Economic Modelling 126 (2023) 106437
A ailable online 13 July 2023
0264-9993/© 2023 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/).
Does employmen p o ec ion legisla ion a ec employmen and
unemploymen ?
1
Philip A es is
a
, Jesus Fe ei o
b
,
*
, Ca men Gomez
b
a
Uni e si y o Camb idge, Uni ed Kingdom
b
Uni e si y o he Basque Coun y UPV/EHU, Spain
ARTICLE INFO
Handling Edi o : Sushan a Mallick
JEL classi ica ion:
E24
J21
J41
Keywo ds:
Employmen p o ec ion legisla ion
Employmen
Employees
Unemploymen a e
ABSTRACT
The a icle analyses he impac o employmen p o ec ion legisla ion (EPL) on labou ma ke ou comes. Despi e
widesp ead e o ms ha ha e educed employmen p o ec ion, he e idence on he e ec s o such e o ms is
inconclusi e. Using da a om six een Eu opean coun ies o e he pe iod 1985–2019, we analyse he impac o
EPL on he dynamics o employmen , employees and unemploymen a es. In con as o exis ing s udies, we
analyse bo h he exis ence o a linea ela ionship be ween EPL and labou ma ke ou comes and he exis ence o
a non-linea ela ionship, as well as in e ac ion e ec s be ween EPL and economic g ow h. Ou esul s show ha
employmen p o ec ion does no explain he changes in employmen , employees and unemploymen a es.
The e o e, labou e o ms ha ha e educed employmen p o ec ion by educing dismissal cos s and acili a ing
he use o empo a y con ac s ha e no had he p esumed posi i e e ec s on employmen and unemploymen
a es.
1. In oduc ion
Acco ding o New Keynesian Economics, employmen and unem-
ploymen ou comes a e explained by he in e ac ion o economic shocks
wi h labou ma ke ins i u ions. Poo job c ea ion and high unemploy-
men a es would be gene a ed by he combina ion o low a es o eco-
nomic g ow h and ine icien -unp oduc i e labou ma ke ins i u ions
ha gene a e igidi ies in he unc ioning o labou ma ke s (Blancha d
and Wol e s, 2000). On he con a y, coun ies wi h e icien labou
ins i u ions, i.e. mo e lexible labou ma ke s, would ha e he bes
employmen and unemploymen ou comes. The policy ecommenda-
ions a e ob ious: in o de o enjoy low and s able unemploymen a es,
labou ma ke s should be e o med o make hem mo e lexible by
add essing hose legal and ins i u ional elemen s ha gene a e igidi ies
in he wage-se ing p ocess and in he adjus men o i ms’ wo k o ces.
Spu ed by hese a gumen s and he ecommenda ions o in e na-
ional o ganisa ions such as he Eu opean Commission, he In e na ional
Mone a y Fund and he O ganisa ion o Economic Co-ope a ion and
De elopmen (OECD), many coun ies ha e adop ed e o ms o make
hei labou ma ke s mo e lexible wi h he aim o educing unem-
ploymen a es in he long un. These e o ms a ge ed wha we e
conside ed o be he main sou ces o labou ma ke igidi ies: unem-
ploymen bene i sys ems, collec i e ba gaining and employmen p o-
ec ion legisla ion (B ancaccio e al., 2018; Geh ke e al., 2019; Kugle ,
2019; McB ide and Wa son, 2019; T idico and Pa iboni, 2017).
Howe e , he e idence on he impac o labou ins i u ions on
employmen and unemploymen is inconclusi e (A dagic and Sala di,
2013; Be ola, 2017; Kugle , 2019). Fo pos -Keynesian economis s, la-
bou ins i u ions a e no a key de e minan o labou ma ke ou comes
and only an inc ease in capi al accumula ion, uelled by expansiona y
demand-side policies, inc eases employmen and educe unemploymen
a es (Gi a di e al., 2020; Hein, 2017; S ockhamme e al., 2014). This
ecommenda ion is sha ed by mains eam economis s, such as Ball
(2009, 2014) and Blancha d and Summe s (2017), who a gue ha he
high unemploymen a es in many Eu opean coun ies can be explained
by he hys e esis e ec s gene a ed by es ic i e demand-side policies
and ha a change in he ele an mac oeconomic policy s a egies is
he e o e necessa y.
* Co esponding au ho . Depa men o Public Policies and Economic His o y, Facul y o Economics and Business, Uni e si y o he Basque Coun y UPV/EHU,
A enida Lehendaka i Agi e 83, 48015, Bilbao, Spain.
E-mail add ess: [email p o ec ed] (J. Fe ei o).
1
We acknowledge he commen s o an edi o and an associa e edi o o he jou nal and h ee e iewe s. Thei sugges ions and ecommenda ions we e ex emely
help ul in imp o ing he a icle. The usual disclaime applies.
Con en s lis s a ailable a ScienceDi ec
Economic Modelling
jou nal homepage: www.jou nals.else ie .com/economic-modelling
h ps://doi.o g/10.1016/j.econmod.2023.106437
Recei ed 19 Sep embe 2022; Recei ed in e ised o m 5 July 2023; Accep ed 5 July 2023
Economic Modelling 126 (2023) 106437
2
Fu he mo e, many s udies a gue ha labou ma ke ins i u ions
ha e posi i e e ec s on he labou ma ke and economic ac i i y, such as
lowe unemploymen , highe employmen , smoo he luc ua ions o
economic ac i i y, mo e egali a ian dis ibu ion o income, highe
accumula ion o human and physical capi al, as well as mo e inno a ion
(B ancaccio e al., 2018; Ciminelli e al., 2018; Dosi e al., 2017, 2018;
Eu opean Commission Di ec o a e-Gene al o Employmen , Social A -
ai s and Inclusion, 2015; Flaschel e al., 2012; Kugle , 2019; La oie,
2017).
Mains eam s udies ha e paid pa icula a en ion o he impac o
employmen p o ec ion legisla ion (EPL) on employmen and unem-
ploymen . Based on he a gumen ha high employmen p o ec ion has
nega i e mic o and mac oeconomic e ec s, many coun ies ha e passed
e o ms o educe such p o ec ion, making i easie and cheape o
dismiss pe manen wo ke s and acili a ing he use o ixed- e m con-
ac s and agency wo ke s (Piasna and Myan , 2017).
The aim o his pape is o es he New Keynesian hypo hesis o he
exis ence o a nega i e e ec o EPL on he e olu ion o employmen and
unemploymen , in es iga ing whe he EPL is a signi ican de e minan
o he dynamics o employmen , employees and unemploymen a es in
Eu ope o e he pe iod 1985–2019. The esul s o his analysis a e
impo an om a policy poin o iew. As no ed abo e, ecen publica-
ions ha e highligh ed he nega i e consequences o excessi e labou
lexibili y. I could be a gued ha hese nega i e e ec s could be ou -
weighed by he bene i s o highe employmen and lowe unemploy-
men a es, bu i hese posi i e e ec s a e no ound he labou e o ms
ha ha e educed employmen p o ec ion o wo ke s could be quali ied
as nega i e.
The pape is s uc u ed as ollows. In sec ion 2, we p o ide a b ie
li e a u e e iew on he impac o employmen p o ec ion legisla ion on
employmen and unemploymen a es. Sec ion 3 p esen s he me hod-
ology o ou empi ical esea ch. Sec ion 4 p esen s he da a o he a -
iables used in ou es ima ions. Sec ion 5 p esen s he esul s o he
es ima ions o he impac o EPL on unemploymen a es g ow h. Sec ion
6 p esen s he esul s o he es ima ions o he impac o EPL on
employmen and employee g ow h. The inal sec ion summa izes and
concludes.
2. Li e a u e e iew
Acco ding o New Keynesian economics, which is la gely based on
he mone a is app oach o he exis ence o a na u al a e o unem-
ploymen (F iedman, 1968), he e is a long- un equilib ium a e o un-
employmen , he non-accele a ing in la ion a e o unemploymen
(NAIRU), which is de e mined by s uc u al-ins i u ional elemen s ha
p e en wages om adjus ing quickly in esponse o demand and supply
shocks. In he absence o changes in hese s uc u al elemen s, he
NAIRU emains s able in he long un and he cu en unemploymen
a e empo a ily de ia es om he equilib ium a e as a esul o demand
shocks. Bo h mone a is and New Keynesian economis s a gue ha he
impe ec ions in he labou ma ke ha lead o igidi ies in nominal and
eal wages and sluggish adjus men s o economic shocks de e mine he
unemploymen a e in he long un, and he e o e he highe he igid-
i ies, he highe he NAIRU (Fe ei o and Gomez, 2020). This easoning
implies ha labou ins i u ions ha inc ease he lexibili y o he labou
ma ke , allowing o a quick adjus men o wages in he p esence o an
economic shock, lead o highe employmen le els and lowe unem-
ploymen a es.
While mone a is s a gue ha empo a y changes in he cu en un-
employmen a e ela i e o he na u al a e o unemploymen do no
a ec he na u al a e o unemploymen , New Keynesian au ho s a gue
ha changes in economic ac i i y caused by demand shocks, especially i
hey a e long-las ing and in ense, can a ec he NAIRU. In his app oach,
he igidi ies c ea ed by labou ma ke ins i u ions, bo h in e ms o he
beha iou o nominal and eal wages and in e ms o hi ing and i ing,
a e a key de e minan o high and pe sis en unemploymen . Thus, i is
a gued ha he dynamics o unemploymen a e explained by he
in e ac ion o ad e se shocks wi h ad e se labou ma ke ins i u ions.
These ine icien ins i u ions ampli y he du a ion o he e ec s o shocks
on cu en unemploymen and hus, h ough hys e esis e ec s, inc ease
he NAIRU. This implies ha he g ea e he igidi ies c ea ed by labou
ins i u ions, he g ea e he nega i e impac o demand shocks on un-
employmen and employmen in bo h he sho un and long un.
Fo New Keynesian economis s, he e o e, he le el (and changes) o
employmen and he unemploymen a e, in bo h he sho and long un,
depend on he igidi ies c ea ed by labou ma ke ins i u ions. One o
hese ins i u ions is employmen p o ec ion legisla ion (EPL), which is
he se o ules ha go e n he hi ing and i ing o wo ke s in each
coun y. The hi ing ules a e he condi ions o he use o s anda d ( ull-
ime pe manen con ac s) and non-s anda d (pa - ime, ixed- e m and
empo a y agency wo ke s, e c.) employmen con ac s. The dismissal
ules go e n he indi idual and collec i e dismissal o wo ke s on pe -
manen con ac s. This legisla ion aims o p o ide wo ke s wi h a
ce ain le el o p o ec ion and secu i y in hei jobs by se ing ou he
equi emen s ha employe s mus obse e and espec when hi ing and
dismissing wo ke s.
Blancha d and Wol e s (2000) a gued in hei seminal a icle ha he
inc ease in s uc u al unemploymen expe ienced since he oil c ises by
Eu opean economies was he esul o he implemen a ion o employ-
men p o ec ion measu es adop ed o mi iga e he ad e se e ec s o
hese c ises on unemploymen . They a gued ha al hough his highe
p o ec ion could ha e educed he nega i e impac o down u ns on
unemploymen in he sho e m, i had a nega i e impac on hi ing in
he long e m (and on capi al accumula ion and p oduc i i y g ow h),
leading o highe unemploymen . O e all, he magni ude o hys e esis
e ec s, and hence equilib ium o s uc u al unemploymen , would be
di ec ly ela ed o s ong employmen p o ec ion (Ande on e al.,
2012). These a gumen s we e accep ed by in e na ional o ganisa ions,
which ecommended educing employmen p o ec ion, mainly o pe -
manen wo ke s, o ensu e lowe and mo e s able unemploymen a es
(Eu opean Commission, 2012; OECD, 2006, 2012, 2017, 2018).
2
Despi e he gene alisa ion o hese e o ms, he e is no clea empi -
ical e idence on he impac o hese measu es on employmen and un-
employmen (Be ola, 2017; Boe i e al., 2015; Heimbe ge , 2017; Heyes
and Lewis, 2015; OECD, 2018; Pa e nesi Meloni e al., 2022), and ecen
s udies conclude ha high employmen p o ec ion has no nega i e
impac on employmen and unemploymen (Adams e al., 2019; A da-
gic, 2015; A dagic and Sala di, 2013; Be ola, 2017; Boe i e al., 2015;
Fe ei o and Gomez, 2020, 2022; Flaschel e al., 2012; Heimbe ge ,
2017; Heyes and Lewis, 2015; Myan and B andhube , 2016; Piasna and
Myan , 2017). This would mean ha labou ma ke e o ms imple-
men ed since he 1980s would ha e no con ibu ed o educing high
unemploymen a es. Indeed, many con ibu ions ocus on he nega i e
economic consequences o hese e o ms, highligh ing he impac on
labou segmen a ion,
3
unemploymen sca s, income dis ibu ion, job
quali y, household consump ion and bo owing, in e na ional ade and
o eign di ec in es men lows, inno a ion, compe i i eness, p oduc-
i i y g ow h and po e y (A es is e al., 2020a; B ancaccio e al., 2018;
Damiani e al., 2016; Gonalons-Pons and Gangl, 2022; Gu ie ez-Ba -
ba usa, 2016; Heyes and Lewis, 2015; Kleinknech , 2020; OECD, 2018;
Oli ei a and Fo e, 2021; Roy, 2021; T idico, 2017).
I should be no ed ha mains eam economis s do no ca ego ically
claim ha employmen p o ec ion has a nega i e impac on he labou
ma ke . Blancha d and Wol e s (2000) and Blancha d (2018) a gue ha
i is likely ha he highe employmen p o ec ion egis e ed in he
2
Recommenda ions o educe employmen p o ec ion ha e also been
di ec ed o eme ging and de eloping economies (Du al and Loungani, 2021).
3
In cases whe e e o ms ha e encou aged he use o empo a y con ac s, and
whe e se e ance paymen s o e mina ing empo a y con ac s a e lowe han
dismissal cos s o open-ended con ac s.
P. A es is e al.
Economic Modelling 126 (2023) 106437
3
se en ies inc eased he na u al a e o unemploymen . Fo he OECD
(2018), employmen p o ec ion o pe manen wo ke s “ ends o ha e
ei he no o a small nega i e e ec on employmen ” (p. 124); and
excessi e employmen p o ec ion o hese wo ke s can ha e nega i e
consequences o job quali y, inclusi eness and p oduc i i y i i is
accompanied by lowe p o ec ion o empo a y wo ke s. O he s udies
a gue ha he e ec s o he EPL di e be ween g oups o wo ke s,
depending on gende , age, skills o ype o employmen con ac , wi h
unce ain e ec s on agg ega e employmen o unemploymen (A es is
e al., 2020b; Boe i e al., 2015; d’Agos ino e al., 2018; Gal and The-
ising, 2015).
Recen s udies ocusing on he pe iod a e he onse o he Global
Financial C isis ha e ein o ced he doub s abou he e ec s o EPL.
Ande on e al. (2012), Boe i and Jimeno (2016), and Sha ma and
Winkle (2018) a gue ha high employmen p o ec ion o pe manen
wo ke s is associa ed wi h a highe inc ease in unemploymen in Eu ope
du ing ha pe iod. In con as , S ockhamme e al. (2014) ind no sig-
ni ican e ec o EPL on unemploymen a es in OECD coun ies. Fo
Blancha d (2018), eplica ing he wo k o Blancha d and Wol e s
(2000), EPL is no a signi ican de e minan o unemploymen a es
when he pe iod analysed is ex ended o 2015. Fe ei o and Gomez
(2022) show ha , du ing he G ea Recession, employmen p o ec ion
did no ha e a signi ican impac on employmen g ow h and ha , in
e ms o unemploymen , only employmen p o ec ion o pe manen
wo ke s agains indi idual dismissals had a signi ican impac on un-
employmen , wi h highe employmen p o ec ion leading o lowe un-
employmen a es.
Finally, some ecen pape s (Boe i and Jimeno, 2016; De Almeida
and Balasundha am, 2018; Du al and Fu ce i, 2018; Du al e al., 2020;
OECD, 2012, 2017) a gue ha he impac o employmen p o ec ion
depends on he phase o he business cycle; hence, i does no a ec
employmen and unemploymen in he long un.
Ou pape a emp s o ad ance he s udy o he e ec s o employ-
men p o ec ion on labou ma ke ou comes by ca ying ou an empi -
ical analysis ha can be conside ed no el. Unlike mos exis ing
empi ical s udies, ou pape analyses no only he impac o EPL on he
a ia ion o unemploymen a es, bu also on employmen g ow h.
Mo eo e , ou pape analyses he impac o EPL on bo h o al
employmen and employees. Mos exis ing pape s analyse he impac o
EPL on o al employmen ( he sum o employees and sel -employmen ).
Howe e , he EPL egula es he hi ing and dismissal condi ions o em-
ployees, so he di ec impac should be on employees. Only i we assume
ha sel -employmen is no a ec ed by he EPL, he change in employees
mus lead o a simila change in o al employmen . In o he wo ds, i he
coe icien o EPL is signi ican in es ima ing he de e minan s o em-
ployees, he coe icien o EPL should be equally signi ican and wi h he
same sign in es ima ing o al employmen .
P oblems would a ise i he signi icance le el o he coe icien s is no
simila and/o he sign o he coe icien is opposi e. This esul implies
ha he e ec s o he EPL on employmen and employees a e opposi e,
wi h changes in sel -employmen mo e han o se ing changes in em-
ployees. To he bes o ou knowledge, he e is no s udy ha igo ously
analyses he mechanisms h ough which employmen p o ec ion a ec s
he g ow h o sel -employmen . This means ha we lack he analy ical
ools necessa y o in e p e such a esul , beyond concluding ha he
esul s a e no obus and ha no ca ego ical conclusion can be d awn
abou he e ec s o EPL on job c ea ion.
Ano he no el y o ou pape is he use o wo di e en sou ces o
da a on employmen and employees. All exis ing s udies use a single
da a sou ce, ei he Labou Fo ce Su eys (LFS) o Na ional Accoun s
(NA) da a. This implies he assump ion ha he choice o he da a sou ce
does no a ec he alidi y o he esul s ob ained. Howe e , ou pape
analyses da a om bo h sou ces. This app oach allows us o adequa ely
es he obus ness o he esul s ob ained. I he sign o he deg ee o
signi icance o he di e en explana o y a iables we e di e en
depending on he da a sou ce, his would aise se ious doub s abou he
ue e ec o employmen p o ec ion legisla ion.
Ano he con ibu ion o ou pape is he use o G oss Domes ic
P oduc (GDP) and G oss Fixed Capi al Fo ma ion (GFCF) g ow h a es
as a iables ela ed o economic g ow h. Fu he mo e, we do no only
in es iga e a linea ela ionship be ween he dependen a iables and
employmen p o ec ion, bu also he exis ence o a non-linea ela ion-
ship, as well as he exis ence o in e ac ion e ec s be ween employmen
p o ec ion legisla ion and economic g ow h. In his way, we es he
alidi y o he hypo heses ha employmen p o ec ion has nega i e e -
ec s when i s le el is excessi e and ha he e ec s o employmen
p o ec ion depend on he phase o he cycle in which he economy inds
i sel .
We would also like o poin ou ha almos all exis ing pape s
analyse sho ime pe iods, using me hods such as Gene alized Me hod
o Momen s (GMM) models. Howe e , ou pape ocus on a e y long
pe iod (35 yea s). The exis ence o a long panel a oids he p oblem aced
by mos s udies which, by ocusing on a sho e pe iod, a e aced wi h
he ques ion o whe he he esul s ob ained a e condi ioned by he
speci ic choice o he da es analysed. On he o he hand, he a ailabili y
o long- ime se ies allows us o use me hods o he han he usual GMM
models o analyse dynamic models, as in ou case.
3. Empi ical me hodology
3.1. Baseline speci ica ion
The aim o his pape is o analyse whe he employmen p o ec ion
legisla ion is a signi ican de e minan o he dynamics o employmen ,
employees and unemploymen a es in Eu opean coun ies. The e o e,
he g ow h a es o employmen and employees and he g ow h in pe -
cen age poin s o unemploymen a es a e he explained a iables o he
empi ical analyses ha we will ca y ou in he pape . The choice o he
g ow h a e o he dependen a iables, a he han hei le el, is based
on he ac ha New Keynesian models sugges ha labou ma ke
ou comes a e explained by he in e ac ion be ween economic g ow h
and labou ins i u ions, in his case EPL. Tha is, o a gi en a e o
economic g ow h, he a ia ion in employmen and unemploymen a es
would depend on he le el o employmen p o ec ion. Thus, o a gi en
(posi i e) g ow h a e o economic ac i i y, he lowe he EPL indices,
he highe he g ow h in employmen and employees and he highe he
all in he unemploymen a e, and ice e sa.
Al hough he New Keynesian app oach a gues ha he impac o EPL
on employmen and employees’ g ow h is nega i e, wi h EPL educing
employmen and employee g ow h a es, exis ing pape s only es he
impac o EPL on one o he wo a iables. This implici ly assumes ha ,
gi en he high co ela ion be ween he wo a iables,
4
i EPL has a
signi ican impac on one a iable (e.g. employmen ), he impac on he
o he a iable (employees) should also be signi ican and i s coe icien
should ha e he same sign. Howe e , i he sign, magni ude and sig-
ni icance o he coe icien s o he EPL indices we e signi ican ly
di e en in he case o employmen and employee g ow h, his
disc epancy would aise se ious doub s abou he obus ness o he e-
sul s, calling in o ques ion any conclusions abou he ue e ec o EPL.
In ou analysis we use wo sou ces o da a on employmen and em-
ployees: Labou Fo ce Su eys (LFS) and Na ional Accoun s (NA) s a-
is ics. Al hough he da a a e e y simila and highly co ela ed,
5
he
esul s o he es ima ions may di e depending on he sou ce o he da a.
The e o e, we es sepa a ely he de e minan s o he employmen and
employee g ow h a es measu ed by he LFS and he NA, which will
4
In ou sample, he co ela ion be ween he employmen and employee
g ow h a es is 0.829 (da a om LFS) and 0.813 (da a om Na ional Accoun s).
5
In ou sample, he co ela ion be ween he employmen g ow h a es
calcula ed using LFS and NA is 0.857, and he co ela ion be ween he
employee g ow h a es calcula ed using LFS and NA is 0.834.
P. A es is e al.
Economic Modelling 126 (2023) 106437
4
allow us o assess he obus ness o he esul s o he impac o he
explana o y a iables. The employmen and employee g ow h a es
based on he LFS a e aken om he OECD da a on o al employmen and
employees. The excep ion is Swi ze land, whose da a on employees a e
aken om he ICTWSS da abase (Visse , 2019). Na ional Accoun s da a
on employmen and employees a e aken om he AMECO da abase.
Unemploymen a es a e aken om he AMECO da abase, excep o
Ge many, o which da a a e aken om he OECD.
The mains eam pos ula es ha , o a gi en a e o economic g ow h,
coun ies wi h mo e lexible labou ma ke s, i.e. wi h lowe employmen
p o ec ion, will expe ience highe employmen and employee g ow h
and a la ge decline in unemploymen . Howe e , he li e a u e sugges s
ha labou ma ke ou comes can be in luenced by o he ac o s o a
demog aphic, economic and ins i u ional na u e. Fo his eason, ou
model will include con ol a iables ela ed o he g ow h o he
wo king-age popula ion (Pop), ade openness (T ade), a se o a iables
ela ed o labou ins i u ions (LabIns ) and, inally, a iables ela ed o
he s uc u e and coo dina ion o collec i e ba gaining (CollBa g).
Yi, =β0+β1Economicg ow hi, +β2EPLi, +β3Popi, +β4T adei,
+β5LabIns i, +β6CollBa gi, +
ε
i,
In ou s udy we use G oss Domes ic P oduc (GDP) and G oss Fixed
Capi al Fo ma ion (GFCF) g ow h a es as explana o y a iables ela ed
o economic ac i i y. Al hough mos pape s use GDP g ow h, pos -
Keynesian s udies emphasise he ole o capi al accumula ion as he
main d i e o employmen and unemploymen . In any case, we expec
ha bo h GDP and GFCF g ow h a es ha e a signi ican posi i e impac
on employmen , employees, and unemploymen a es. Gi en he high
co ela ion be ween GDP and GFCF g ow h a es,
6
hese wo a iables
canno be included in he same equa ion, and, consequen ly, o each
dependen a iable we es wo equa ions ha di e in he a iable used
o measu e he e ec o economic g ow h: GDP o GFCF g ow h a e.
This aises he possibili y ha he esul s on he impac o EPL may be
in luenced by he chosen a iable ela ed o economic g ow h. In his
way, we es no only he alidi y o he pos -Keynesian s udies, bu also
he obus ness o he conclusions on he impac o EPL on employmen ,
wo ke s and unemploymen . Thus, a obus e ec o employmen p o-
ec ion on he labou ma ke exis s i he sign and signi icance o he
coe icien s on he EPL indices a e he same ega dless o whe he we use
GDP o GFCF g ow h as he explana o y a iable. Real GDP and GFCF
da a a e aken om he AMECO da abase, excep o Ge many whe e
da a a e aken om he OECD.
In o de o analyse he e ec s o employmen p o ec ion on labou
ma ke ou comes we use he Employmen P o ec ion Legisla ion (EPL)
s ic ness indica o s de eloped by he OECD. The OECD EPL indica o s
measu e he s ic ness o employmen p o ec ion o egula (pe ma-
nen ) and empo a y con ac s by cons uc ing syn he ic indica o s
based on he alues assigned o di e en i ems. Each indica o is
measu ed on a scale om 0 o 6, wi h highe alues ep esen ing s ic e
egula ion and hence a mo e igid labou ma ke . The sco e o each
index is calcula ed on he basis o he legisla ion in o ce on 1 Janua y o
each yea .
Gi en he common me hodology used o cons uc he indexes, hey
make i possible o compa e employmen p o ec ion legisla ion be ween
coun ies and o ack he e olu ion o na ional EPL indexes, which a e
linked o legal e o ms a ec ing he condi ions o dismissing o hi ing a
wo ke using one o he a ailable employmen con ac s. Al hough hese
indexes ha e p oblems in measu ing he ue lexibili y- igidi y o la-
bou ma ke s, such as he inabili y o measu e employmen p o ec ion
based on no ms o he han legal ones, and he ailu e o ake in o ac-
coun p ocedu al equi emen s in assessing he di icul ies and cos s o
ca ying ou indi idual and collec i e dismissals (Ha cou e al., 2021;
Myan and B andhube , 2016), hei use in empi ical analyses is wide-
sp ead and hus allows o he compa ison o he esul s o di e en
s udies.
The OECD calcula es se e al indices: he EPRC index, which mea-
su es he p o ec ion o egula -pe manen employees agains indi idual
and collec i e dismissal, and he EPT index, which measu es he egu-
la ion o empo a y o ms o employmen , mainly ixed- e m and em-
po a y agency wo ke s. In addi ion, he EPRC index is spli in o wo
indexes: he EPR index, which ela es o he p o ec ion o pe manen
employees agains indi idual dismissal; and he EPC index, which e-
la es o he speci ic addi ional equi emen s o collec i e dismissals o
pe manen employees. In ou case, in o de o analyse a long pe iod o
ime, we will use e sions 1 o he EPRC and EPT indices, which co e
he pe iod 1985–2019.
An analysis o labou ma ke pe o mance based only on changes in
economic ac i i y and labou ma ke lexibili y assumes ha labou
supply is cons an . Howe e , changes in labou supply due o de-
mog aphic changes in he popula ion, mig a ion lows, ageing, e c., can
a ec labou ma ke s. In o de o accoun o hese p oblems, we include
in ou es ima ions he g ow h o he wo king age popula ion as an
explana o y a iable. The a iable Pop
i,
is he g ow h a e o he pop-
ula ion aged 15–64 in coun y i in yea (da a om he AMECO
da abase).
In o de o con ol o he possible in luence o o he a iables on he
e olu ion o employmen and unemploymen a es, we ha e included
se e al a iables ha ha e been highligh ed in he li e a u e as possible
de e minan s o labou ma ke ou comes. The i s a iable is ade
openness (T ade
i,
), measu ed as he pe cen age o GDP o he sum o
expo s and impo s o goods and se ices. This pe cen age was calcu-
la ed using da a on expo s and impo s om he AMECO da abase ( om
he OECD o Ge many).
In he case o popula ion g ow h, he sign o he coe icien is ex-
pec ed o be posi i e, so ha g ow h in labou supply is e lec ed in
highe employmen and a highe unemploymen a e. Rega ding he
sign o he coe icien on ade openness, se e al s udies poin o he
nega i e e ec s o he globalisa ion p ocess on he labou ma ke , so ha
he expec ed sign would be nega i e o employmen and employee
g ow h and posi i e o he unemploymen a e, slowing down job
c ea ion and inc easing he unemploymen a e.
The emaining con ol a iables co espond o a iables ela ed o
labou ma ke ins i u ions. One such a iable is union densi y (Union
i,
),
which measu es he pe cen age o employees who a e membe s o a
ade union. The da a a e aken om he OECD and AIAS, Ins i u ional
Cha ac e is ics o T ade Unions, Wage Se ing, S a e In e en ion and
Social Pac s (ICTWSS) da abase (OECD and AIAS, 2021). Ano he con-
ol a iable is public expendi u e on ac i e labou ma ke policies
(ALMP
i,
), measu ed as a pe cen age o GDP. Da a o his a iable come
om he OECD.
The las a iables included in he models ela e o collec i e ba -
gaining, bo h in e ms o he s uc u e and cen alisa ion o wage ba -
gaining and he coo dina ion o he wage-se ing p ocess. We include he
a iable Coo d, which measu es he deg ee o coo dina ion in he wage-
se ing p ocess. This index, a ailable in he OECD/AIAS ICTWSS da a-
base, anges om 1 (co esponding o agmen ed wage ba gaining a
he en e p ise o plan le el wi h no coo dina ion) o 5 (co esponding o
he exis ence o binding no ms esul ing om cen alised ba gaining
be ween unions and employe s’ associa ions o go e nmen -imposed
wage g ow h guidelines).
Rega ding he cen alisa ion o collec i e ba gaining, he OECD/
AIAS ICTWSS da abase p o ides se e al indices. Cen al is a summa y
index ha akes in o accoun he incidence and con ol o addi ional
ba gaining a he en e p ise le el, he ‘space’ ha cen al o sec o al
ag eemen s alloca e, delega e o allow o such addi ional ba gaining,
and he ex en o which ag eemen s can be pe o a ed h ough he use o
‘opening clauses’. Cen al anges om 0 o 5, wi h highe alues
6
In ou sample, he co ela ion be ween he GDP and GFCF g ow h a es is
0.708.
P. A es is e al.
Economic Modelling 126 (2023) 106437
5
co esponding o a ba gaining s uc u e whe e cen alised ba gaining o
na ional sec o al ag eemen s p edomina e. Le el is an index ha e lec s
he p edominan le el a which ba gaining akes place (in e ms o
co e age o employees), wi h alues anging om 1 (company o en-
e p ise le el) o 5 (cen alised o c oss-indus y ba gaining). Ano he
index is Mul ile el, which e lec s he combina ion o le els a which
collec i e ba gaining on pay akes place, wi h alues anging om 1
(company le el) o 7 (c oss-sec o al, wi h cen ally de e mined binding
no ms o ceilings ha all o he ag eemen s mus espec ). The high
co ela ion be ween he h ee a iables means ha hey canno all be
included in he same equa ion. On he o he hand, gi en ha he e may
be a ela ionship be ween he deg ee o cen alisa ion o collec i e
ba gaining and he coo dina ion o wage ba gaining, we ha e chosen o
include he Mul ile el a iable in ou ini ial model, as i is he a iable
wi h he lowes co ela ion wi h Coo d (see Table 1).
In he case o a iables ela ed o labou ma ke ins i u ions,
ollowing an o hodox poin o iew, we would expec he sign o be
nega i e o employmen and employee g ow h and posi i e o unem-
ploymen g ow h, i.e., hey would slow down job c ea ion and inc ease
he unemploymen a e.
The dynamics o labou ma ke ou comes a e cha ac e ised by ine ia
and high pe sis ence o labou ma ke ou comes. This leads o a po en-
ial p oblem o se ial co ela ion, which a ec s he consis ency o he
esul s. Indeed, he exis ence o se ial co ela ion was ound in all he
models examined.
7
In o de o co ec his p oblem, he one-pe iod lag
o he explained a iable was included as an explana o y a iable in all
he equa ions. In his way, we ans o m he es ed models in o dynamic
models:
Yi, =β0+β1Yi, −1+β2Economicg ow hi, +β3EPRCi, +β4EPTi, +β5Popi,
+β6T adei, +β7Unioni, +β8ALMPi, +β9Coo di, +β10Mul ile eli, +
ε
i,
Mos empi ical s udies o he impac o EPL on he labou ma ke use
GMM models o sol e he p oblem o in oducing he lagged dependen
a iable as an explana o y a iable. GMM models a e sui able o sho
panels wi h a la ge numbe o coun ies and a small numbe o ime
pe iods. Howe e , o long panels whe e he numbe o yea s (T) is la ge,
abo e 30, and he numbe o indi iduals (N) is signi ican ly smalle han
he numbe o pe iods, being he T/N a io abo e 2, he es ima ion o
dynamic models using ixed e ec s p o ides much mo e consis en e-
sul s han al e na i e p ocedu es, such as ins umen al a iables (IV) o
GMM es ima o s, because he bias o he GMM es ima o s inc eases wi h
he numbe o pe iods (Bal agi, 2005; Hsiao, 2014; Kennedy, 2008;
Pesa an, 2015; Woold idge, 2010). Gi en ha we ha e a panel wi h 35
yea s and 16 coun ies, ou models a e es ima ed using ixed e ec s.
On he o he hand, Eu opean economies a e highly in e linked and
can be a ec ed by common shocks. The e o e, in cases whe e he panel
es s con i med he exis ence o c oss-sec ional dependence, we ha e
used SUR es ima o s o co ec o con empo aneous co ela ion be-
ween c oss-sec ions
8
(Kennedy, 2008).
One issue ha may a ec he alidi y o he esul s is he possible
endogenei y o EPL (Roy, 2021). EPL indices a e no immune o he
c i icism ha hey may be subjec ed o measu emen e o and hus may
no co ec ly e lec he ue deg ee o lexibili y in hi ing and i ing.
Mo eo e , i is plausible ha he e is an in e se causal ela ionship
be ween labou ma ke ou comes and EPL indices, as policymake s may
e o m hi ing and i ing condi ions based on labou ma ke pe o -
mance. The e o e, ollowing Woold idge (2010, 2013), we es o
possible endogenei y o he EPRC and EPT indices using wo ins u-
men al a iables. The i s is Compensa ions o Employees as a pe -
cen age o GDP (da a om he OECD), and he second is he Type index
om he OECD/AIAS ICTWSS da abase, which measu es he ype o
wage-se ing coo dina ion. The es s ca ied ou show ha hese in-
s umen s a e ela ed o he EPL indica o s and ha hey a e exogenous,
since hey do no a ec he g ow h o employmen and unemploymen .
9
The es s ca ied ou allow us o conclude ha he EPRC index is exog-
enous in all he es ima ions. In he case o he EPT index, we only ind an
endogenei y p oblem in he es ima ions o unemploymen g ow h. This
esul would imply ha Eu opean coun ies ha e made he labou
ma ke mo e lexible in o de o educe he unemploymen a es, and o
do so hey ha e acili a ed he use o empo a y con ac s. This hy-
po hesis is consis en and compa ible wi h he inc easing segmen a ion
and dualisa ion o a la ge pa o Eu opean labou ma ke s (Eichho s
and Ma x, 2021). Indeed, in ou sample o coun ies and yea s, he
a e age alue o he EPRC index ell om 2.56 o 2.23 be ween 1985
and 2019, while he a e age alue o he EPT index ell om 2.825 o
1.724, indica ing he in ensi y o e o ms ha ha e acili a ed empo a y
hi ing.
The e o e, in he case o unemploymen g ow h es ima es, in addi-
ion o OLS models, we es ima e he equa ions using an ins umen al
a iable (IV) app oach based on a Two-S age Leas Squa es (TSLS)
speci ica ion wi h ixed e ec s. In his case, we use he Compensa ion o
Employees and Type a iables as ins umen s o he EPT index. The
choice o his p ocedu e, as explained abo e, is based on he exis ence o
a la ge panel wi h long ime se ies
10
.
3.2. Robus ness checks
Ou baseline models assume a linea ela ionship be ween EPL in-
dica o s and employmen and employee g ow h and he unemploymen
a e. Howe e , exis ing s udies sugges he exis ence o non-linea e -
ec s be ween EPL and employmen and unemploymen g ow h. By
including a quad a ic ela ionship be ween employmen p o ec ion and
labou ma ke ou comes, we es he hypo hesis whe he he e is an
inc easing o dec easing ma ginal ela ionship be ween hese a iables,
and whe he he e is a h eshold a which he e ec s o EPL on
employmen and unemploymen inc ease o dec ease. We he e o e es
he ollowing equa ion:
Yi, =β0+β1Yi, −1+β2Economicg ow hi, +β3EPRCi, +β4EPRC2
i,
+β5EPTi, +β6EPT2
i, +β7Popi, +β8T adei, +β9Unioni, +β10ALMPi,
+β11Coo di, +β12Mul ile eli, +
ε
i,
Mo eo e , gi en ha some empi ical s udies sugges ha he e ec s
o employmen p o ec ion depend on he economic con ex o he phase
o he cycle in which he economy inds i sel , we also es o he ex-
is ence o in e ac ion e ec s be ween employmen p o ec ion and eco-
nomic g ow h. The exis ence o hese e ec s would indica e ha he
impac o employmen p o ec ion on he labou ma ke di e s
depending on he a e o economic g ow h and, consequen ly, on he
phase o he cycle in which he economy inds i sel . The equa ion o be
es ima ed is he e o e:
Yi, =β0+β1Yi, −1+β2Economicg ow hi, +β3EPRCi,
+β4Economicg ow hi, *EPRCi, +β5EPTi, +β6Economicg ow hi, *EPTi,
+β7EPopi, +β8T adeomici, +β9Unioni, +β10ALMPi, +β11 Coo di,
+β12Mul ile eli, +
ε
i,
7
Rele an da a a e a ailable upon eques .
8
Rele an da a a e a ailable upon eques .
9
Da a a ailable upon eques .
10
The exis ence o long ime se ies aises he possibili y o non-s a iona i y o
he a iables, which would equi e he use o o he me hods (coin eg a ion
me hods and e o co ec ion models) o analyse he sho - and long- e m e-
la ionships be ween he a iables in he model. Howe e , he es s ca ied ou
show ha he se ies o g ow h a es o employmen and employees and he
g ow h o unemploymen a es a e s a iona y. Da a a ailable upon eques .
P. A es is e al.
Economic Modelling 126 (2023) 106437
6
4. Da a
To analyse he impac o employmen p o ec ion on he dynamics o
employmen , employees and unemploymen a es o e he longes
possible pe iod, we use e sions 1 o he EPRC and EPT indices. Gi en
he a ailabili y o da a, we analyse he de e minan s o labou ma ke
pe o mance in 16 Eu opean coun ies (Aus ia, Belgium, Denma k,
Finland, F ance, Ge many, G eece, I eland, I aly, he Ne he lands,
No way, Po ugal, Spain, Sweden, Swi ze land and he Uni ed Kingdom)
be ween 1985 and 2019 (35 yea s). As no all a iables a e a ailable in
all coun ies o he o al numbe o yea s analysed, we ha e an unbal-
anced panel.
Table 2 p esen s he main desc ip i e s a is ics o he a iables
included in ou analysis. To al employmen and employees g ew a an
a e age annual a e o jus unde 1%, while he unemploymen a e
emained i ually unchanged. Economic ac i i y g ew a an a e age
annual a e o 2.2% and p oduc i e in es men a a sligh ly highe a e
o 2.7%, al hough in his case wi h g ea e dispe sion.
Rega ding wage ba gaining, he coo dina ion o he wage-se ing
p ocess (Coo d) is cha ac e ised by he exis ence o non-binding
no ms and guidelines issued by he go e nmen and/o employe s’ as-
socia ions and ade unions. The deg ee o cen alisa ion o ba gaining
(Mul ile el) is cha ac e ised by an in e media e s uc u e in which
ba gaining by sec o o indus y p edomina es.
Table 2 does no p o ide any in o ma ion on he empo al dynamics
o hese a iables, in pa icula on he exis ence o ends o b eaks in
hei e olu ion ha could lead o signi ican di e ences a di e en
poin s du ing he pe iod analysed. Wi h his objec i e in mind, we ha e
plo ed in Fig. 1 he e olu ion o e ime o he a e age alue o he
a iables included in ou models o he 16 coun ies s udied.
The h ee dependen a iables luc ua e in a s able, cyclical manne
a ound he a e age a es o he pe iod. A simila esul can be obse ed
o he GDP and GFCF g ow h a es. Fo he emaining explana o y
a iables, howe e , he e a e clea ends. Fo example, he g ow h a e
o he wo king age popula ion ell sha ply om 2008 onwa ds. T ade
openness, on he o he hand, is cha ac e ised by a con inuously
inc easing pa h, e lec ing he accele a ion o he p ocess o economic
globalisa ion. Rega ding he a iables ela ed o labou ma ke in-
s i u ions, a clea downwa d end can be obse ed o all o hem. The
decline in he EPRC and EPT indices e lec s a gene al end o educe
employmen p o ec ion and he commi men o make he use o em-
po a y con ac s mo e lexible. The da a also show a decline in ade
union densi y. As ega ds expendi u e on ac i e labou ma ke policies,
public expendi u e on hese i ems has been declining since 1993, when i
peaked (1% o GDP), and s ood a 0.7% o GDP in 2018. Finally, wi h
ega d o collec i e ba gaining, he deg ee o coo dina ion o he wage-
se ing p ocess is g adually declining, along wi h a mo e owa ds g ea e
decen alisa ion o collec i e ba gaining.
F om a New Keynesian pe spec i e, he decline in employmen
p o ec ion should ha e ansla ed in o an accele a ion in job c ea ion
and a educ ion in unemploymen a es, a p ocess acili a ed by he
g ea e lexibili y in wage ba gaining p ocesses and he loss o wo ke s’
ba gaining powe associa ed wi h lowe union densi y. Howe e , as
Fig. 1 shows, he g ow h o employmen , employees and unemploymen
a es ha e emained ai ly s able o e he long e m.
5. Employmen p o ec ion legisla ion and g ow h o
unemploymen a e
Acco ding o he New Keynesian app oach, he coe icien s o he
EPRC and EPT indica o s should always be signi ican , wi h a nega i e
sign in he es ima ions o he de e minan s o employmen and employee
g ow h a es, and a posi i e sign in he case o he g ow h o he un-
employmen a e. This implies ha employmen p o ec ion has a
nega i e impac on employmen , employees and unemploymen a es,
and ha hose labou e o ms ha ha e educed employmen p o ec ion
would ha e con ibu ed o speeding up he p ocess o job c ea ion and
educing unemploymen a es.
Table 3 shows he esul s o he equa ions es ing he de e minan s o
he e olu ion o he unemploymen a e. The da a show he high ine ia
o he change in he unemploymen a e, gi en he posi i e alue o he
lagged change in he unemploymen a e. As expec ed, economic
g ow h, whe he measu ed by GDP o p oduc i e in es men g ow h
a es, has a signi ican in e se e ec on he e olu ion o he unem-
ploymen a e, con ibu ing o i s educ ion
11
.
Rega ding he impac o EPL, he esul s o ou analysis clea ly show
ha employmen p o ec ion o pe manen and empo a y wo ke s does
no ha e a signi ican impac on he g ow h o he unemploymen a e.
These esul s can be conside ed obus and conclusi e, as hey a e no
a ec ed by he use o GDP o in es men g ow h a es as explana o y
a iables, o by he ype o model (OLS wi h ixed e ec s o IV-TSLS).
11
Rega ding he con ol a iables, he e ec o he wo king-age pop-
ula ion on he dynamics o he unemploymen a e is signi ican and
di ec , so ha he g ow h o he wo king-age popula ion inc eases he
unemploymen a e. In con as , he emaining con ol a iables a e no
signi ican , so hey would no a ec changes in he unemploymen a e.
Only in Mul ile el, he a iable ela ed o he cen alisa ion o collec i e
Table 1
Co ela ion among explana o y a iables.
T ade ALMP Cen al Coo d EPRC EPT GDP GFCF Le el Mul ile el Pop Union
T ade 1.00 0.07 0.04 0.36 −0.34 −0.38 0.23 0.25 0.12 0.04 0.21 0.01
ALMP 0.07 1.00 0.06 0.29 −0.05 0.00 0.01 −0.02 0.17 −0.00 −0.09 0.49
Cen al 0.04 0.06 1.00 0.49 0.11 0.32 0.11 −0.01 0.94 0.75 0.29 0.18
Coo d 0.36 0.29 0.49 1.00 −0.14 −0.11 0.04 0.00 0.62 0.41 0.19 0.42
EPRC −0.34 −0.05 0.11 −0.14 1.00 0.350 −0.14 −0.11 0.07 0.08 −0.22 −0.24
EPT −0.38 0.00 0.32 −0.11 0.35 1.00 −0.11 −0.04 0.30 0.34 −0.09 −0.06
GDP 0.23 0.01 0.11 0.04 −0.14 −0.11 1.00 0.68 0.04 0.05 0.31 0.03
GFCF 0.25 −0.02 −0.01 0.04 −0.11 −0.04 0.68 1.00 −0.04 −0.02 0.19 0.01
Le el 0.12 0.17 0.94 0.62 0.07 0.30 0.04 −0.04 1.00 0.74 0.23 0.25
Mul ile el 0.04 −0.00 0.75 0.41 0.08 0.34 0.05 −0.02 0.74 1.00 0.13 0.29
Pop 0.21 −0.09 0.29 0.19 −0.22 −0.09 0.31 0.19 0.23 0.13 1.00 −0.05
Union 0.01 0.49 0.18 0.42 −0.24 −0.06 0.03 0.00 0.25 0.29 −0.05 1.00
Sou ce: Own calcula ions
11
I is impo an o no e ha he coe icien s o GFCF g ow h a e smalle han
hose o GDP g ow h. This esul is also ob ained in he es ima ions o he de-
e minan s o he g ow h o employmen and employees. One explana ion o
his esul could be ha in es men is basically gea ed owa ds p omo ing a less
labou -in ensi e p oduc ion model, a phenomenon uelled by economic glob-
alisa ion and he eloca ion o he mos labou -in ensi e s ages o p oduc ion o
eme ging economies. Howe e , i could be due o he g ea e dispe sion o he
da a o GFCF g ow h, as shown in Table 2, and he consequen exis ence o
ex eme da a o ou lie s in his a iable. In any case, despi e he ele ance o
his esul , an explana ion o he lowe alue o he coe icien s o GFCF g ow h
is beyond he scope o his pape .
P. A es is e al.
Economic Modelling 126 (2023) 106437
7
Table 2
Summa y s a is ics.
Mean Median Maximum Minimum S d. De . Obs.
ΔEmploymen LFS 0.91 0.93 13.51 −8.86 2.11 560
ΔEmploymen NA 0.87 0.99 8.15 −7.84 1.75 553
ΔEmployees LFS 1.14 1.14 29.57 −13.37 2.70 560
ΔEmployees NA 1.03 1.16 7.27 −8.01 1.91 546
ΔUnemploymen a e −0.03 −0.10 6.60 −3.30 1.14 560
ΔGDP 2.16 2.20 25.18 −10.15 2.56 560
ΔGFCF 2.68 2.80 74.87 −25.37 7.64 560
EPRC 2.39 2.41 5.00 1.10 0.82 560
EPT 2.08 1.63 5.25 0.25 1.25 560
ΔPop15-64 0.40 0.36 3.30 −1.33 0.57 560
T ade openness 80.23 69.77 252.34 34.33 35.97 560
Union densi y 38.97 33.80 86.60 8.50 21.46 521
ALMP 0.81 0.74 2.70 0.06 0.45 533
Coo d 3.29 4.00 5.00 1.00 1.06 560
Mul ile el 3.00 2.00 6.00 1.00 1.42 560
Sou ce: Own calcula ions
Fig. 1. E olu ion o he a e age.
Sou ce: Own calcula ions.
P. A es is e al.
Economic Modelling 126 (2023) 106437
8
ba gaining, is ound o be a signi ican e ec . Howe e , his e ec only
appea s in he equa ion es ima ed by OLS ha includes GDP g ow h as
an explana o y a iable and, he e o e, canno be conside ed obus .
Gi en ha he impac o employmen p o ec ion may a y depending
on he le el o employmen p o ec ion o i s in e ac ion wi h economic
g ow h, we ha e es ima ed he impac o EPL on he g ow h o he un-
employmen a e, es ing o he exis ence o non-linea e ec s o EPL
and in e ac ion e ec s be ween employmen p o ec ion and economic
g ow h (Table 4).
Rega ding non-linea e ec s (equa ions 1 and 2), he esul s ule ou
he exis ence o a non-linea ela ionship be ween he EPRC index and
he g ow h o unemploymen a es. In he case o p o ec ion o em-
po a y wo ke s, we ha e ound he exis ence o a dec easing ma ginal
e ec , such ha employmen p o ec ion o empo a y wo ke s con-
ibu es o educing he unemploymen a e when he EPT index is abo e
3.15. Howe e , his is no a obus and conclusi e esul , as no such
e ec is ound when GDP g ow h is used as explana o y a iable.
Rega ding he in e ac ion e ec s (equa ions 3 and 4), he esul s
change depending on whe he we use GDP o in es men g ow h. I we
use GDP g ow h, bo h economic g ow h and he g ow h o he wo king
age popula ion a ec employmen g ow h as expec ed. Employmen
p o ec ion o pe manen wo ke s has no signi ican e ec on he un-
employmen a e, and p o ec ion o empo a y wo ke s alone has no
e ec , al hough we do ind an in e ac ion e ec , meaning ha he highe
he le el o employmen p o ec ion o empo a y wo ke s he g ea e
he educ ion in he unemploymen a e as a esul o economic g ow h.
All o he con ol a iables ha e no e ec , excep o mul ile el, which
implies ha g ea e cen alisa ion o collec i e ba gaining con ibu es o
highe unemploymen .
Howe e , he esul s change i we use he g ow h o p oduc i e in-
es men . Wi h GFCF g ow h, he g ow h o wo king-age popula ion and
he o he con ol a iables a e no longe signi ican . The es ima es show
ha EPL indices and he in e ac ion be ween EPL indices and GFCF
g ow h a e signi ican . Employmen p o ec ion would inc ease unem-
ploymen a es, al hough his e ec would be smalle he highe he
in es men g ow h, especially in he case o he p o ec ion o empo a y
wo ke s. Howe e , he esul s a e a om conclusi e o se e al ea-
sons. Fi s , he g ow h o in es men alone has no e ec on he g ow h o
he unemploymen a e; so, hese esul s should be ea ed wi h cau ion.
Mo eo e , he coe icien s co esponding o he EPRC index and i s
in e ac ion wi h GFCF g ow h a e no signi ican a he usual 5%
p obabili y le el. As a as he EPT index is conce ned, al hough p o-
ec ion o empo a y wo ke s inc eases he unemploymen a e, his
e ec is smalle he highe he g ow h o p oduc i e in es men , so ha
a in es men g ow h a es o 8% o mo e, p o ec ion o empo a y
wo ke s educes he unemploymen a e. Thus, he e ec o EPT on he
unemploymen a e depends on he g ow h o in es men .
In summa y, hese esul s allow us o conclude, i s , ha employ-
men p o ec ion does no inc ease unemploymen a es, and, second,
ha labou ma ke e o ms aimed a making he labou ma ke mo e
lexible by educing employmen p o ec ion o pe manen and empo-
a y wo ke s ha e no con ibu ed o educing unemploymen a es in
Eu ope. The e o e, he only e ec i e and iable s a egy o educe un-
employmen a es in Eu ope is o implemen economic policies ha
a ou economic g ow h.
Table 3
EPL and g ow h o unemploymen a e.
OLS IV TSLS
(1) (2) (3) (4)
C −0.003
(0.175)
−0.252
(0.182)
0.377
(0.643)
−0.083
(0.370)
Unemploymen (-1) 0.462***
(0.063)
0.469***
(0.063)
0.460***
(0.063)
0.472***
(0.064)
GDP g ow h −0.170***
(0.024)
−0.181***
(0.034)
GFCF g ow h −0.047***
(0.008)
−0.047***
(0.008)
EPRC 0.068
(0.051)
0.070
(0.050)
0.123
(0.121)
0.030
(0.137)
EPT −0.024
(0.031)
0.026
(0.033)
−0.253
(0.405)
0.034
(0.040)
Popula ion 15–64 0.314***
(0.097)
0.229**
(0.095)
0.333***
(0.105)
0.211**
(0.10)
T ade openness 0.001
(0.001)
0.001
(0.001)
−0.001
(0.003)
0.001
(0.002)
Union densi y 0.001
(0.002)
0.002
(0.002)
−0.001
(0.003)
0.001
(0.003)
ALMP 0.064
(0.080)
0.002
(0.080)
0.156
(0.182)
0.014
(0.091)
Coo d −0.070
(0.053)
−0.053
(0.054)
−0.118
(0.091)
−0.048
(0.056)
Mul ile el 0.057*
(0.030)
0.023
(0.031)
0.145
(0.150)
0.030
(0.034)
Coun y ixed e ec No No No No
Yea ixed e ec Yes Yes Yes Yes
P ob. J-S a is ics 0.646 0.130
R
2
0.677 0.672 0.647 0.671
Obs. 484 484 482 482
Robus s anda d e o s in pa en hesis.
***p <0.01, **p <0.05, *p <0.1.
Sou ce: Own es ima ion
Table 4
EPL and g ow h o unemploymen a e: non-linea and in e ac ion e ec s.
(1) (2) (3) (4)
C 0.177
(0.328)
−1.344
(1.388)
−0.180
(0.196)
−0.388
(0.725)
Unemploymen (-1) 0.461***
(0.064)
0.459***
(0.064)
0.451***
(0.061)
0.451***
(0.060)
GDP g ow h −0.169***
(0.025)
−0.104**
(0.044)
GFCF g ow h −0.048***
(0.008)
−0.004
(0.015)
EPRC −0.094
(0.248)
0.567
(0.848)
0.102
(0.066)
0.442*
(0.263)
EPRC
2
−0.031
(0.047)
−0.024
(0.120)
EPRC*GDP −0.011
(0.021)
EPRC*GFCF −0.013*
(0.007)
EPT −0.016
(0.118)
0.660***
(0.227)
0.069
(0.048)
0.127**
(0.061)
EPT
2
−0.000
(0.025)
−0.105***
(0.039)
EPT*GDP −0.038**
(0.017)
EPT*GFCF −0.016***
(0.005)
Popula ion 15–64 0.309***
(0.097)
0.158
(0.131)
0.307***
(0.098)
0.203
(0.127)
T ade openness 0.001
(0.001)
0.001
(0.004)
0.001
(0.001)
−0.001
(0.004)
Union densi y 0.001
(0.002)
−0.015
(0.010)
0.001
(0.002)
−0.013
(0.008)
ALMP 0.067
(0.081)
−0.084
(0.150)
0.067
(0.080)
−0.090
(0.143)
Coo d −0.058
(0.057)
−0.051
(0.104)
−0.073
(0.052)
−0.116
(0.098)
Mul ile el 0.054*
(0.030)
0.074
(0.056)
0.054*
(0.030)
0.078
(0.057)
Coun y ixed e ec No Yes No Yes
Yea ixed e ec Yes Yes Yes Yes
R
2
0.679 0.689 0.687 0.708
Robus s anda d e o s in pa en hesis.
***p <0.01, **p <0.05, *p <0.1.
Sou ce: Own es ima ion
P. A es is e al.
Economic Modelling 126 (2023) 106437
9
6. Employmen p o ec ion legisla ion and employmen and
employee g ow h
Table 5 analyses he de e minan s o employmen g ow h
12
. As ex-
pec ed, employmen g ow h is di ec ly ela ed o economic g ow h,
whe he measu ed by GDP o in es men g ow h. The g ow h o he
wo king-age popula ion has a posi i e e ec on employmen g ow h.
Rega ding employmen p o ec ion, he coe icien o he EPRC index
is no signi ican . This esul implies ha p o ec ion o pe manen
employmen does no explain employmen dynamics and ha poo job
c ea ion canno be blamed on high p o ec ion o pe manen wo ke s. In
he case o he EPT index, i is only signi ican , wi h a posi i e sign, when
we use LFS da a and GDP g ow h is used as an explana o y a iable. This
esul implies ha g ea e p o ec ion o empo a y wo ke s is associ-
a ed wi h g ea e job c ea ion, he opposi e o wha mains eam econ-
omis s a gue. Howe e , his esul is no obus because EPT is no
signi ican when in es men g ow h and Na ional Accoun s da a a e
used. Wi h ega d o he o he con ol a iables, he esul s a e incon-
clusi e, because hey change wi h he use o GDP o GFCF g ow h and
he use o da a om Labou Fo ce Su eys o Na ional Accoun s.
F om a policy pe spec i e, as in he case o unemploymen , hese
esul s imply ha employmen p o ec ion has no impac on job c ea ion,
and ha , con a y o he expec a ions o hei p omo e s, labou e o ms
ha ha e educed employmen p o ec ion o pe manen and empo a y
wo ke s in Eu ope ha e no con ibu ed o accele a ing employmen
g ow h and ha only policies aimed a accele a ing economic g ow h
can inc ease employmen .
These conclusions a e de i ed om es ima es based on a linea
ela ionship be ween EPL and he g ow h o employmen . Howe e , as
no ed abo e, some s udies sugges ha nega i e e ec s o EPL on labou
ma ke ou comes a e gene a ed when employmen p o ec ion is exces-
si e. Mo eo e , he e ec s o EPL may di e depending on he g ow h
a e o he economy. The e o e, as we made in he case o unemploy-
men , in Table 6 we now p esen he esul s ob ained when es ima ing
he models by es ing, i s , o he exis ence o a non-linea quad a ic
ela ionship be ween he EPL indices and he g ow h o employmen
and, second, o he exis ence o in e ac ion e ec s be ween economic
g ow h and he EPL indices.
In hese es ima es, he only signi ican a iables a e GDP g ow h and
wo king-age popula ion g ow h, bo h o which ha e a di ec e ec on
employmen g ow h. S ikingly, in es men g ow h is only signi ican
when es ing o he exis ence o non-linea EPL e ec s using LFS da a
13
.
The esul s ule ou he exis ence o a non-linea ela ionship be-
ween EPL and employmen g ow h, a esul ha is obus as i is
egis e ed in all models ega dless o he p oxy o economic g ow h and
he sou ce o he employmen da a. This inding unde mines he a gu-
men ha excessi e employmen p o ec ion hampe s job c ea ion, while
a he same ime p o iding no jus i ica ion o ad oca ing labou e-
o ms ha educe employmen p o ec ion in coun ies wi h high le els
o employmen p o ec ion.
Rega ding he in e ac ion e ec s be ween EPL and economic g ow h,
in he case o employmen p o ec ion o pe manen wo ke s, he EPRC
index and he in e ac ion e ec a e signi ican only when using Na ional
Accoun s da a and he in es men g ow h a e. The e o e, we can
conclude ha EPRC does no a ec employmen g ow h. As o he
in e ac ion e ec s be ween employmen p o ec ion o empo a y
wo ke s and economic g ow h, he es ima es show ha , al hough he
EPT index alone is no signi ican , he sign o he in e ac ion coe icien
be ween EPT and economic g ow h is always posi i e and signi ican .
This means ha p o ec ion o empo a y wo ke s con ibu es o
inc easing he posi i e e ec o economic g ow h on job c ea ion.
None heless, he e a e doub s abou he obus ness o his esul . The
eason is ha when we use GFCF g ow h as he explana o y a iable,
in es men g ow h alone has no impac on job c ea ion, as is he case
wi h GDP g ow h.
Finally, wi h ega d o he emaining con ol a iables, he es ima-
ion esul s do no allow us o each a i m conclusion, since he esul s
change depending on he sou ce o he employmen da a, which p e-
en s us om ob aining obus esul s, i.e. independen o he speci ic
speci ica ion o he model.
In summa y, he esul s o ou s udy allow us o conclude ha
employmen p o ec ion is no a signi ican de e minan o employmen
g ow h. This means ha he e o ms implemen ed o educe employ-
men p o ec ion ha e no had he expec ed e ec o speeding up he job
c ea ion p ocess and, i any hing, may ha e con ibu ed o slowing i
down, especially hose e o ms ha ha e educed p o ec ion o em-
po a y employmen . As in he case o unemploymen , a as e pace o job
c ea ion would imply highe economic g ow h, hence he need o
measu es o s imula e economic ac i i y.
Finally, we ha e analysed he impac o employmen p o ec ion
legisla ion on employee g ow h
14
(see Table 7). As expec ed, whe he
we ocus on GDP o GFCF, economic g ow h has a signi ican di ec
Table 5
EPL and employmen g ow h.
Employmen LFS Employmen Na ional Accoun s
(1) (2) (3) (4)
C −0.171
(0.330)
0.344
(0.352)
2.657**
(1.070)
2.957***
(1.065)
Employmen (-1) 0.284***
(0.051)
0.293***
(0.054)
0.427***
(0.058)
0.448***
(0.059)
GDP g ow h 0.358***
(0.051)
0.299***
(0.045)
GFCF g ow h 0.089***
(0.016)
0.081***
(0.013)
EPRC −0.045
(0.093)
−0.046
(0.096)
−0.439
(0.307)
−0.433
(0.301)
EPT 0.112**
(0.056)
0.009
(0.060)
−0.049
(0.093)
−0.113
(0.097)
Popula ion 15–64 0.443**
(0.172)
0.636***
(0.180)
0.380**
(0.186)
0.425**
(0.192)
T ade openness −0.004
(0.003)
−0.003
(0.003)
−0.011
(0.007)
−0.013*
(0.007)
Union densi y −0.008**
(0.004)
−0.010**
(0.004)
−0.020*
(0.012)
−0.017
(0.011)
ALMP −0.290**
(0.167)
−0.162
(0.172)
0.082
(0.198)
0.220
(0.203)
Coo d 0.273***
(0.100)
0.224**
(0.107)
0.221
(0.168)
0.253
(0.170)
Mul ile el −0.104*
(0.061)
−0.038
(0.067)
−0.331***
(0.089)
−0.317***
(0.091)
Ne he lands 1987 12.639***
(1.641)
12.375***
(1.747)
Coun y ixed e ec No No Yes Yes
Yea ixed e ec Yes Yes Yes Yes
R
2
0.663 0.642 0.729 0.723
Obs. 484 484 477 477
Robus s anda d e o s in pa en hesis.
***p <0.01, **p <0.05, *p <0.1.
Sou ce: Own es ima ion
12
We ha e included a dummy a iable o cap u e he e ec o a change in he
Du ch employmen se ies in 1987, based on LFS due o pu ely s a is ical ac o s.
13
See oo no e 10.
14
Among he de e minan s o employee g ow h based on Labou Fo ce Su -
eys we ha e included h ee dummies o cap u e he e ec o a change in
Ne he lands in 1987 and in Swi ze land in 2010 due o pu ely s a is ical ac o s,
and o cap u e he e ec on he Ge man employees se ies in 1991 due o he
euni ica ion p ocess.
P. A es is e al.
