What drives the intention of Bavarian crafts apprentices to change employer or occupation?

18 Pages·2013·0.56 MB·English

JLabourMarketRes.(2013)46:43–60
DOI10.1007/s12651-012-0109-7
ARTICLE
What drives the intention of Bavarian crafts apprentices to change
employer or occupation? An empirical study in the crafts sector
RobertWagner·MaximilianWolf
Accepted:9August2012/Publishedonline:7September2012
©InstitutfürArbeitsmarkt-undBerufsforschung2012
Abstract This paper combines several job related factors Keywords Jobrelatedfactors·Jobchange·crafts
such as occupational enjoyment, job security, financial at- apprentices
tractiveness,workingconditionsandsocialrelationsforthe
firsttimeandestimatestheirimpactontheintentionofcrafts JELClassification J24·J28·J62
apprentices to leave their training establishment or occu-
WelcheFaktorenbeeinflussendieAbsichtvon
pation. In contrast to previous analyses, we consider the
bayerischenHandwerkslehrlingendieFirmaoderden
dimension of the intended change by taking into account
Berufzuwechseln?EineempirischeStudieim
whether crafts apprentices intend to change establishment
Handwerkssektor
and/oroccupation.Wefindthatoccupationalenjoyment,re-
gional proximity to the employer and job security are the
Zusammenfassung IndieserStudieuntersuchenwir,wel-
most important drivers for the intention to stay with the
cheFaktorenimKontexteinerdualenAusbildungimbayeri-
training establishment. In contrast to other studies, mone-
schenHandwerkdieAbsichtvonAuszubildendenbeeinflus-
taryincentivesdonot drivetheintentiontoleavethetrain-
sen, nach Abschluss der Ausbildung entweder das Ausbil-
ingestablishmentoroccupation.Ourresultshavefarreach-
dungsunternehmenoderdenAusbildungsberufzuwechseln.
ingimplicationsfortrainingestablishmentsandinstitutions
Exemplarische Faktoren sind Freude an der Arbeit, regio-
in the Bavarian crafts sector concerning incentive systems, nale Nähe zum Ausbildungsbetrieb, Arbeitsplatzsicherheit
initial screening of apprentices, further training as well as oder das Verhältnis zu Kollegen und Vorgesetzten. Anders
improvementoftheimageofthecraftssectortothepublic als frühere Studien zu Job-Mobilität unterscheiden wir ex-
throughcraftschambers. plizitzwischendenMöglichkeitennurdasUnternehmenzu
wechseln,jedochdemAusbildungsberuftreuzubleibenund
derAlternative,sowohldasUnternehmenalsauchdenBe-
rufzuwechseln.BasierendaufeinemmultinomialenLogit
(cid:2)
Dipl.-Hdl.R.Wagner( )
ModellzeigenunsereErgebnisse,dassFreudeanderArbeit,
InstitutfürWirtschaftspädagogik,
regionaleNähezumArbeitgeberundArbeitsplatzsicherheit
Ludwig-Maximilians-UniversitätMünchen,
Geschwister-Scholl-Platz1,80539München,Germany die wichtigsten Faktoren sind, um weiterhin beim Ausbil-
e-mail:[email protected] dungsunternehmentätigseinzuwollen.UnsereErgebnisse
haben praktische Relevanz für zahlreiche Institutionen im
Dr.M.Wolf
Handwerk.
LehrstuhlfürProduktionswirtschaftundControlling,
Ludwig-Maximilians-UniversitätMünchen,Ludwigstr.28RG/V,
80539München,Germany
e-mail:[email protected] 1 Introduction
Dr.M.Wolf
In spite of intensive efforts by employers to reduce labour
Ludwig-Fröhler-Institut,Max-Joseph-Str.4,80333München,
Germany turnover, employees frequently change employer or career
44 R.Wagner,M.Wolf
(Neal 1999). Jovanovic (1979) argues that workers decide affect the intention to change employer or occupation than
tochangejobsinwhichtheirproductivityisrevealedtobe previousstudies.
low.Assumingestablishmentspayemployeestheirmarginal Ourresultshavepracticalimplicationsforseveralinstitu-
product,workersmoveoutofjobsiftheiroutsideoptionis tionsintheBavariancraftssector.1 Firstofall,theanalysis
better(McLaughlin1991)orthecurrentwageislowrelative of job related factors for apprentices allows training estab-
toalternativewages(Parsons1991).Therefore,wagesarea lishmentstodevelopsystematicstrategiestoretainahigher
key determinant in job changing decisions (e.g. Topel and proportionofapprenticeshipgraduates(e.g.initialscreening
Ward 1992; Fitzenberger and Spitz 2003). However, Free- of new apprentices). Second, organizations such as cham-
man(1978)identifiesjobsatisfactionasadrivingforcebe- bers(Handwerkskammer)thatareresponsiblefortheexecu-
hindthedecisiononfuturejobmobility.Jobsatisfactionin tionofapprenticeshiptraininginthecraftssectorcanmake
generalandseveraljobsatisfactiondimensionssignificantly use of the findings in order to adapt the dual apprentice-
influence an employee’s decision on job mobility (Clark ship systems to the needs of juveniles. Third, policy mak-
2001; Kristensen and Westergård-Nielsen 2004) as well as ers and the German Confederation of Skilled Crafts (Zen-
theintentionforjobmobility(ShieldsandPrice2002;Delf- tralverbanddesDeutschenHandwerks)cantacklethelack
gaauw2007;BöckermanandIlmakunnas2009).Thismeans ofskilledlabourbyenhancingtheimageofthecraftssector
that,evenifemployersfaceashortageofskilledlabour,they basedonourresults.
cannotpreventlabourturnoverifemployeesaredissatisfied. Theremainderofthispaperisstructuredasfollows.Sec-
All studies analysing the effects of pecuniary as well tion2providesanoverviewofthemostimportantliterature
as non-pecuniary job factors on job mobility focus on the onjob satisfactionand jobmobility.In Sect. 3, weprovide
labour market in general. As far as we know, the effect of anoverviewofthedatasetbeforeanalysingtheconnection
job related factors on the intention of crafts apprentices in betweenjobrelatedfactorsandjobmobilityinthenextsec-
Germany to change employer or occupation after appren- tions,usingamultinomiallogitmodel.Afterdiscussingthe
ticeship training has not yet been analysed. Low job sat- results,Sect.6summarizesthemostimportantfindingsand
isfaction during apprenticeship training could be one rea- providesaconclusion.
son for the low retention rate of around 30 % in the Ger-
man crafts sector after apprenticeship graduation (Harhoff
and Kane 1997; Büchel and Neubäumer 2001). On aver- 2 Background
age,establishmentsfromthecraftssectorpursueaproduc-
tivityorientatedtrainingstrategyandinvestlessintraining One strand of previous literature examines the impact of
than establishments from other sectors (Mohrenweiser and wages on job mobility. Topel and Ward (1992) identify
Backes-Gellner2008;MohrenweiserandZwick2009).Low the wages of skilled workers as the key determinant in job
investment in training decreases training quality (Soskice changingdecisions.FitzenbergerandSpitz(2003)showthat
1994), reduces apprentices’ job satisfaction and increases lowapprenticeshipwagesincreasetheprobabilityofchang-
involuntarylabourturnover.IntheGermancraftssector,in- ing occupation after apprenticeship graduation.2 However,
voluntarylabourturnoverleadstoalackofskilledworkers Freeman (1978) identifies job satisfaction as an economic
(Haverkampetal.2009). variableandshowsthattheeffectofjobsatisfactiononquit-
Inthispaper,weanalysewhichjobrelatedfactorsaffect tingremainsstableevenaftercontrollingforothereconomic
theintentionofcraftsapprenticestochangeemployeroroc- variables such as wages or working hours. According to a
cupation after finishing apprenticeship training. In contrast model developed by Neal (1999), a worker derives utility
to previous studies, we do not just estimate the intention from establishment and career matches. The first captures
for worker-initiated job mobility—we simultaneously esti- how well a worker interacts with his colleagues as well as
matewhetherapprenticeshipgraduatesintendtochangeem- how well the employee is suited to the environment cre-
ployer or occupation. Applying a multinomial logit model, ated by the employer. The career match captures how well
we show that occupational enjoyment, job security and re-
gional proximity to the current employer are the most im-
1Becauseofthehighspecificityofouranalysis,ourresultscannotbe
portant factors that affect the intention to change employer
generalizedorappliedtoothersectors.
oroccupation.Weuseanewdatasetcontaininguniquein-
2Other studies examine the effect of job and establishment changes
formation about future career plans and job related factors
of German apprenticeship graduates on wages in order to show the
driving the intention of apprentices to change employer or applicabilityofapprenticeshiptrainingcontentstothelabourmarket
occupation in the crafts sector. Our analysis contributes to (Dustmannetal.1997;HarhoffandKane1997;Werwatz1997;Ace-
mogluandPischke1998;ClarkandFahr2001;FitzenbergerandSpitz
the previous literature on job mobility by focusing on the
2003;BougheasandGeorgellis2004;EuwalsandWinkelmann2004;
specific group of German apprentices in the crafts sector. vonWachterandBender2006;GeelandBackes-Gellner2009;Göggel
Furthermore,wetestabroadersetofjobrelatedfactorsthat andZwick2012).
WhatdrivestheintentionofBavariancraftsapprenticestochangeemployeroroccupation? 45
the worker is suited to the type of work he or she is per- and personnel management are important. At the industry
forming.Iftheworkerhasthechancetoimproveoneofthe level, financial prospects, working conditions and unpleas-
matches,heorshewillchangeestablishmentorcareer.The antworkdutiesincreaseanemployee’sdesiretochangein-
definitions of establishment and career match allow non- dustry. Clark (2001) argues that the intention to quit might
pecuniary job related factors to affect a worker’s decision bejustadifferentwayofexpressingdissatisfactionandcrit-
onjobmobility. ically asks whether we can learn something from estimat-
Theliteratureevaluatingtheeffectofnon-pecuniaryjob ingtheeffectofjobsatisfactiondimensionsontheintention
related factors on job mobility shows that job satisfaction toquit.However,BöckermanandIlmakunnas(2009)show
affectsjobmobility.Clark(2001)usespromotionprospects, that job search, as an extreme form of a quit intention, in-
pay, hours worked, relations with supervisors, job security, creasestheprobabilityofjobmobility.Hence,jobdissatis-
ability to work on one’s own initiative and the actual work factionleadstoquitintentions,andtheseemployeessearch
itselfasjobsatisfactiondimensionsandarguesthatlowjob fornewmatchesmoreoften.6
satisfactionisagoodindicatorforpooremployer–employee Someofthejobrelatedfactorsinpreviousstudiesover-
matchquality.Heranksthejobsatisfactiondimensionsand lapacrossstudies,butthevastmajorityofjobrelatedfactors
shows that income and job security are the most important differ.7 Thismeansthatthereisnoconsistencywithregard
predictors of future quits.3 A comparable study was con- tojobsatisfactionorjobrelatedfactors.Previousstudiesuse
ductedbyKristensenandWestergård-Nielsen(2004).They different items for and different aspects of job satisfaction.
find comparable results; however, the ranking of the job Yet, the results are comparable in terms of low job satis-
satisfaction dimensions is different from the ranking sug- faction increasing job mobility. We argue that job satisfac-
gested by Clark (2001).4 In the paper by Kristensen and tioncanbeseenasajobrelatedfactoraffectingjobmobil-
Westergård-Nielsen,thetypeofworkisthemostimportant ity.Jobsatisfactionisasubjectivevariable(Freeman1978)
indicator of future quits. Mendius (2002) analyses which that captures what people think about their current em-
job related factors are especially important in leading to a ploymentmatch(KristensenandWestergård-Nielsen2004).
changeinemployerintheGermancraftssector.Conducting However, this subjective evaluation of the current employ-
expertinterviews,5heshowsthatlimitedcareerandtraining mentmatchcontainsseveraljobrelatedfactors(Kristensen
andWestergård-Nielsen2004).Therefore,wedefinejobsat-
prospects,physicalstress,workinghoursandlowjobsecu-
isfaction as all pecuniary as well as non-pecuniary job re-
rity compared with other sectors are important motives for
lated factors capturing the employment match that can be
skilled workers to leave. Interviews among German crafts
evaluatedbyworkers.Thisdefinitionincludesoccupational
workers reveal low salary and low job security as impor-
enjoyment, wages, career and training prospects, social re-
tantdrivingfactorsforthedecisiontoleavethecraftssector
lations, working conditions, working environment and job
(Haverkampetal.2009).
security.
In contrast to Clark (2001) as well as Kristensen and
We focus on the effect of job related factors on the in-
Westergård-Nielsen (2004), who estimate the effect of job
tentionofcraftsapprenticestochangeestablishmentoroc-
satisfaction on future quits, other studies measure the im-
cupation.Therefore,weusejobrelatedfactorsthatarespe-
pactofjobsatisfactionontheintentiontoquitandinitiating
cific for the crafts sector (e.g. Mendius 2002; Haverkamp
asearchforanewjob.Delfgaauw(2007)estimatesthelike-
et al. 2009) and important job related factors from studies
lihoodofaworkersearchingforanewjoboutsidethecur-
outside the crafts sector (e.g. Clark 2001; Kristensen and
rentorganizationaswellasoutsidethecurrentindustry.He
Westergård-Nielsen 2004; Delfgaauw 2007). None of the
identifies different job satisfaction dimensions driving the
studies presented above measures the effect of job related
intention to leave the establishment or the industry. At the
factors on the intention to change employer or occupation.
establishmentlevel,factorssuchasearnings,workpressure
6Furtherliteratureexaminingtherelationshipbetweenquitintentions
3Thisstatementisonlyvalidforthewholesample.Clark(2001)de-
andactualquitscanbefoundinDelfgaauw(2007).
rivesdifferentresultsbydividingthesampleaccordingtogender,age
7TopelandWard(1992)aswellasFitzenbergerandSpitz(2003)es-
andhoursworked.
timate the effect of wages on changing occupation. Clark (2001) as
4Clark (2001) uses data from the UK; Kristensen and Westergård-
wellasKristensenandWestergård-Nielsen(2004)operationalizedjob
Nielsen (2004) use data from Denmark. Kristensen and Westergård-
satisfactionbypromotionprospects,hoursworked,relationswithsu-
Nielsen(2004)explainthisdifferencebylowerunemploymentinsur-
pervisors,jobsecurity,abilitytoworkonone’sowninitiativeandthe
ancebenefitsintheUKcomparedwithDenmark.
actual work itself. Delfgaauw (2007) identifies earnings, work pres-
5447 crafts establishments, research institutes, government agencies, sureandpersonnelmanagementasfactorsinchangingemployerand
federal employment agencies and crafts organizations (Handwerk- financialprospects,workingconditionsandunpleasantworkdutiesas
skammern)participatedintheexpertinterviews.Theywereaskedto factorsinchangingindustry.AccordingtoMendius(2002),limitedca-
evaluatethefuturesupplyofskilledlabourincraftsestablishmentsus- reerandtrainingprospects,physicalstress,workinghoursandlowjob
ingaquestionnaire. securityarerelevantjobrelatedfactorsinthecraftssector.
46 R.Wagner,M.Wolf
Asfarasweknow,wearethefirsttodifferentiatebetween ing the crafts sector leads to a shortage of skilled labour
theintentionofapprenticestochangeemployeroroccupa- (Haverkamp et al. 2009). Therefore, we have variation in
tionafterapprenticeshipgraduation.Wearguethatasimul- intendedjobmobilityinthecraftssectorastherearealotof
taneous consideration of changing employer or occupation apprenticeswhostaywiththeirtrainingfirmbutalsoalotof
inoneestimationmodelisamorerealisticapproachtodeter- apprenticeswholeavetheirtrainingfirm.
minethespecificjobchangeintentionsofcraftsapprentices. Furthermore,trainingqualityvariesbetweentraininges-
Delfgaauw(2007)showsthatthereasonsforchangingem- tablishments and sectors. On the one hand, cost–benefit
ployeraredifferentfromthereasonsforchangingindustry. analysis of the Federal Institute for Vocational Educa-
The reasons for changing employer might also be different tion and Training (Bundesinstiut für Berufsbildung; BIBB)
fromthoseforchangingoccupation. shows that establishments in the crafts sector invest less in
To sum up, workers optimize their employer or career training than other sectors (Wenzelmann et al. 2009). Sec-
match by changing employer or occupation (Neal 1999). ond,asurveyamongaround6,000apprenticesinGermany
Empirical evidence shows that job related factors captur-
illustratesthattheperceivedtrainingqualitydiffersbetween
ing the current employment match are important determi-
occupationsandislowincraftsoccupations.However,there
nants in the intention to leave an establishment or industry
arealsoqualitydifferencesbetweencraftsoccupations(Be-
as well as for actual quits. Therefore, we deduce the hy-
ichtetal.2009).Consequently,weassumethatjobsatisfac-
pothesis that job related factors affect a crafts apprentice’s
tion among apprentices varies and affects their decision to
decision to leave the training firm or training occupation
change employer or occupation after apprenticeship train-
after finishing apprenticeship training. This means that the
ing.
higherthesatisfactionwithaspecificjobrelatedfactor,the
lower the probability of changing employer or occupation
afterapprenticeshipgraduation.However,thereasonsmight
3 Dataandsamplecharacteristics
differ depending on whether apprentices intend to change
employerand/oroccupation.
The dual apprenticeship system in Germany8 forms a Our data were collected in 2009 via online and paper-
good setting for our analysis. Crafts (Handwerk) estab- basedquestionnairesamongcraftsapprenticesinvocational
lishments, like other small and medium sized enterprises schools. As the Bavarian ministry of education supported
(SME),dependonskilledapprenticesinordertoensuretheir the data collection, we focused exclusively on apprentices
economic growth and handle the latest innovations (Kath inBavaria.ThecraftssectorinBavariaisquitestrongcom-
1996). The number of apprentices in crafts establishments pared with Germany as a whole (2009: 19.5 % of 975,316
traditionallyexceedsitsowndemandforskilledhumancap- crafts firms in Germany were located in Bavaria). Further-
ital, so that other sectors of the economy benefit from this more, apprenticeship training in the Bavarian crafts sec-
sourceofqualifiedlabour(SmitsandZwick2004).Costand tor plays an important role in professional training (2009:
benefitanalysesshowthattrainingabovedemandand,there- 18.5 % of the 461,502 crafts apprentices in Germany were
fore,ahighturnoverwasaprofitablestrategyforcraftses- trained in Bavaria) (Deutscher Industrie- und Handelskam-
tablishments(MohrenweiserandZwick2009;Wenzelmann mertag 2010; Zentralverband des Deutschen Handwerks
et al. 2009). Yet, in recent years, the lack of workers at an 2010).ThequestionnairesweresenttoBavarianvocational
intermediateskilllevelinGermancraftsoccupationshasin- schools,whichdistributedthemrandomlytocraftsappren-
creased.In1999,20%ofapprenticeshipgraduatesaged15– tices in specific occupations. In accordance with our re-
25yearsleftcraftsestablishments;in2006,thisproportion search,theBavarianvocationalschoolswereaskedtohand
increasedto50%(Haverkampetal.2009).Despitetraining out the questionnaires only to apprentices who were just
above demand being a profitable strategy for crafts estab-
abouttofinishtheirapprenticeshiptraining.Inaddition,vo-
lishments, the increasing number of skilled workers leav-
cational schools were asked to hand out the questionnaires
onlytoapprenticesfromsevenspecificcraftssectors(auto-
8The dual apprenticeship system is the main source of professional motive trade, crafts for commercial needs, finishing trade,
training in Germany: in 2008, around 50 % of men and women be- food trade, health trade, main construction trade and per-
tween30and35yearsofagehadsuccessfullycompletedapprentice-
sonal services). The apprentices were asked about their fu-
shiptraining,whereasonlyaround20%hadauniversitydegree;18%
turecareerplans,thereasonswhytheychoseanapprentice-
donothaveadegreeatall;12%hadcompletedaprofessionalschool
degree (Autorengruppe Bildungsberichterstattung 2010). Because of ship in a crafts occupation and to what degree specific job
itsefficiencyingeneratingqualifiedpersonnelaccordingtotheneeds relatedfactorsarefulfilledinthecraftssector.Allitemshad
oftheeconomy,ithasfrequentlybeenusedasarolemodelinother
tobeansweredonafive-pointLikertscale,where1stands
countries(e.g.Baileyetal.1992;GitterandScheuer1997;Harhoffand
for ‘does not apply at all’ and 5 means ‘applies to a very
Kane1997;Gospel1998;Steedmanetal.1998;HamiltonandHamil-
ton1999;Lehmann2000;Bosch2010. highdegree’.
WhatdrivestheintentionofBavariancraftsapprenticestochangeemployeroroccupation? 47
The questionnaire was designed based on the relevant Table1 Descriptivestatisticsonsamplecharacteristics
literature and discussed with several experts (such as re-
Variable Share(%) Sharein
searchers in the handicraft sector as well as apprenticeship Germany
training supervisors). Two pre-tests among eight appren- (%)
tices in the main construction trade guaranteed the com-
Male 77.02 84.98a
prehensibility and completeness of the questionnaire. Out
of 2,560 apprentices, 1,792 participated in the online sur- Noschooling 2.20 2.85b
vey (non-response rate of 30.0 %), and out of 734 appren-
Secondarygeneralschool 68.96 54.84
tices, 408 completed the paper based questionnaires (non-
Intermediateschool 24.80 36.74
responserateof44.4%).Afterremovingincompleteandin-
Grammarschool 4.04 5.56
consistentonlineaswellaspaper-basedquestionnairesand
questionnairesfromparticipantsinthesecondyearoftrain- Mainconstructiontrade 5.54 5.20
ing,wehaveasamplesizeofN =1,088. Finishingtrade 10.11 17.00
Descriptivestatisticsonindividual,occupationalandfirm Craftsforcommercialneeds 18.02 19.43
characteristics as well as the distribution of the factors in Automotivetrade 35.11 30.80
Germany can be found in Table 1.9 Table 1 shows that Foodtrade 11.12 6.27
the proportion of males in our sample is around 8 percent- Healthtrade 2.85 3.23
agepointslowerthanthefigureforGermany.Furthermore,
Personalservices 17.37 18.07
theproportionofcraftsapprenticeswhoattendedsecondary
grammarschoolishigherinoursamplethaninGermanyas Establishmentsize1–4 15.06 55.63
awhole.Thedistributionofoccupationalgroupsdiffersonly Establishmentsize5–9 21.85 23.24
slightlyfromthedistributioninGermany.However,thereis Establishmentsize10–19 19.38 12.36
alargebiasinestablishmentsize.Largeestablishmentsare Establishmentsize20–49 20.11 6.37
over-represented in our data, and only around 15 % of the Establishmentsize50–100 11.75
apprenticesworkinverysmallestablishments.However,the Establishmentsize101–499 8.36 2.40
majority(around55%)ofallcraftsapprenticesinGermany Establishmentsize500+ 3.49
workinSMEs. Samplesize=1,088
aOwncalculationsbasedonthenumberofapprenticesintherespec-
4 Measurementofjobmobilityandjobsatisfaction tivecraftssectors(N=192,241)on31December2010inGermany
(seehttp://www.zdh-statistik.de/application/stat_det.php?LID=1&ID=
MDE3OT%20E=&cID=,accessed13December)
Weusefouritemsregardingfurthercareerplanstocreatea
bAs there is no information about the educational background of
variableyindicatingtheworker-initiatedintentiontochange
recent apprentices in the relevant jobs, we present the number of
employer or occupation.10 All four items on further career new apprentices entering their training firm in 2010 (N =52,440)
planshadtoberatedonafive-pointLikertscale.Oneitem (seehttp://www.zdh-statistik.de/application/stat_det.php?LID=1&ID=
indicates the intention to stay with the training firm, two MDE4MzQ=&cID=00338,accessed13December2011)
itemstheintentiontochangeemployerbutstayinthesame
occupation, and one item indicates the intention to change iftheratingtostaywiththetrainingfirmishigherthanthe
employerandoccupation.11Thejobmobilityvariabley=1 ratingontheotherthreevariables,y=2iftheratingofone
of the two items on changing employer but staying in the
9Asnotalltheinformationabouttheapprenticesisavailableforthe same occupation is higher than the rating of the other two
Bavarianapprenticeshipsystem,wealternativelypresentthedescrip- variables, and y =3 if the rating of the item on changing
tivestatisticsforGermanytoshowtheappropriatenessofoursample. employer and occupation is higher than the rating on the
10Thesurveyitemsunderlyingthejobmobilityvariablecanbefound othertwoitems.12Some31.25%ofoursamplewanttostay
inAppendixAinEnglishaswellasGerman.Thefouritemsregarding
futurecareerplansdonotdirectlyaskfortheintentionofapprentices
tochangeemployeroroccupationbutfortheattractivenessofseveral 12Incaseofties(twoalternativeswereratedequally),wedecidedto
jobalternatives(seeAppendixA).Wearguethatapprenticesdonottry
puttheapprenticeintothehigherjobmobilitycategory.Thisallowsus
torealizeunattractivealternatives.Therefore,theattractivenessitems
tokeep294observationswhichwouldbelostotherwise.Toensurethat
mirroraconsiderablepartoftheintentionoftheapprenticestochange
ourfindingsarenotdrivenbythedefinitionofthevariable,werepeated
employeroroccupation.
ouranalyseswithallcasesthatcanbeuniquelydefined.(Weerasedall
11Otherquestionsonfuturecareerplanswereeitheraboutfurtheredu- 294 cases with ties). A multinomial logit estimation shows that the
cation(e.g.wanttogotouniversity)ortoounspecificforourpurposes resultsfromTable5onlymarginallychangeandthatthesignificance
(e.g.wanttolearnanewoccupationorwanttoworkinanewoccu- levelsareidentical(theadditionaloutputcanbeprovidedonrequest).
pation does not necessarily imply that participantswant to leavethe Weappliedtheidenticallogicalsoincaseswereallcategorieswere
trainingestablishment). ratedwiththevalues1or2.However,only13respondentsratedall
48 R.Wagner,M.Wolf
Table2 Crosstabulationofjobmobilitymeasures
Employmentoffer Jobofferfrom Communicated Noinformation
bytraining anotherfirm separationfromthe aboutjoboffersor
establishment training communicated
establishment separations
Intentiontostay 245 12 4a 79 340
withestablishment 22.52% 1.10% 0.37% 7.26% 31.25%
andkeepoccupation
Intentiontochange 144 31 14 190 379
establishmentbut 13.24% 2.85% 1.29% 17.46% 34.83%
keepoccupation
Intentiontochange 93 45 21 210 369
establishmentand 8.55% 4.14% 1.93% 19.30% 33.92%
occupation(change
career)
482 88 39 479 1,088
44.30% 8.09% 3.58% 44.03%
aThisfigureiscounterintuitiveatfirstglance.Onewouldexpectthat,ifanapprenticehasalreadytoldthetrainingestablishmentthatheorshe
wantstoleaveafterapprenticeshiptraining,theapprenticehasnointentionofstayingwiththetrainingfirm.However,thesefourcasescanbe
explainedbytemporaryseparationsfromthetrainingfirm(e.g.militaryorsocialservice).Theapprenticesleaveforacoupleofmonthsbutwantto
returnafterwards.Tocheckwhetherthefourobservationsaffecttheresults,wedeletedthesecasesandrepeatedouranalyses.Allresultsremained
constant
withtheirtrainingestablishment,34.83%intendtochange atelyaftergraduation.15Wearguethatthevariablezmirrors
establishmentbutstayinthesameoccupation,and33.92% theseparationintentionmorepreciselythanthejobmobility
wanttochangeemployerandoccupation(Table2).13 variabley.Ourdataallowustospecifythevariablezfor566
To verify whether the intended job mobility variable y respondents,becausearound44%oftheapprenticesinour
is a valid criterion for the worker-initiated change of em- sample had neither received an employment offer from the
ployer or occupation, we use information on employment trainingfirmoranoutsidefirmnorsaidthattheyintendedto
offersoftrainingaswellasoutsideestablishmentsandcom- leave the training establishment after graduation (Table 2).
municatedseparationsfromtheapprentices.14Basedonthis Some 15 % of the 566 respondents intended to leave the
information,wegenerateadummyvariablezindicatingan trainingestablishmentorhadjoboffersfromotherestablish-
impending separation and compare the variable z with the ments,andaround85%receivedanemploymentofferfrom
intendedjobmobilityvariabley.Table2showsthat44.30% the training establishment (Table 3). Table 3 shows a cross
of our sample had already received an employment offer tabulation for the variables of intended job mobility y and
fromthetrainingfirm,8.09%receivedanemploymentoffer impending separation z with absolute and relative values.
fromanoutsidefirm,and3.58%hadalreadytoldthetrain- Some 43.29 % of the reduced sample had an employment
ingestablishmentthattheywantedtoleaveimmediatelyaf- offer from the training establishment and intended to stay
tertraining.Thevariableforimpendingseparation z=0 if with the current employer after apprenticeship graduation.
theapprenticereceivedanemploymentofferfromthetrain- On the other hand, 78 out of 84 apprentices who either re-
ingestablishmentandz=1iftheapprenticeeitherhasajob ceivedanemploymentofferfromanoutsidefirmortoldthe
offerfromanoutsidefirmorhastoldthetrainingestablish- incumbentemployerthattheywishedtoleaveafterappren-
mentthatheorsheintendstoleavethetrainingfirmimmedi- ticeshipgraduationwanttochangeemployeroroccupation
afterapprenticeshipgraduation.Pearson’schi-squarestatis-
ticrevealsthattheassumptionofindependencybetweenthe
itemswithsuchlowvalues.Additionalanalysesshowthatallfindings variables z and y cannot hold.16 This means that the vari-
remainrobustaftererasingthese13cases.(Additionaloutputcanbe
providedonrequest).
13Ourdatadonotprovideinformationonchangingoccupationwithin 15Weexcludeapprenticeswhoreceivedanemploymentofferfromthe
thetrainingestablishment.AppendixDshowsdescriptivestatisticsfor trainingestablishmentaswellasfromanoutsidefirmasthesecases
allfoursub-groupsofTable2. cannot be uniquely assigned. However, if an apprentice received an
14Communicated separations from the training firm do not include employmentofferfromthetrainingfirmbutalsotoldthetrainingfirm
dropping out of the apprenticeship training. The apprentices finish thatheorsheleavesaftergraduation,thevariablez=1.
trainingbutintendtoleaveimmediatelyaftergraduation. 16Pearsonχ2(2)=70.79.
WhatdrivestheintentionofBavariancraftsapprenticestochangeemployeroroccupation? 49
Table3 Crosstabulationof
Employmentofferfrom Communicatedseparationfrom
intendedjobmobilityand
trainingestablishment thetrainingestablishmentorjob
impendingseparation
(z=0) offerfromanotherfirm(z=1)
Intentiontostaywith 245 6 N=251
establishmentandkeep 43.29% 1.06% 44.35%
occupation(y=1)
Intentiontochange 144 31 N=175
establishmentbutkeep 25.44% 5.48% 30.92%
occupation(y=2)
Intentiontochange 93 47 N=140
establishmentand 16.43% 8.30% 24.73%
occupation(y=3)
N=482 N=84 N=566
85.16% 14.84%
ablesofintendedjobmobilityyandimpendingseparationz The original questions as well as the corresponding items
measurecomparableseparationintentions.17 canbefoundinAppendixBinEnglishaswellasGerman.18
Incontrasttopreviousstudieswhichanalysetheconnec- We chose these 13 particular items for several reasons.
tion between job satisfaction and job mobility (e.g. Clark First,wechoseitemsthathavebeenusedinpreviousstud-
2001;KristensenandWestergård-Nielsen2004;Delfgaauw ies on job satisfaction and job mobility. Second, we argue
2007; Böckerman and Ilmakunnas 2009), our data provide that questions about why a specific occupation was chosen
nodirectinformationaboutjobsatisfaction.Thismeansthat are more closely related to the current situation of an ap-
there are no questions such as ‘How satisfied are you with prentice than questions about the crafts sector in general.
your job’. Instead, we use questions about the reasons for Therefore,ifquestionsinthetwoquestionblocksareredun-
choosingapprenticeshiptraininginaspecificoccupationas dant, we chose the question about the reason for choosing
well as questions about how certain criteria are fulfilled in an occupation and not the items about how certain criteria
thecraftssector.Therespondentshadtorate23reasonsfor arefulfilledinthecraftssector.19Fourth,alltheapprentices
choosinga specific occupationon a five-pointLikert scale. inoursampleareintheirthirdorfourthyearoftraining.We
These23questionsincludestatementssuchas‘Ichosethis arguethatratingsonallitemsareinfluencedbyexperience
occupationtotakeoverthefamilybusiness’,whichareirrel- gained during the training period. Therefore, if apprentices
evantforourpurposes.Otherquestionsaskforcomparable are satisfied with the items mentioned above, they choose
aspects (e.g. financial attractiveness or working hours) and ahighvalueontheLikertscaleirrespectiveofwhetherthe
are redundant as they are either part of the question block questions are retrospective or not. This means that, even if
aboutthereasonsforchoosingaspecificoccupationorpart ourquestionsdonotspecificallyaskforjobsatisfaction,the
ofthe16questionsabouthowcertaincriteriaarefulfilledin ratingsoftheparticipantsarecloselyrelatedtojobsatisfac-
thecraftssector. tion.Fortheremainderofthispaper,wecalltheseitemsjob
In a first step, we conducted an explorative factor anal- relatedfactors.Adescriptiveoverviewofthejobrelatedfac-
ysestoreducethenumberofdimensions.However,wedid torsaswellastheproportionofrespondentswhochosea4
not find factors with an intuitive interpretation. Therefore, or5ontheLikertscalecanbefoundinTable4.
we selected five items from the question block about why ThemeansofalljobrelatedfactorsareabovetheLikert
aspecificoccupationwaschosen(occupationalenjoyment, scalemiddlevalueof3,andwehavevariationinallitems.
imageofthetrainingestablishment,jobsecurity,promotion In all but two criteria, more than 50 % of the respondents
prospectsandfinancialattractiveness).Anothereightitems choseatleastthevalue4onthefive-pointLikertscale.The
werechosenfromthequestionblockabouthowspecificfac- figuresinTable4suggestthat,onaverage,craftsapprentices
tors are fulfilled in the crafts sector (relation with supervi- inBavariaaresatisfiedwiththeircurrentsituation.
sor, relation with colleagues, regional proximity of estab-
lishment,physicallyhardwork,noroutineinwork,further 18AppendixCshowsthecorrelationsbetweenthechosenitems.
training, fixed working hours and flexible working hours). 19Tomakesureourresultsarenotcausedbytheitemswechose,weran
ourregressionswithdifferentitemsforidenticalfactors.Forinstance,
weusedfivedifferentitemsthatallcapturefinancialattractivenessand
17Ourdataprovidenoinformationonrealizedjobmobility.Therefore, repeatedourregressions.Wefoundnoeffectirrespectiveofwhichitem
wecannotsaywhethertheparticipantswereabletorealizetheirinten- weused.Weusedthesamemethodforworkinghours,regionalprox-
tionsafterapprenticeshipgraduationornot. imitytoworkplaceandoccupationalenjoyment.
50 R.Wagner,M.Wolf
Table4 Descriptivestatistics
Variable Mean SD Percentage
onjobrelatedfactors
choosing4
or5
Occupationalenjoyment 4.238 0.897 85.48
Imageoftrainingestablishment 3.798 1.039 69.67
Financialattractiveness(salary) 3.431 1.162 52.57
Promotionprospectsincrafts 4.001 1.010 74.17
Thecolumn‘Percentage Jobsecurityincrafts 4.072 1.003 76.93
choosing4or5’showsthe Relationwithsupervisorinestablishment 3.388 0.988 47.34
proportionofparticipantswho
Relationwithcolleaguesinestablishment 3.947 0.997 71.60
chosethevalue4or5onthe
Likertscale.Forthemultivariate Regionalproximitytoestablishment 3.608 0.987 55.88
analyses,wetransformedthe Physicallyhardworkincrafts 3.986 1.054 72.24
Likertscaledvariablesinto
Noroutineinworkincrafts 3.883 1.069 68.84
dummies.Thedummiesequal1
ifaparticipantchosethevalue4 Furthertrainingpossibilities 3.055 1.217 36.40
or5ontheLikertscaleand0 Fixedworkinghours 4.200 1.024 78.31
otherwise.Wearguethat,ifa
Flexibleworkinghours 3.958 1.071 70.13
participantchoseatleastthe
value4,heorsheissatisfied Samplesize=1,088
withthejobrelatedfactor
5 Estimationmodelandfindings ventory.22Dohmenetal.(2005)showthatattitudestowards
taking risks in general as well as taking risks in one’s ca-
To estimate the effect of job related factors on the inten- reerdiffersignificantlybygender.Womenhaveahigherrisk
tion to change employer or occupation, we apply a multi- aversionthanmen.Ahighriskaversionmightnegativelyin-
nomial logit model. The dependent variable y has m=3 fluence the decision to change firm or job. We try to take
mutually exclusive and non-ordered categories20 (Winkel- riskaversionintoaccountbycontrollingforgender.Büchel
mannandBoes2009).Wechoose y=1 (‘intentiontostay and Neubäumer (2001) show that apprenticeship graduates
with training firm and keep occupation’) as the reference with a low educational background have a lower probabil-
category. Then, we calculate regressions for the remaining ityofworkinginanoccupationcomparabletotheirtrained
m−1 categories(y=2 ‘intentiontochangefirmbutkeep occupation. This means that apprenticeship graduates with
occupation’ and y =3 ‘intention to change firm and occu- a low educational background have a higher probability of
pation’)inreferencetoy=1.Thecoefficientsβmk andγml changingoccupationthanmorehighlyeducatedapprentice-
show changes in the log odds for the categories y =2 and ship graduates. Franz and Zimmermann (1999) find a pos-
y =3 relative to the base category y =1.21 The multino- itive correlation between establishment size and retention
miallogitmodelallowsfordifferentalternativecoefficients rate, and Harhoff and Kane (1997) show that the immedi-
depending on which of the remaining m−1 categories is atedeparturerateofapprenticeshipgraduatesdiffersbysec-
compared with the base category y =1. This means, β2k tor.Thismeansthatestablishmentsizeaswellasthesector
candifferfromβ3k. mightinfluencethedecisiontochangeemployerand/oroc-
(cid:2) (cid:3)
P(y =m) (cid:4)K (cid:4)L cupation. As our analysis focuses on the crafts sector, we
ln i =α + β X + γ W do not differentiate by sector but control for occupational
P(y =1) m mk ik ml il
i k=1 l=1 groupswithinthecraftssector.Theoccupationalgroupscan
X representsavectorofalljobrelatedfactorspresented befoundinTable1.
inTable4andW isavectorcapturingallotherfactorsthat All factors in vector W mightalso affect the job related
mightinfluencethedecisiontochangeemployeroroccupa- factors in vector X. For instance, the image of the training
tionforeachindividuali.Inthispaper,W includessex,edu- firm might correlate with firm size. Thus, the estimates for
cationalbackground,firmsize,occupationandabig-fivein- thejobrelatedfactorsinvectorXmightbebiasedifwene-
glectthefactorsinvectorW.Weconductedseveralregres-
20As the dependent variable is characterized by non-ordered cate-
gories,wedonotchooseorderedresponsemodelssuchastheordered 22The questionnaire included a 10-item short version of the big-five
probitmodel. inventory.Itmeasurespersonalitywith2itemsperscale(extraversion,
21Therefore,allrawestimatesinTable5havetointerpretedrelativeto agreeableness,conscientiousness,neuroticismandopenness)(Ramm-
thebasecategoryy=1. stedtandJohn2007).
WhatdrivestheintentionofBavariancraftsapprenticestochangeemployeroroccupation? 51
Table5 Rawestimatesaftermultinomiallogit on theintentionto changeemployeras well as occupation.
Variables ln(P(yi=2)) ln(P(yi=3)) However,apprenticeswhoworkinlargecompaniesareless
P(yi=1) P(yi=1) likely to change firm and occupation. Furthermore, our re-
Occupationalenjoyment –0.288 –1.001∗∗∗ sults show that apprentices in food trade occupations are
(0.305) (0.283) morelikelytochangeoccupationthanapprenticesfromper-
Imageoftrainingfirm –0.572∗∗∗ –0.536∗∗∗ sonal services and that a high degree of conscientiousness
(0.197) (0.207) negativelyaffectsthedecisiontochangeemployeroroccu-
Relationwithsupervisor –0.373∗∗ –0.567∗∗∗ pation.24
(0.180) (0.193)
Occupational enjoyment has a significant negative ef-
Relationwithcolleagues 0.0135 –0.201
fect on the intention to change occupation but no effect on
(0.227) (0.230)
changingoccupation.Thisfindingisveryintuitive.Ifanap-
Regionalproximitytofirm –0.103 –0.679∗∗∗
prentice does not like his or her occupation, it would not
(0.177) (0.185)
make sense to change employer but stay in the same oc-
Physicallyhardworkincrafts 0.0893 0.189
(0.198) (0.211) cupation. The image of the training firm as well as the re-
Noroutineinworkincrafts 0.368 0.135 lation with the supervisor have significant negative effects
(0.204) (0.210) on the intention to change employer and occupation. Re-
Furthertrainingpossibilities –0.164 –0.0450 gionalproximitytotheincumbentemployerhasanegative,
(0.174) (0.191) yet insignificant, effect on changing employer but a signif-
Fixedworkinghours –0.0451 –0.0628 icant negative effect on the intention to change employer
(0.251) (0.255)
andoccupation.Jobsecurityhasasignificantnegativeeffect
Flexibleworkinghours –0.261 –0.344 onchangingemployerandoccupation.Promotionprospects
(0.225) (0.229)
have a negative effect on the intention to change occupa-
Jobsecurity –0.559∗∗ –1.026∗∗∗
tion, but no significant effect on changing employer. This
(0.242) (0.240)
result might be caused by limited promotion prospects in
Promotionprospects 0.0355 –0.452∗∗
certainoccupations.Therefore,badpromotionprospectsin-
(0.224) (0.220)
volvechangingoccupationaswellasemployer.
Financialattractiveness(salary) 0.154 0.0622
(0.168) (0.183) Surprisingly,financialattractivenessdoesnothaveasig-
Sex Yes Yes nificant effect on the intention to change employer or oc-
Schooling Yes Yes cupation. Previous studies show that low wages are one of
themostimportantdriversforjobmobilityafterapprentice-
Firmsize Yes Yes
ship training (Fitzenberger and Spitz 2003). In contrast to
Occupation Yes Yes
otherstudies,weuseanitemaskingfortheperceivedfinan-
Big-fiveinventory Yes Yes
Constant 0.781 1.496∗ cial attractiveness of a particular crafts occupation instead
(0.826) (0.859) of using real apprenticeship wages.25 Therefore, measure-
LRχ2(72) 280.25 ment errors might be one explanation for our insignificant
PseudoR2 0.1173 findings. However, it is more likely that our findings are
driven by our homogeneous sample. We only look at one
Observations 1,088
specific sector (crafts) and one specific group (apprentices
in Bavaria). Fitzenberger and Spitz (2003) use a sample of
sions and added the control variables from vector W step-
WestGermanmalesfromdifferentsectorsandoccupations.
wise.ThecoefficientsforthejobrelatedfactorsinvectorX
Theyshowthatapprenticeshipgraduateschangeoccupation
remainverystableafteraddingadditionalcontrolvariables
whenaverageapprenticeshipwagesinthetrainingoccupa-
fromvectorW,indicatingalowcorrelationbetweenthefac-
tionarelow.Moreoccupationsandsectorsimplymorevari-
torsinXandW.Therefore,weonlyshowtheresultsforour
ationinapprenticeshipwagesbecausethereareconsiderable
full specification.23 The raw estimates for the multinomial
wage differences between occupations and sectors (BIBB
logitspecificationcanbefoundinTable5.
2009).Asweonlyfocusononesectorandacoupleofoc-
We find no gender effect but positive effects for edu-
cupations,thevariablemeasuringfinancialattractivenessis
cational background and a negative establishment size ef-
unlikelytohaveacomparablevariationtothewageinforma-
fect.Ahighereducationalbackgroundhasapositiveeffect
tionintheanalysisbyFitzenbergerandSpitz(2003).There-
fore,itisverylikelythatfinancialattractivenesswouldhave
23Thedatadonotprovideinformationaboutwhetherapprenticesare
trainedinthesametrainingestablishmentornot.Therefore,wecannot
controlforspecificfirmdummiesorclusterthestandarderrorsatthe 24ResultsarenotshowninTable5butcanbeprovidedonrequest.
trainingestablishmentlevel. 25Ourdataprovidenoinformationaboutapprenticeshipwages.
52 R.Wagner,M.Wolf
Table6 Predictedprobabilitiesforjobrelatedfactors
Intentiontostaywith Intentiontochangefirmbut Intentiontochangefirm
trainingfirmandkeep keepandoccupation(%) occupation(%)
occupation(%)
Sampleaverage 29.12∗∗∗ 37.96∗∗∗ 32.92∗∗∗
(0.016) (0.017) (0.016)
Occupationalenjoyment=0 18.77∗∗∗ 31.29∗∗∗ 49.94∗∗∗
(0.310) (0.044) (0.047)
Occupationalenjoyment=1 30.98∗∗∗ 38.73∗∗∗ 30.28∗∗∗
(0.018) (0.018) (0.017)
Discretechangeinthepredicted 12.21 7.44 –19.66
probabilityforoccupational
enjoyment∗
Regionalproximity=0 24.80∗∗∗ 34.29∗∗∗ 40.97∗∗∗
(0.023) (0.025) (0.026)
Regionalproximity=1 32.43∗∗∗ 40.40∗∗∗ 27.17∗∗∗
(0.022) (0.023) (0.021)
Discretechangeinthepredicted 7.63 6.11 –13.8
probabilityforregional
proximity∗
Jobsecurity=0 18.21∗∗∗ 36.47∗∗∗ 45.32∗∗∗
(0.029) (0.023) (0.038)
Jobsecurity=1 32.92∗∗∗ 37.72∗∗∗ 29.36∗∗∗
(0.019) (0.019) (0.018)
Discretechangeinthepredicted 14.71 1.25 –15.96
probabilityforjobsecurity∗
AllcalculationsarebasedonestimatesinTable4(N=1,088)
Allexplanatoryvariablesarefixedattheirmeans
Standarderrors(afterdelta-method)inparentheses:∗∗∗p<0.01,∗∗p<0.05
∗Discretechangeinthepredictedprobabilitylevelsshowsthechangeinthepredictedprobabilitylevelsinpercentagepointsifthecorresponding
dummychangesfrom0to1
an effect on job mobility in an inter-sector analysis but in abilityofaround29.12%.Thisfigureisconsistentwithan
notinanintra-sectoranalysis.26 immediateaverageretentionrateofroughly30%ofappren-
To get a feeling for the economic relevance of the job tices in the crafts sector (Harhoff and Kane 1997; Büchel
relatedfactorsoccupationalenjoyment,jobsecurityandre- andNeubäumer2001).However,ifapprenticesaredissatis-
gionalproximitytothecurrentemployer,Table6illustrates fiedwithoccupationalenjoyment,theprobabilityofstaying
howtheprobabilityvaluesofthethreecategoriesofthejob with the training firm is just 18.77 %. However, if appren-
mobility variable y change if one or more of the dummy ticesarehappywithoccupationalenjoyment,theprobability
variablesmeasuringtheratingofjobrelatedfactorschanges tostaywiththetrainingfirmsincreasesby12.21percentage
from0to1.TheresultsinTable6allowustodrawimportant pointsandtheprobabilitytochangefirmandoccupationde-
conclusions about the job mobility intentions of crafts ap- creases by 19.66 percentage points. The smallest discrete
prentices.ThefirstlineinTable6showstheprobabilityval- changesforthepredictedprobabilitiescanbefoundforthe
uesforthesampleaverage.Theprobabilitylevelsofallthree
intentiontochangefirmbutkeepoccupation,rangingfrom
categories of the job mobility variable y are around one-
1.25 to 7.44 percentage points. Especially the effect of job
third.Thismeansthattheaverageapprenticeinoursample
security on the intention to change firms but keep occupa-
intendstostaywiththetrainingestablishmentwithaprob-
tion(1.25percentagepoints)isofhardlyanyeconomicrel-
evance.However,theidentifieddiscretechangesforthere-
26Anotherexplanationcouldbethatworkerscomparetheirwageswith mainingtwocategoriesrangefrom14.71to−19.66percent-
areferencegroupofco-workers.Inmostcases,therearenogreatdif- agepoints.Therefore,theidentifiedeffectsforoccupational
ferencesinthewagesofapprentices,buttherearedifferencesinthe
enjoyment,jobsecurityandregionalproximityhaveahigh
wagesofskilledworkers.Consequently,thewageargumentmaynot
economicrelevancefortheintentiontochangeemployeror
berelevantattheendoftheapprenticeshipbutatthebeginningofthe
firstjobasskilledworkers. occupation.