,-
investigates the determinants of labour turnover using establishment2level surve data
for the @;. Martin indicated that there is an inverse relationship between relative
wages and turnover *ie establishments with higher relative pa had lower turnover-.
Pay and performance
4riffeth et al *+(((- noted pa and pa2related variables have a modest effect on
turnover. Their analsis also included studies that examined the relationship between
pa! a person8s performance and turnover. The concluded that when high performers
are insufficientl rewarded! the leave. The cite findings from Milkovich and
Newman *%&&&- that where collective reward programs replace individual incentives!
their introduction ma lead to higher turnover among high performers.
Flat-rate versus piece-rate pay systems
Taplin et al *+((,- conducted a large2scale turnover stud in the )ritish clothing
industr. Two factors emerged as the most significant reasons for emploees leaving
the industr. <ne was the low level of wage rates in the clothing industr relative to
other manufacturing sectors. The other reason referred to industr image with staff
leaving because of fears relating to the long2term future of clothing manufacture in
the @;. In this stud! turnover rates were highest among the most skilled workers.
The stud also examined the role of pament sstems in turnover. The researchers
found that where there were flat2rate pament sstems alone! average turnover
exhibited a statisticall significant difference from the industr mean *ie the were A.'
per cent lower-. 0owever! most firms in the clothing industr adhered to piece rate
pament sstems finding it to be the most effective wa of regulating the effort2
bargain. This is! in the authors8 view! despite anecdotal evidence that man skilled
workers dislike its unpredictabilit and new entrants to the workforce lack the skills to
maximise their earnings potential.
Attitudes to money
"or some individuals pa will not be the sole criterion when people decide to continue
within an existing /ob. In their stud of mental health professionals! Tang et al *+(((-
examined the relationship between attitudes towards mone! intrinsic /ob satisfaction
and voluntar turnover. <ne of the main findings of this stud is that voluntar
turnover is high among emploees who value mone *high mone ethic
endorsement-! regardless of their intrinsic /ob satisfaction. 0owever! those who do not
value mone highl but who have also have low intrinsic /ob satisfaction tended to
have the lowest actual turnover. "urthermore! emploees with high intrinsic /ob
satisfaction and who put a low value on mone also had significantl higher turnover
than this second group. The researchers also found that placing a high value of mone
predicted actual turnover but that withdrawal cognitions *ie thinking about leaving-
did not.
Training and career development
Martin *+((,- detected a complex relationship between turnover and training. 0e
suggested that establishments that enhance the skills of existing workers have lower
turnover rates. 0owever! turnover is higher when workers are trained to be multi2
skilled! which ma impl that this tpe of training enhances the prospects of workers
to find work elsewhere. The literature on the link between lower turnover and training
has found that off2the2/ob training is associated with higher turnover presumabl
because this tpe of training imparts more general skills *Martin! +((,-.
Impact of training on mobility
7hah and )urke *+((,- reviewed some of the literature on the relationship between
turnover and training. In a )ritish stud examining the impact of training on mobilit!
4reen et al *+(((- concluded that! in aggregate! training has on average no impact on
mobilit. 0owever! training that is wholl sponsored b the individual *or their
families- is on balance likel to be a prelude to /ob search. In contrast! when
emploers pa for training the downward effect on mobilit is more likel.
9nch *%&&%! %&&+- concluded that both on2the2/ob and off2the2/ob training have a
significant effect on /ob mobilit. Bhile formal on2the2/ob training reduces the
likelihood of mobilit! particularl for oung women! off2the2/ob training increases
the likelihood of mobilit. In a stud of six local labour markets in )ritain! $lias
*%&&A- found that women who received emploer2provided and /ob2related training
had a lower probabilit of changing emploer or making the transition to non2
emploment! but for men training made no significant difference to this tpe of
turnover.
Effect of vocational training
In a stud examining the effect of apprenticeships on male school leavers in the @;!
)ooth and 7atchell *%&&A- found that completed apprenticeships reduced voluntar
/ob2to2/ob! voluntar /ob2to2unemploment and involuntar /ob termination rates. In
contrast! incomplete apprenticeships tended to increase the exit rate to these
destinations relative to those who did not receive an training. Binkelmann *%&&C-
reported that in 4erman apprenticeships and all other tpes of vocational training
reduce labour mobilit in spite of the fact that the 4erman apprenticeship training is
intended to provide general and thus more transferable training.
Career commitment
Chang *%&&&- examined the relationship between career commitment! organisational
commitment and turnover intention among ;orean researchers and found that the role
of career commitment was stronger in predicting turnover intentions. Bhen
individuals are committed to the organisation the are less willing to leave the
compan. This was found to be stronger for those highl committed to their careers.
The author also found that emploees with low career and organisational commitment
had the highest turnover intentions because the did not care either about the
compan or their current careers.
Individuals with high career commitment and low organisational commitment also
tend to leave because the do not believe that the organisation can satisf their career
needs or goals. This is consistent with previous research that high career committers
consider leaving the compan if development opportunities are not provided b the
organisation. 0owever! this group is not apt to leave and is likel to contribute to the
compan if their organisational commitment is increased. Chang found that
individuals become affectivel committed to the organisation when the perceive that
the organisation is pursuing internal promotion opportunities! providing proper
training and that supervisors do a good /ob in providing information and advice about
careers.
Rural or remote areas and lifestyle factors
A +((% stud of the factors influencing the recruitment and retention of nurses in rural
and remote areas in Dueensland found that overall work2related factors were
considered to be more important in decisions b nurses to leave rural and remote
nursing practice. The five ma/or factors influencing decisions to leave rural or remote
area nursing practice were management practices! emotional demands of work!
workable communication! management recognition of work and famil
responsibilities.
The findings regarding lifestle factors appear to be mixed. <n the one hand! Erural
lifestle8 was ranked as the third most important factor for staing in rural and remote
practice and! similarl! Esense of belonging to the communit8 was ranked fifth.
0owever! when respondents were asked to identif the most important factors that
influenced them to leave rural and remote health services! /ust under A( per cent of
respondents cited issues related to the isolation caused b distance from basic
amenities as one inducement for them to resign. These issues included travelling long
distances to basic social and commercial activities! distance from famil! friends and
medical specialists! the comparativel high cost of living and a lack of
communication facilities such as the Internet! which would mitigate personnel and
professional isolation.
The emploment difficulties of non2metropolitan life were also a contributing reason
for nurses8 decisions to leave practice. "or example! unavailabilit of suitable
emploment or career development opportunities for their spouse was cited b +% per
cent of respondents. 7ome respondents with children *%C per cent- also identified
access to suitable education! childcare facilities and specialist medical expertise!
which are not usuall available outside densel populated areas! as factors influencing
leaving decisions.
Other factors contributing to turnover
Turnover studies have highlighted the
relationship between turnover and a range
of other factors. 7ome of these findings
are presented briefl below.
The role of shocks!
9ee 6 Mitchell8s *%&&A- Eunfolding model8 of emploee turnover represented a
significant departure from the previous labour market2 and pschological2oriented
turnover literature. This model is based on the premise that people leave organisations
in ver different was and it outlines four decision pathwas describing different
kinds of decisions to #uit. A notable feature of the unfolding model is its emphasis on
an event or Eshock8 *positive or negative- that prompts some decisions to #uit.
Morrell et al *+((A- tested the unfolding model b studing the voluntar turnover of
nurses in the @;. Their findings indicated that shocks pla a role in man cases where
people decide to leave. "urthermore! the found that shocks not onl prompted initial
thoughts about #uitting but also tpicall had a substantial influence over the final
leaving decision. The also noted that decisions to #uit prompted b a shock are
tpicall more avoidable. The authors suggest that their research illustrates the
importance for managers of understanding avoidabilit i.e. the extent to which
turnover decisions can be prevented.
"rganisational si#e
;irschenbaum 6 Mano2Negrin *%&&&- indicated that turnover is affected b
organisational siFe! with siFe being the ke mediator of an organisation8s internal
labour market. The suggest that organisational siFe impacts on turnover primaril
through wage rates but also through career progression paths. ?eveloped internal
organisational labour markets produce lower departure rates since promotion
opportunities have a strong negative influence on departures for career2related
reasons.
$nionisation
Martin *+((,- looked at the effect of unions on labour turnover and found clear
evidence that unionism is associated with lower turnover. 0e suggested that lower
turnover is a result of the abilit of unions to secure better working conditions thus
increasing the attractiveness for workers of staing in their current /ob. According to
Martin! the relationship between lower turnover and unionisation has been well
established b researchers using both industr2level and individual data.
Influence of co-workers
A +((+ stud b ;irschenbaum and Beisberg of A11 emploees in %' firms examined
emploees8 /ob destination choices as part of the turnover process. <ne of their main
findings was that co2workers8 intentions have a ma/or significant impact on all
destination options 2 the more positive the perception of their co2workers desire to
leave! the more emploees themselves wanted to leave. The researchers suggest that a
feeling about co2workers8 intentions to change /obs or workplace acts as a form of
social pressure or /ustification on the emploee to make a move.
%upervision&management
Moble et al *%&1&- concluded that a number of studies offered moderate support for
a negative relationship between satisfaction with supervision and turnover *ie the
higher the satisfaction with supervision! the lower the turnover-.
'ehavioural predictors
7ome research *0ulin cited in 4riffeth et al! +(((- implies that lateness and absence
can be predictors of turnover because the represent withdrawal responses from the
organisation.
Mobility in the Australian labour maret
7hah and )urke8s *+((,- report provides estimates of /ob and occupational labour
mobilit in Australia. It also provides information on the effects of demographic!
educational and labour market variables on the probabilit of /ob separation using
data from the A)7 Labour Mobility surve for +((+.
It is important to note that the A)7 surve does not directl measure turnover but
asks respondents whether the have changed /obs in the preceding %+ months. As
such! the concept of mobilit used in this stud is not directl comparable with
turnover but can arguabl give some indication of the nature of /ob separation in
Australia. In the A)7 surve! movers are defined as those people who stopped
working in a /ob sometime in the current ear! and can be /ob leavers or /ob losers.
7ome of the main conclusions of the report with relevance for MINTRAC are
outlined below.
Types of employment
>art2time workers have a much higher rate of /ob separation than full2time workers.
The effect is larger for men probabl because part2time work is likel to be casual and
hence short2term. At an given age! /ob separation is more likel for male part2time
workers than for female part2time workers.
(igrants
The chances of /ob separation are significantl higher for recentl arrived migrants
than for Australian2born workers. Recentl arrived migrants are less likel to have
established roots in a particular location or with a particular emploer and therefore
more likel to consider alternative /ob offers. Migrants from the main $nglish2
speaking countries had higher separation rates than other migrants! who presumabl
had a greater need to improve their $nglish language competenc! and to accumulate
work experience with a single emploer.
)egional areas
Borkers in non2metropolitan areas are significantl less likel to remain in the same
occupation than workers in metropolitan areas. This means that workers in non2
metropolitan areas either have higher chances of not being emploed or have higher
chances of moving to another occupation.
"ccupation
In general! the lower the skill levels of the occupation *short2term and casual /obs are
concentrated in lower2skill occupations- the higher the rate of /ob separation.
Retention strategies
!nderstanding turnover
The academic literature *?alton et al cited in Abelson! %&31- suggests that
differentiating avoidable and unavoidable turnover *from the organisation8s point of
view- can help organisations to understand voluntar turnover more full. Avoidable
reasons include emploees leaving to find better pa or working conditions elsewhere!
problems with management or leaving for better career opportunities. @navoidable
reasons 2 which are beond the organisationGs control 2 include! for example! an
emploee having to move because of relocation b a spouse or leaving to fulfil famil
or caring responsibilities.
If an organisation can identif that much of its voluntar turnover is unavoidable it
ma profit better from initiatives that seek to manage turnover after the event rather
than expend resources on implementing preventative measures. <n the other hand! if
the bulk of turnover is avoidable this offers the potential for targeted intervention.
0owever! if managers assume the turnover problem to be largel unavoidable! the
ma fail to recognise turnover as a smptom of underling problems within the
organisation.
*abour market issues
Another step towards understanding turnover within an organisation is to determine
whether retention difficulties are caused b internal or external factors. Bhile the role
of labour market conditions in causing turnover ma preclude the use of targeted
human resource strategies! this information ma be useful in analsing to what extent
turnover is due to outside factors. 0owever! although tight labour markets affect an
emploer8s abilit to attract and retain staff! looking outwards at the local labour
market cannot be a substitute for understanding what is going on within the
organisation *I?7! +(((-
Measuring turnover
Research suggests that to gain an accurate perspective of internal causes of turnover!
it is useful to look at both #uantitative and #ualitative information *I?7! +((A-. To
identif underling reasons for turnover! #ualitative information on the reasons wh
emploees have left is necessar. The @; Chartered Institute of >ersonnel and
?evelopment *CI>?- suggests that it is important emploers have an understanding of
their rates of labour turnover and how the affect the organisation8s effectiveness
*CI>?! +((A-. ?epending on the siFe of the business! understanding the levels of
turnover across occupations! locations and particular groups of emploees *such as
identified high performers- can help inform a comprehensive retention strateg. )
understanding the nature of the turnover problem an organisation can decide whether
to adopt targeted retention initiatives! for example at particular sites or groups of
emploees! or to manage overall levels so that there is sufficient labour.
Wastage rates
Tpicall! organisations use the crude wastage rate for measuring turnover. This
calculates the number of leavers in a given period as a percentage of the average
number of emploees during the same period. To calculate the average number of
emploees during a given period! organisations often add together the number of
emploees at the beginning of the period with those emploed at the end of the period
and divide it b two *I?7! +((A-.
Crude wastage rate
The simplicit of this measure means there is less risk of different parts of the
organisation suppling inconsistent data. 0owever! the crude wastage rate has its
limitations because it includes all tpes of leavers 2 involuntar leavers. A single
measure of turnover that does not distinguish between cases where people left because
the were dissatisfied and where people left because of ill health or retirement will be
inade#uate because it treats leavers as a homogeneous group *Morrell et al! +((A-.
"urthermore! when calculating replacement figures it ma be misleading to base them
on crude wastage rates which include emploees that do not need replacing *I?7!
+((A-. The crude wastage rate also makes no distinction between functional *ie
beneficial- and dsfunctional turnover *CI>?! +((A-.
)esignation rates
Another wa of measuring turnover is to base turnover rates on voluntar leavers or
resignation rates onl! thus excluding emploees who have left for other reasons such
as retirement! redundanc! dismissal or redeploment to another part of the
organisation. 0owever! basing turnover rates on voluntar leavers can also have its
drawbacks because it does not indicate how man staff need recruiting to cover those
emploees who have left because of retirement or voluntar internal transfers. <ne
solution is to record separate turnover rates for voluntar and involuntar leavers
*I?7! +((A-.
Bastage rates can also be used for specific groups of emploees or different business
units! which allows an organisation to detect differences in turnover within different
parts of the organisation. <verall figures tend to mask potentiall significant
differences in turnover within an organisation *I?7! +((A-. "or example! high
turnover in one area of the business could produce the same overall rate as a small
number of leavers distributed evenl across the organisation! but the actions re#uired
to deal with these situations would be #uite different. $xamining turnover b
department can identif an local issues or possible problems concerning particular
line managers or to monitor turnover among groups of emploees with scarce skills.
Bastage rates can also be applied to emploees with a certain length of service *eg
less than one ear- which can help pinpoint ineffective recruitment! selection or
induction processes *I?7! +((A-.
+acancy rate
Another approach is to place a greater emphasis on the number of vacancies that need
to be filled. The vacanc rate is based on the number of positions an organisation
activel wishes to recruit to as a percentage of the number of overall emploees.
%tability inde,
This measure gives an indication of the extent to which experienced emploees are
being retained. It can be used to calculate the stabilit of the whole organisation or of
a particular group of emploees. It is usuall calculated as the number of emploees
with one ear8s service or more as a percentage of the number of people emploed a
ear ago. This formula can be varied according to particular circumstances *eg basing
it on a longer period of service-. A rise in the stabilit index indicates the compan is
improving retention of more experienced staff. Normall! a wastage rate would be
expected alongside a low level of stabilit. If both percentages are high! this indicates
the organisation is experiencing problems with a small number of high turnover /obs
*I?7! +((A-.
Cohort analysis
This techni#ue enables an organisation to understand service2related leaving patterns
b taking the leaving rates of a *usuall homogeneous- group of emploees who
/oined at the same time. The resulting leaving rates can be plotted on a wastage or
survival curve. Cohort analsis is a useful tool for organisations concerned about
turnover costs due to high expenditure on recruitment! induction and training.
-astage and survival curves
"or wastage curves the number of leavers is plotted against their length of service on
leaving. A characteristic turnover pattern shows a high level for new starters! which
then decreases with length of service. Alternativel! survival curves represent the
number of people who sta against length of service! providing a measure of retention
instead of turnover.
"#it intervie$s and surveys
<rganisations tpicall obtain #ualitative information on turnover through exit
interviews and surves. 0owever! it is important to appreciate that the reasons people
give for their resignations are fre#uentl untrue or onl partiall true *CI>?! +((A-.
The use of exit interviews is widespread et the can be unreliable! particularl when
conducted b someone who ma later be asked to write a reference for the departing
emploee. Bhere exit interviews are used it is best to conduct them a short time after
the emploee hands in their notice. The interviewer should be someone who has not
had direct responsibilit for the individual *ie as their line manager- and who will not
be involved in future reference writing. Confidentialit should be assured and the
purpose of the interview explained *CI>?! +((A-.
Alternative approaches involve the use of confidential attitude surves which include
#uestions about intention to leave and #uestionnaires sent to former emploees on a
confidential basis about six months after their departure.
Costing turnover
The extent of the impact of turnover on an organisation cannot be full understood if
there is no attempt to #uantif the costs. The more complex approaches to costing
turnover give a more accurate and higher estimate of the costs. 7uch approaches often
take into account the costs associated with lost productivit *ie the productivit of a
new emploee during their first few weeks or months in the role and that of resignees
during the notice period- and the effect on morale of the remaining workforce.
<ne such framework is that proposed b TFiner and )irati *%&&C- which builds on the
earlier Cascio model of separation costs! replacement costs and training costs. The
authors demonstrate how their conceptual framework can be translated into a formula
and applied in practice. The TFiner and )irati framework includes5

direct costs incurred in the replacement process *recruiting! hiring! training and
socialising new emploees! including the extra effort b supervisors and co2
workers to integrate them
1 indirect costs and losses relating to interruptions in production! sales and the
deliver of goods to customers
2 financial value of the estimated effect on performance as a result of the drop in
morale of the remaining workforce following dsfunctional turnover.
Bhile such approaches are arguabl more accurate in that the cover all the costs
associated with turnover! in practice these can prove too complex and time2consuming
for man organisations. The @; Chartered Institute of >ersonnel and ?evelopment
*CI>?- suggests that because of the difficulties involved in estimating and
#uantifing some of the indirect costs man organisations prefer to take a Enot less
than8 approach in attempting to cost turnover. According to the CI>? *+((A-! it is
possible to compute a Enot less than8 figure b working out what it costs on average to
replace a leaver with a new starter in each ma/or emploment categor. This figure
can then be multiplied b the crude turnover rate for that emploee group to calculate
the total annual costs of turnover. The CI>? suggests that the ma/or turnover costs
are5
1 administration of the resignation *including exit interviews-
2 recruitment costs *including advertising-
3 selection costs
4 costs of cover *temporar emploees or overtime- during the vacanc period
5 administration of recruitment and selection process
6 induction training for new emploees.
Recruitment practices
In high turnover industries in particular! a great deal of emploee turnover consists of
people resigning or being dismissed in the first few months of emploment *CI>?!
+((A-. The costs of recruitment and turnover per individual become much greater
when new staff leave after onl a short period of time.
Bhere new emploees leave after a short period in the /ob! poor recruitment and
selection decisions both on the part of the emploer and emploee are usuall the
cause! along with poorl designed or non2existent induction programs *CI>?! +((A-.
If expectations are raised too high during the recruitment process this can result in
people accepting /obs for which the ma be unsuited. <rganisations often do this to
ensure that the fill their vacancies with sufficient numbers of well2#ualified people
as #uickl as possible. 0owever! this can be counterproductive over the longer2term!
as it can lead to costl avoidable turnover and to the development of a poor reputation
in local labour markets. The CI>? *+((A- suggests that emploers give emploees a
realistic /ob preview at the recruitment stage and take care not to raise expectations. It
ma also be useful to invite applicants to work a shift before committing themselves.
Retention strategies
I?7 *+((A- suggests that where there is a general turnover problem within the
organisation! companies often take a holistic approach aimed at full engaging with
staff. This ma encompass a wide range of measures such as5
improving recruitment procedures to ensure candidates receive a realistic
impression of the compan and the /ob
1 supporting new recruits during the critical first few weeks in the /ob
2 providing clear career paths! interesting work and support for personal
development
3 considering work2life balance issues
4 keeping pa in line with appropriate market rates
5 offering an attractive emploee benefits package
6 creating a pleasant working environment
7 communicating and consulting effectivel with emploees.
$xamples of such measures are illustrated in more detail in the industr and
organisational studies below.
Industry studies
Clothing industry
In Taplin et al8s *+((,- stud of the )ritish clothing industr! emploer initiatives to
reduce turnover included a range of measures. Improving remuneration packages was
the single most common initiative with changes to the paments sstems made so that
workers could increase their earnings. Also important were more rigorous screening
procedures for new hires! and improved training programs designed to bring new
workers up to speed so that the could maximise their piece rate earnings potential.
Harious #ualit of working life initiatives were introduced! such as flexible working
hours and forms of emploee participation *eg works councils-.
Accommodation industry
A recent stud *?avies et al! +((%- examined the effect of three human resource
strategies *performance appraisal! salar and benefit strategies-! and training and
development initiatives in the Bestern Australian accommodation industr. The
authors concluded that onl training and development indicated a reduced turnover of
emploees.
Mining
A stud of turnover in E"l2In "l2<ut8 *"I"<- mining operations in Australia *)each
et al +((,- showed that turnover rates between mines operating within the same
general labour market varied considerabl! suggesting that the main drivers of
turnover were often internal! rather than external to mine sites. In terms of managing
turnover! sites which had lower turnover! was attributed to a combination of four
factors5 e#uitable remuneration= commitment to training and skills development= good
management= and developing and maintaining a positive organisational culture.
Managers at some sites indicated the tried to control turnover b recruiting for a
good Eperson2organisation8 fit.
Case studies of organisations
The retention strategies outlined here are based on organisational2specific initiatives.
<ne of the problems with organisational2case studies is that their experience ma not
be generalisable! given that the causes of turnover and the resulting strategies are
likel to be specific to a particular organisation! a site or even a particular group of
emploees. Another problem is that where a combination of measures are used *which
is often the case- it is ver difficult to attribute success to one particular element of the
strateg. Nevertheless! some common practices can be drawn from these experiences
that appear to be successful in helping to improve retention.
The first two examples are academic studies while the following table summarises the
main features of the retention strategies adopted b a range of @; organisations to
successfull improve retention.
%ospital
A Canadian stud b 9um et al *%&&3- assessed the impact of certain pa policies
upon the turnover intentions of paediatric nurses. Two tpes of salar supplements
were introduced I bonuses to intensive care nurses onl and a ' per cent salar
differential for all staff nurses I to reduce turnover. The supplements were structured
in such a wa as to have the most favourable influence on the senior staff nurses who
were presumed to be the most experienced and those most valuable emploees.
7atisfaction with pa had both direct and indirect effects on turnover intent. The
found that although pa satisfaction *unlike /ob satisfaction- was significantl
associated with reducing intended turnover! its indirect effect upon turnover intent!
mediated through /ob satisfaction and organisational commitment was weaker. In
particular! nurses with greater experience were more satisfied with their pa and were
less likel to leave! which was the anticipated effect of the salar supplements. The
anecdotal evidence showed that the senior nurses perceived the pa supplements to be
an important recognition of their contribution to the organisation.
&epartment store
In a stud of retail salespeople *"irth et al! +((,- found that /ob stressors *eg work
overload! /ob ambiguit- are the factors that trigger the chain of pschological states
that lead to intention to #uit. The suggested that supervisor support can reduce the
impact of stressors on pschological states and intentions to #uit. Monitoring
workloads and supervisor2subordinate relationships b management ma not onl
reduce stress but increase /ob satisfaction and commitment to the organisation.
Causes of turnover and retention strategies adopted by
!' organisations
Organisation/ Retention problem/ Retention initiatives
industry causes
Standard Life High turnover in Guildford Streamlining recruitment and
Healthcare
1
office induction
Private medical Strategy based around Providing access to learning resource
insurance provider prioritising sales centres
750 employees Ineffectual leadership Competitive salary and benefits
eactive employee pac!age
resourcing "ocusing performance management
#ac! of customer focus on employee development
#o$ employee morale Improving communications
%easuring employee engagement
&ncouraging employee health and
$ell'being through targeted
interventions
Hillarys
1
etention of shopfloor (e$ employee benefits program $ith
)linds manufacturer employees at (ottingham a focus on family'friendly benefits and
*50 employees site lin!ing re$ard to length of service +eg
,issatisfaction $ith $or!ing increasing annual leave for longer
environment service-
,issatisfaction $ith Improving communications
communications ,evelopment of a ne$ internal brand
High number of .ob Improvements to the physical $or!ing
vacancies for lo$er's!illed environment +eg repainting and minor
$or!ers in one particular renovations-
area &ngaging employees through social
events and community activities
First Choice National High turnover/ particularly Increasing base pay in line $ith local
Sales Centre
1
among front'line sales staff call centre mar!et
0ravel call centre 1d hoc employee (e$ pay progressions system based
250 employees resourcing on achieving !ey competencies +$ith
3nderstaffing during reduced emphasis on sales-
operational pea!s Improving resource management +ie
)elo$ mar!et pay lin!ing recruitment activity to
)onus scheme too focused pea!s4troughs5 introducing annual
on sales targets hours-
Poor communications Streamlining recruitment and
0ight local call centre induction procedures/ lin!ing them to
labour mar!et the competency frame$or!
Happy
1
etain staff in I0 mar!et 6or!'life balance and fle7ible $or!ing
0raining company arrangements
87 staff &mpo$ering employees to ma!e !ey
decisions
Promoting an open culture
&ncouraging employee volunteering/
$hich offers development
opportunities
Kent County Council
1
#ocation +&ast 9ent (e$ competency'based career grade
5/:2; staff isolated/ 6est 9ent structure
pro7imity to #ondon- %ore coordinated recruitment
High turnover of social program +based on attracting ne$ly
$or!ers <ualified social $or!ers-
Heavy $or!loads and high Scheme to recruit and train
stress un<ualified social $or!ers
#imited opportunities for evie$ $or!loads and reduce
career progression administrative burdens
BAE Systems
1
39 graduate attrition Conducted research into
1erospace =uestions over effectiveness of graduate
c>25/000 +39- effectiveness of graduate development program
development program ealistic recruitment message to
emphasise the defence'related
manufacturing nature of its business
&ducating managers about their
responsibilities in the graduate
scheme
Seeboard Energy
:
High turnover in customer %ove to fi7ed shift patterns
3tility contact centre ecruitment targeted at $or!ing
,issatisfaction $ith 8'$ee! mothers to improve the response
rolling shift system
Marina e!elopments
8
' Introduced on'site/ internally
%arina group delivered training course for yard staff
:75 employees 0raining lin!ed to achievement of
competency targets and salary
increments
Ma"ita Manufacturing #ocal labour mar!et Improved induction process by setting
Europe
2
shortages of uns!illed and up a training scheme for ne$ recruits
Po$er tools semi's!illed staff $ith both off' and on'the'.ob training
manufacturing Culture of .ob'hopping in %ore targeted recruitment +ie
2;0 employees local area employees that can demonstrate
Ineffective induction some stability/ older $or!ers-
%ore stringent chec!ing procedures
during selection
%ore aptitude testing during
recruitment
e$arding long service
Asda
2
#ocal labour mar!et issues &7tended access to share schemes
etailer &7pansion highlighted through an employee share option
100/000 employees greater need to retain staff plan
%ore promotion options for staff
(e$ management development
program
#la$oSmithKline S!illed staff not returning eturners bonus upon return from
%esearch & after maternity leave maternity leave
e!elopment 'formerly ?pportunities for more fle7ible
#la$o (ellcome)
2
$or!ing/ including phased return to
2/;00 staff $or! and .ob'sharing
1ssistance in ma!ing childcare
arrangements
*rice+aterhouseCoop ,ifficulties in balancing Introduced $or!'life balance
ers Management $or! and home life among initiatives +eg $or!ing one day a
Consultancy Ser!ices
2
consultants $ee! at home-
8/200 staff 6or!'life balance $or!shops
Coaching scheme
Improve graduate retention
1
1dapted from@ I,S H Studies 7;5/ Improving staff retention/ Aanuary :002
:
Politt ,> +:008- BShift'pattern s$itch improves staff turnover and recruitment at SeeboardC
Human %esource Management ,nternational igest Dol> 11+1-/ pp1:'12>
8
Politt ,> +:002- B%arina ,evelopments staff get a lift from boatyard trainingC Human
%esource Management Dol> 1:/ (o> :/ pp12'1;>
2
1dapted from I,S Studies ;*:/ Improving staff retention/ Auly :000>
%igh performance $or practices and turnover
There have been a number of studies into the impact of high performance work
practices on a number of organisational outcomes. In a large2scale surve of 33' @7
firms! 0uselid *%&&'- concluded that the use of such practices had a statisticall
significant impact on turnover concluding that high performance work practices lead
to lower turnover. 0owever! a later stud of New .ealand firms came to a slightl
different conclusion. 4uthrie *+((%- suggested that the use of high performance work
practices ma have implications for the effect of turnover on productivit. 0is
explanation is that the use of such practices increases the value and importance of
human capital *ie emploees become more valuable to the organisation- and hence the
cost of emploee departures. A Canadian stud *7tatistics Canada! +((,- found that
the use of such practices appeared to be related to lower #uit rates in high2skill service
industries. 0owever! while there was some evidence of this link in lower skill
services! it was not as great as for high2skill services. The authors also found ver
little evidence that such practices reduce #uit rates in manufacturing. 0owever! one
particular practice I self2directed work groups I appeared to be associated with lower
#uit rates in manufacturing.
Conclusions
This review of turnover literature identifies a range of factors that have been shown to
be consistentl linked to turnover. These include organisational commitment! /ob
satisfaction! alternative opportunities and intentions to #uit. $vidence on the role of
pa is still somewhat inconclusive! although keeping pa in line with market rates is
certainl critical to retaining staff. Apart from age and tenure! personal characteristics
of emploees appear to have little relationship to turnover.
Bhile these factors can help emploers understand the general nature of turnover and
its likel causes! the retention strategies adopted within industries and organisations
tend to cover a uni#ue mix of measures and approaches specificall targeted at the
particular problem the face. @nderstanding the problem is ke to devising an
effective retention strateg. Access to both #uantitative and #ualitative data is
necessar for understanding levels of turnover across occupations! sites and for
particular groups of emploees as well as for identifing the underling causes of
turnover.
"urther investigation of turnover in the meat processing industr could include an
examination of turnover data to establish whether turnover is uniforml high across
the industr or whether there are differences between establishments in the same local
labour market. The collation of #ualitative data through emploee surves either at
industr level or within particular establishments ma be useful for identifing
sources of dissatisfaction! intentions to leave! and an underling causes of turnover.
Bibliography
Abelson M. A. *%&31-! E$xamination of avoidable and unavoidable turnover8 Journal
of Applied Psychology! vol. 1+*,-! pp. ,3+2,3C.
Allen N. J. 6 Meer J. >. *%&&(-! EThe measurement and antecedents of affective!
continuance and normative commitment on the organiFation8! Journal of
Occupational Psychology! vol. C,*%-! pp.%2%3.
)each R.! )rereton ?. 6 Cliff ?. *+((,-! Workforce turnover in FIFO mining
operations in Australia An e!ploratory study! Centre for 7ocial Responsibilit
in Mining! @niversit of Dueensland 6 7ustainable Minerals Institute.
http5::www.csrm.u#.edu.au. Accessed November +((A.
)oxall >.! Mack ;. 6 Rasmussen $. *+((,-! E9abour turnover and retention in New
.ealand= the causes and conse#uences of leaving and staing with emploers8!
Asia Pacific Journal of "uman #esources! vol.A%*+-! pp.%&C2+%A.
Chang $. *%&&&-! ECareer commitment as a complex moderator of organiFational
commitment and turnover intention8! "uman #elations! vol. '+ *%(-! pp.%+'12
%+13.
Chartered Institute of >ersonnel and ?evelopment *+((A-! "act sheet on emploee
turnover and retention.
?avies ?.! Talor R. 6 7aver 9. *+((%-! EThe role of appraisal! remuneration and
training in improving staff relations in the Bestern Australian accommodation
industr5 a comparative stud8! Journal of $uropean Industrial %raining!
vol.+'*C:1-! pp. ,CC2,1,.
$langovan A. R. *+((%-! ECausal ordering of stress! satisfaction and commitment! and
intention to #uit5 a structural e#uations analsis8! Leadership & Organi'ation
(evelopment Journal! vol.++*A-! pp.%'&2%C'.
"irth 9.! Mellor ?. J.! Moore ;. A. 6 9o#uet C. *+((A-! E0ow can managers reduce
emploee intention to #uitK8! Journal of Managerial Psychology!
vol.%&*+-! pp.%1(2%31.
4riffeth R. B.! 0om >. B. 6 4aertner 7. *+(((-! EA Meta2analsis of antecedents
and correlates of emploee turnover5 @pdate! moderator tests! and research
Implications for the next millennium8! Journal of Management
vol. +C*,-! pp.AC,2A33.
4uthrie J. >. *+((%-! E0igh2involvement work practices! turnover and productivit5
$vidence from New .ealand8! Academy of Management Journal vol. AA*%-!
pp.%3(2%&(.
0egne ?.! Rogers2Clark C.! 4orman ?.! )aker 7. 6 McCarth! A. *+((%-! Factors
influencing the recruitment and retention of nurses in rural and remote areas
in )ueensland* ?epartment of Nursing! @niversit of 7outhern Dueensland.
0uselid M. A.! *%&&'-! EThe impact of human resource management practices on
turnover! productivit and corporate financial performance8! Academy of
Management Journal! vol. ,3*,-! pp.C,'2C1+.
I?7 *+(((-! Improving staff retention! I(+ +tudies No. C&+! Jul +(((.
I?7 *+((A-! Improving staff retention! I(+ "# +tudies No. 1C'! Jan +((A.
I?7 *+((A-! Improving staff retention! I(+ "# +tudies No. 1C'! Jan +((A.
;irschenbaum A. 6 Mano2Negrin R. *%&&&-! E@nderling labor market dimensions of
LopportunitiesM5 The case of emploee turnover8! "uman #elations* vol.
'+*%(-! pp.%+,,2%+''.
;irschenbaum A. 6 Beisberg J. *+((+-! E$mploee8s turnover intentions and /ob
destination choices8! Journal of Organi'ational ,ehavior!
vol. +,*%-! pp.%(&2%+'.
9ee T. B. 6 Mitchell T. R. *%&&A-! EAn Alternative approach5 The unfolding model
of voluntar emploee turnover8! Academy of Management #evie- vol.%&*%-!
pp.'%23&.
9um 9.! ;ervin J.! Clark ;.! Reid ". 6 7irola B. *%&&3-! E$xplaining nursing
turnover intent5 /ob satisfaction! pa satisfaction or organiFational
commitmentK8! Journal of Organi'ational ,ehavior! vol.%&*,-! pp.,('2,+(.
Martin C. *+((,- E$xplaining labour turnover5 $mpirical evidence from @;
establishments8! Labour! vol.%1*,-! pp.,&%2A%+.
Moble B. 0.! 4riffeth R. B.! 0and! 0. 0. 6 Meglino! ). M. *%&1&-EReview and
conceptual analsis of the emploee turnover process8! Psychological
,ulletin vol. 3C *,-! pp.A&,2'++.
Morrell ;.M.! 9oan2Clarke J. 6 Bilkinson J *+((A-! E<rganisational change and
emploee turnover8! Personnel #evie-! vol. ,,*+-! pp.%C%2%1,.
>olitt ?. *+((,- E7hift2pattern switch improves staff turnover and recruitment at
7eeboard8! "uman #esource Management International (igest ! vol.%%*%-!
pp.%+2%A.
>olitt ?. *+((A- EMarina ?evelopments staff get a lift from boatard training8 "uman
#esource Management! vol.%+*+-! pp.%A2%C.
7hah C. 6 )urke 4. *+((,-! Labour mobility demographic* labour force and
education effects and implications for .$%! Report to ANTA! Centre for the
$conomics of $ducation and Training! Monash.
7tatistics Canada *+((,-! EInnovative work practices and labour turnover in Canada8
%he $volving Workplace +eries/
Tang T. 9. >.! ;im J. B. 6 Tang ?. 7. 0 *+(((-! E?oes attitude toward mone
moderate the relationship between intrinsic /ob satisfaction and voluntar
turnoverK8 "uman #elations! vol. ',*+-! pp.+%,2+A'.
Taplin I. M.! Binterton J. 6 Binterton R. *+((,-! E@nderstanding labour turnover in a
labour intensive industr5 $vidence from the )ritish clothing industr8 Journal
of Management +tudies! vol. A( *A-!pp.%(+%2%(AC.
TFiner A. 6 )irati ). *%&&C-! EAssessing emploee turnover costs5 A revised
approach8 "uman #esource Management #evie-! vol. C*+-! pp.%%,2%++.