Ramos, Vincen Je ald; Sugui an, Edga An onio
A icle — Published Ve sion
The Incomple e Leap: On he T ansi ion F om Union
Regis a ion o he Fi s Collec i e Ag eemen
Indus ial Rela ions Jou nal
P o ided in Coope a ion wi h:
John Wiley & Sons
Sugges ed Ci a ion: Ramos, Vincen Je ald; Sugui an, Edga An onio (2024) : The Incomple e Leap:
On he T ansi ion F om Union Regis a ion o he Fi s Collec i e Ag eemen , Indus ial Rela ions
Jou nal, ISSN 1468-2338, Wiley, Hoboken, NJ, Vol. 56, Iss. 3, pp. 189-203,
h ps://doi.o g/10.1111/i j.12458
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/323804
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Indus ial Rela ions Jou nal
ORIGINAL ARTICLE
The Incomple e Leap: On he T ansi ion F om Union
Regis a ion o he Fi s Collec i e Ag eemen
Vincen Je ald Ramos
1,2,3
| Edga An onio Sugui an
4
1
He ie School, Be lin, Ge many |
2
Dynamics Resea ch T aining G oup, Humbold Uni e si y Be lin, Be lin, Ge many |
3
ESRC Cen e o Popula ion
Change, Uni e si y o Sou hamp on, Sou hamp on, UK |
4
Depa men o Economics and Poli ical Science, Uni e si y o he Philippines‐Baguio, Baguio,
Philippines
Co espondence: Vincen Je ald Ramos ([email p o ec ed])
Recei ed: 20 Augus 2024 | Re ised: 14 No embe 2024 | Accep ed: 20 No embe 2024
Funding: This esea ch bene i ed om unding om he Philippine Compe i ion Commission's Long‐Te m Resea ch P og am and he co esponding au ho
has ecei ed unding om he Ge man Resea ch Founda ion (DFG), g an numbe 390285477/GRK2458. Funding om he UKRI Economic Social and
Resea ch Council (ES/W002116/1), h ough he Cen e o Popula ion Change‐Connec ing Gene a ions p ojec , is g a e ully acknowledged.
Keywo ds: collec i e ba gaining | e en his o y | union densi y | unioniza ion
ABSTRACT
In decen alized sys ems whe e unioniza ion and ba gaining occu a he es ablishmen le el, wha explains he (speedy)
ansi ion om union egis a ion o he i s collec i e ag eemen ? While p io e idence in some de eloped coun ies (i.e., he
Uni ed S a es) es ima es his ansi ion o be app oxima ely a yea , on a e age, li le o no e idence exis s in o he con ex s
whe e unions play a less cen al ole, as is he case in many de eloping coun ies. Add essing his gap, we analyze he
Philippines whe e na ional unioniza ion and collec i e ba gaining co e age a es a e ela i ely low. Using me hods om
su i al analysis on no el egis e da a o all new union egis a ions om 2016 o 2021, we desc ip i ely demons a e ha : (i)
median union membe ship densi y is low a 40%; (ii) only 20% o all new union egis a ions success ully egis e a con ac
wi hin he i s 5 yea s; and (iii) unions wi h highe densi ies, ha a e independen , and in he manu ac u ing sec o a e
associa ed wi h ele a ed ansi ion a es o a collec i e ag eemen .
JEL Classi ica ion: J51, J52, C14
1 | In oduc ion
In decen alized sys ems whe e ba gaining occu s a he es-
ablishmen le el, unioniza ion is a necessa y bu insu icien
miles one owa ds wo ke s' ep esen a ion. While he implica-
ions o decen alized sys ems on an a ay o ou comes ha e
been s udied ex ensi ely (Ka z 1993; O iguei a 2013; Calm o s
e al. 1988), he dynamics and ansi ions be ween union eg-
is a ion and he signing o he i s collec i e ag eemen ,
especially in de eloping coun y con ex s, ha e ecei ed li le o
no a en ion hus a . Con a y o he Eu opean expe ience wi h
mo e es ablished indus ial ela ions ins i u ions, unions in
Sou heas Asian coun ies, including he Philippines, exe
weake in luence in se ing wages and employmen condi ions,
which was exace ba ed du ing he COVID‐19 pandemic
(Fo d and Wa d 2021). In such con ex s, a puzzle a ises—wha
a e he ‘enabling condi ions’ o unions o secu e a collec i e
ag eemen ?
This pape desc ip i ely ills his gap and con ibu es o ou
unde s anding o his ansi ion. How many unions make he
‘success ul leap’ om unioniza ion o a collec i e ag eemen ?
Which condi ions acili a e (and accele a e) his ansi ion?
While ba gaining dynamics a e a guably he e ogeneous and
shaped by i m‐speci ic and o he exogenous ac o s (i.e.,
nego ia ion s yles), some s uc u al explana ions migh be
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly
ci ed.
© 2024 The Au ho (s). Indus ial Rela ions Jou nal published by B ian Towe s (BRITOW) and John Wiley & Sons L d.
189 o 275Indus ial Rela ions Jou nal, 2025; 56:189–203
h ps://doi.o g/10.1111/i j.12458
ele an isk ac o s. Fo ins ance, unions wi h highe densi ies
(e.g., he sha e o he wo kplace who a e union membe s) migh
expe ience highe ansi ion a es o a collec i e ba gaining
ag eemen (CBA) ela i e o hose wi h lowe densi ies. In ui-
i ely, highe densi ies o e unions s onge le e age in nego-
ia ions as he h ea o ac ion (e.g., s ikes, wo k s oppages) is
p esumably la ge . In neighbo ing Indonesia, i ms a e legally
bound o ba gain in good ai h a e a union is egis e ed o a e
a ce ain h eshold o in e nal suppo is me (P ali asa i, Feby,
and Hamid 2023). As such, ‘size’(o densi y) is a guably c i ical
when s udying union success.
In he Philippines, p i a e sec o unions a e ei he independen
and sel ‐go e ning o a local chap e o a la ge na ional ed-
e a ion. Rega dless, he ba gaining p ocess and he co e age o
any esul ing ag eemen a e con ained wi hin es ablishmen s
(Bi onio 2012). One could a gue ha ede a ed unions may
bene i om e ical and ho izon al coo dina ion and a e
hus mo e likely o succeed in ba gaining (B andl and
B aakmann 2021; O ganisa ion o Economic Co‐ope a ion and
De elopmen 2019). On he con a y, independen unions
likewise bene i om g ea e au onomy and lexibili y as well as
insula ion om poli ical ‘malignmen ’ ha is pe asi e among
Sou heas Asian coun ies (Fo d and Wa d 2021).
This pape ocuses on a speci ic and unde explo ed ou come—
he du a ion be ween union egis a ion and he signing o he
i s collec i e ag eemen . In he Uni ed S a es, Combs (2021)
es ima es his ansi ion o be a li le o e a yea (465 days o
app oxima ely 1.27 yea s). No e ha his is in a con ex whe e
unions mus be sizable enough o win an elec ion and i ms a e
de ju e obliga ed o ba gain in good ai h. To he bes o ou
knowledge, he e is li le o no empi ical e idence add essing
hese ques ions in o he con ex s hus a , owing o he
una ailabili y o ex ensi e egis e da a.
B idging his gap, we collec and compile no el egis e da-
abases o all new union egis a ions and collec i e ag ee-
men s in he Philippines om 2016 o 2021. This da abase
con ains limi ed bu use ul in o ma ion abou he i m and
union names, b oad indus y ca ego ies (1‐digi le el o he
In e na ional S anda d o Indus ial Classi ica ion), geo-
g aphic egion, yea o egis a ion and numbe o membe s
and o al employees. This allows o a sys ema ic and
popula ion‐le el analysis o he ansi ion (bo h selec ion
in o and du a ion) o he i s ag eemen . Using e en his o y
eg ession models on his ma ched egis e da a, we analyze
how ela i e ansi ion a es (o haza d a ios) a e associa ed
wi h h ee s uc u al ea u es—union densi y, union ype
and indus y classi ica ion. These h ee ea u es p oxy o
o ganiza ional and ins i u ional ac o s ha p io li e a u e
has demons a ed o shape i m‐union dynamics. Fi s , un-
ion densi y measu es he in a‐ i m s eng h and le e age o
unions wi h highe densi y ones bene i ing om b oade
wo k o ce suppo . Second, union ype indica es whe he a
union is independen o is a local chap e o a la ge ede -
a ion. Finally, he ype o indus y deno es whe he he un-
ion is in he manu ac u ing sec o . Conside ing ha unions
a e ela i ely mo e common in his sec o , i is in e es ing o
es whe he manu ac u ing unions a e mo e likely o secu e
a collec i e ag eemen .
P e iewing ou indings, we highligh h ee no able esul s and
con ibu ions. Fi s , o he bes o ou knowledge, his pape is
he i s o empi ically u ilize du a ion models in s udying he
ansi ion om unioniza ion o collec i e ba gaining in any
con ex . These me hods a e bes sui ed o esea ch ques ions
whe e he ou come is he ‘ ime‐ o‐e en ’o he du a ion be ween
wo s a es while accoun ing o igh ‐censo ing. Second, we
desc ip i ely demons a e ex emely low CBA egis a ions—
only 1 in 5 o he unions ha ha e egis e ed since 2016 ha e
success ully egis e ed a CBA as o 2021. Finally, ou empi ical
analyses e eal ha high‐densi y, independen and manu ac u -
ing sec o unions ha e obus ly ele a ed ansi ion a es o he
signing o he i s CBA.
Taken al oge he , hese esul s subs an i ely sugges an
‘incomple e leap’ owa ds wo ke s' ep esen a ion in he Phil-
ippines. The majo i y o new union egis a ions do no secu e a
CBA and hose ha we e success ul ook yea s o nego ia e.
Unions wi hou CBAs do no bene i om con ac ually binding
wage inc eases and o he nonwage en i lemen s. In e ec , he
gains o hese ‘con ac less unions’a e inconsis en and limi ed
a bes o nonexis en a wo s . Ou pape emphasizes he need
o indus ial ela ions policies aimed a imp o ing wo ke s'
ep esen a ion o go beyond p omo ing unioniza ion, as has
been he p ac ice in he Philippines. Ins ead, suppo ing unions
h oughou he nego ia ion p ocess will ensu e ha unionized
wo ke s e en ually eap angible and en o ceable gains.
2 | Philippine Con ex
As a o me colony o he Uni ed S a es, he Philippine indus-
ial ela ions landscape is hea ily pa e ned a e he Ame ican
Na ional Labo Rela ions Ac o 1935 (Bi onio 2012)—
unioniza ion and ba gaining a e highly decen alized a he
es ablishmen le el. This is con a y o o he sys ems whe e
‘sup a’ ade unions o ede a ions play a la ge ole, as is he
case in B azil o example (Maia 2019). In ela i ely cen alized
sys ems, ho izon al coo dina ion o unions occu s h ough
ede a ions (i.e., ‘DGB’in Ge many). In heo y, cen alized
ba gaining sys ems unc ion as i he e is one ‘sup a‐union’
(Di ich 2008), whe eas decen alized sys ems de ol e nego i-
a ions a he i m o plan le el, wi h i m‐based unions ope -
a ing independen ly om one ano he (Di ich 2008;
Bi onio 2012). Usual analyses highligh some ine iciencies
b ough abou by cen alized ba gaining, pa icula ly on wages
and p oduc i i y (Calm o s e al. 1988). O iguei a (2013),
Walle s ein and Wes e n (2000) and Ka z (1993) no ed ha
cen alized ba gaining g ew inc easingly common in many
indus ialized socie ies be ween 1950 and 1980, bu ha e since
declined, wi h decen alized s uc u es being a o ed.
These sys ems a e no mu ually exclusi e and, in p ac ice, mos
landscapes a e de ac o ‘hyb id’. Indeed, in some coun ies,
o ganized decen alized s uc u es—whe e sec o le el ag ee-
men s exis bu only o b oad amewo ks o i ms o ollow—
combine sec o al coo dina ion wi h i m‐le el lexibili y o
ba gaining ou comes (O ganisa ion o Economic Co‐ope a ion
and De elopmen 2019) which could yield some e iciencies.
B andl and B aakmann (2021) ind e idence ha e ically co-
o dina ed o mul ile el ba gaining sys ems could lead o be e
190 o 275 Indus ial Rela ions Jou nal, 2025
pe o mance and labo p oduc i i y in Eu opean Union (EU)
coun ies. Howe e , his is con ingen on he ex en o coo di-
na ion be ween ba gaining uni s ac oss he di e en le els
( i m, sec o , coun y), and whe he su ounding ins i u ions
can os e such coo dina ion. I is pe haps mo e app op ia e o
hink o cen alized decen alized sys ems as a spec um a he
han a bina y.
2.1 | Compa isons and Key Di e ences
How does he Philippines' indus ial ela ions landscape di e
om i s neighbo ing Asian coun ies? While unions in Asia
gene ally possess weake in luence in he poli ical sphe e
compa ed o hose in Eu ope (Ku u illa e al. 2002; F enkel and
Ku u illa 2002; Ku u illa and Venka a a nam 1996), ce ain
in a‐ egional ea u es a e no able. In Malaysia, go e nmen
policies p ohibi indus ial o sec o ‐le el unions he eby
esul ing in a ‘de ac o’decen aliza ion a he i m‐le el as
(F eeman 2010; Ku u illa and Venka a a nam 1996). Mean-
while, ansna ional egula o y channels enhance and comple-
men s a e egula ions in Indonesia, consequen ly imp o ing
he en o cemen o and compliance in minimum s anda ds
(Amengual and Chi o 2016; Ca away and Fo d 2017). Finally,
an in e es ing case is also ha o Vie nam, whe e wildca
s ikes—‘unau ho ized’employee ac i i y beyond ailed
ba gaining—in luence s a e policy and he egula ion o labo
s anda ds and ba gaining (Anne and Liu 2016). These ea u es
(e.g., ansna ional ac o s, wildca s ikes and es ic i e go -
e nmen policies) a e a guably less p onounced, i a all, in he
Philippine con ex . Despi e hese di e ences, hese de eloping
Sou heas Asian coun ies a e cha ac e ized by common issues
o agmen a ion, declining union membe ship and suppo and
la ge in o mal sec o s (Asa 2024; McKay 2006; Viaja 2009;
Ku u illa e al. 2002; Ku u illa and Venka a a nam 1996).
As no ed abo e, labo egula ions in he Philippines we e pa e ned
a e he US, conside ing he s ong in luence o pos colonial
Ame ican poli ical and economic ins i u ions (Bi onio 2012;
Hu chison 2016). A p esen , one key di e ence exis s in e ms o
union ecogni ion. In he Philippines, union elec ions a e no held
no manda o y be o e collec i e ba gaining. In he Uni ed S a es
(Bhulle e al. 2022), union ecogni ion consis s o wo main
s eps— i s , wo ke s sign union pe i ions and i hey each a
majo i y o all employees, can seek olun a y ecogni ion di ec ly
wi h he i m. O he wise, hey can ile a pe i ion o hold an elec ion
wi h he Na ional Labo Rela ions Boa d (NLRB) i a leas 30% o
he wo ke s sign o he i m declines o ecognize. Second, a e
winning he elec ions, he i m mus ba gain in good ai h o e
wo king condi ions.
1
In any case, majo i y suppo is gua an eed
be o e ba gaining. In he Philippines, a union can egis e wi h he
Bu eau o Labo Rela ions wi h as li le as 20% o employees
signing up as membe s and could he e o e ba gain wi h he i m
despi e mino i y suppo .
2
His o ically, union in e es , measu ed h ough he median un-
ion densi y (union membe s/employees) has been a a s eady
pe sis en decline. While he e a e ce ainly new union egis-
a ions e e y yea , hei ela i e sizes ha e been sh inking.
Figu e 1shows ha he median union densi y is ini ially
cha ac e ized by some da a noise
3
a e he Philippine inde-
pendence o 1946, ollowed by a ela i ely s agnan phase un il
1990. Since hen, go e nmen e o s o ‘main ain indus ial
peace’occu ed alongside pu suing lexibiliza ion policies (Sale
and Sale 2014). F om 2016 o 2021, ou pe iod o s udy, median
union densi y is a ound 40% bu eached a his o ic low a 27%
in 2021, owing o he COVID‐19 pandemic. As o 2022, union
membe ship and collec i e ba gaining co e age a es (as a sha e
o o al employmen ) in he Philippines a e a his o ic lows a
a ound 6.5% and 8.4% o o al employmen , espec i ely
(Philippine S a is ics Au ho i y 2024).
Finally, ano he challenge ha con on s unions in he Philip-
pines a p esen is he pe sis en ‘ ed‐ agging’, ex a‐judicial
killings and disappea ances o union leade s and o he o ga-
nized wo ke s. Based on ecen epo s om he Human Righ s
Wa ch (2024), 72 union leade s and membe s ha e been killed
since 2016. They ha e also been publicly a acked unde
FIGURE 1 | Median union densi y by yea o egis a ion. Sou ce: Own es ima es using he union egis e da a om he Bu eau o Labo
Rela ions.
191 o 275
accusa ions o being linked o communis g oups as sympa-
hize s o insu gen s. These labels ha e been used o jus i y
iolence and c ackdowns, pu ing unionis s in ac i e dange
(Human Righ s Wa ch 2024). A guably, na ional ede a ions o
la ge unions a e mo e suscep ible o hese isks.
3 | Li e a u e Re iew and Hypo heses
This sec ion begins wi h a ecogni ion ha unioniza ion, in and
o i sel , is a necessa y bu insu icien condi ion owa ds
wo ke s' ep esen a ion. Since i ms inhe en ly possess ou sized
powe ela i e o wo ke s (e.g., using h ea s o e mina ion and
plan shu downs) (Kau man 2012; B on enb enne 2009), un-
ions ha e his o ically coun e ed his asymme y h ough he
signing o binding con ac s, consequen ly imp o ing ea nings
and wo king condi ions (Flanagan 1990; Hi sch 2008;
K is al 2023). D awing om exis ing empi ical li e a u e, we
de elop h ee hypo heses a ound he s uc u al condi ions ha
may acili a e he ansi ion o collec i e ag eemen .
3.1 | Does ‘Size’Ma e ?
As discussed abo e, a key peculia i y in he Philippine case is
ha unions can egis e and seek ecogni ion upon ob aining a
leas 20% o he employee suppo . Hence, i is plausible and
common o see ela i ely ‘small unions’(as a sha e o he i m's
wo k o ce) ba gain wi h he i m. This poses a p oblem, as low‐
densi y unions ce ainly possess less le e age. Indeed, ha ing
mo e o ganized wo ke s leads o la ge poli ical demons a ions
and o e all in luence in public sen imen (Nguyen‐Van and
Te az 2021; Kallas, Pa k, and Aleks 2023). Wi hin i ms, highe
union densi ies can s eng hen how unions can o malize
wo kplace go e nance as mo e wo ke s pa icipa e in he un-
ion's ac i i ies and he ba gaining p ocess (Ed alin 2016;
Kau man 2012).
P io wo k has a gued ha decen aliza ion plays a ole in he
decline o o e all and in a‐ i m union densi y as employe s ake a
mo e ad e sa ial and hos ile s ance o limi he e ec s on hei
p oduc ion p ocess ha hei compe i o s migh no be simila ly
exposed o (Cu che ‐Ge shen eld and Kochan 2004;Schnabel2013).
B on enb enne (2009) no ed ha ad e sa ial employe ac ics such
as h ea s o shu downs, su eillance and e mina ion, among
o he s, impede unioniza ion. Consequen ly, lowe ‐densi y unions
ha e a mo e limi ed ba gaining powe and a mo e p o ac ed
con ac se lemen (Schnabel 2013;Nguyen‐Van and Te az 2021).
In e ms o con ac se lemen , Kallas, Pa k, and Aleks (2023)
ind empi ical e idence ha highe union densi y is posi i ely
associa ed wi h he p obabili y o se ling a i s con ac wi hin
he i s 24 mon hs o union o ma ion. These esul s echo p io
indings ha low union densi y is associa ed wi h low and
p o uded g ow h in ba gaining co e age, despi e he epo ed
numbe o wo ke s who a e membe s o unions (Bi onio 2012;
Lama che 2015).
In he Philippine con ex , Hu chison (2016) poin s ou ha i 's
no jus ha union densi ies a e low, bu also ha mos
es ablishmen s a e small. Hence, he decen alized na u e o
ba gaining pe se cons ains wo ke s' associa ional powe .
Thus, we expec ha low‐densi y unions ha e less ba gaining
le e age and ind i mo e di icul o success ully each an
ag eemen . We deduce his expec a ion om he US expe ience
whe eby he size o he ba gaining uni (ini ial signa o ies o he
union ca ds) is associa ed wi h he elec ion ou come o he
union and hei e en ual ba gaining success (Fe guson 2008;
B on enb enne 2009). As such, we es he hypo hesis ha :
H1: (densi y). Unions wi h highe membe ship densi ies ha e
an ele a ed ansi ion o he i s con ac ag eemen .
3.2 | Does ‘Union Type’Ma e ?
P i a e sec o unions in he Philippines a e classi ied as ei he
o wo ypes—‘independen ’o a ‘local chap e ’o a la ge
ede a ion. Could we expec ha ede a ion chap e s a e mo e
likely o secu e a collec i e ag eemen han independen unions
(o ice‐ e sa)? We d aw on p io li e a u e on union coo di-
na ion and ne wo ks o de elop ou heo e ical expec a ions.
On he one hand, ede a ions possess ad an ages in e ms o
being o mal channels o ho izon al and e ical coo dina ion.
Fede a ions a e mo e ac i ely and di ec ly engaged in ne wo ks
wi h ex e nal pa ies including poli ical and socio‐ci ic o gani-
za ions, and can mo e easily mobilize mo emen s and p o es s
(Kallas, Pa k, and Aleks 2023; Oska sson 2003; Anne , Fische ‐
Daly, and Ma ie 2021; Paolucci and Roche 2024b). In he case
o I eland, ‘con ede a ions’became o mal condui s o ho i-
zon al and e ical coo dina ed ba gaining and in o mal in e ‐
union mechanisms exis ed o coo dina e s a egies in nego ia-
ions (Paolucci and Roche 2024a). G ea e mobiliza ion e o s
inc ease he p essu e on i ms o heed hei demands (e.g.,
s onge e ec s o wo k s oppages) (Kau man 2012). Ku u illa
e al. (2002) and Ku u illa and Venka a a nam (1996) s a ed
ha in many Asian coun ies (i.e., Japan, Ko ea and he Phil-
ippines), he e ha e been a emp s o s eng hen ede a ions in
pu sui o union e i aliza ion, al hough he Philippines has
no ably lagged in his ega d.
Fu he mo e, ede a ions and ne wo ks, in gene al, allow un-
ions o connec wi h in o mal wo ke s h ough socio‐ci ic
o ganiza ions as hey a e o en es ic ed om union membe -
ship (Hu chison 2016; Viaja 2009). Absen he channels o
ede a ion ne wo ks, unions migh no be able o coo dina e
and consolida e he calls o hose in olun a ily le in empo a y
o in o mal wo k a angemen s—‘ emps’o ‘con ac uals’
(Li win and Shay 2022; Asa 2024). In he Philippines, in o mal
wo ke s can only o m wo ke s' associa ions wi h no ba gaining
capaci y, bu hey a e also allowed o a ilia e wi h ede a ions,
hence he usage o ne wo ks by unions (Bi onio 2012). Indeed,
in de eloping coun ies whe e in o mal employmen is ampan
(O ganisa ion o Economic Co‐ope a ion and De elop-
men 2019; Lama che 2015), unions can u ilize ede a ion ne -
wo ks o b oaden hei o ganiza ional ele ance and o he
la ge labo mo emen , as is he case in he Philippines and
Indonesia (Hu chison 2016; Se ano 2018; McKay 2006;
Ca away and Fo d 2017). Being pa o a la ge ede a ion
also allows s onge coali ion‐building h ough coo dina ion,
192 o 275 Indus ial Rela ions Jou nal, 2025
union‐ o‐union ‘lea ning’and mu ual suppo , which can be
use ul du ing he ba gaining s age (Ca dullo, Con i, and
Sulis 2020; Ronnma e al. 2023; Anne , Fische ‐Daly, and
Ma ie 2021; Gumb ell‐McCo mick and Hyman 2019).
The con ex o globaliza ion has also unde sco ed he need o
wo ke s o consolida e hei calls o ac ion h ough ans-
na ional ne wo ks, as hese powe esou ces can also be le e -
aged in con ex s whe e he s a e o e s weak labo egula ions
(Amengual and Chi o 2016; Locke, Amengual, and
Mangla 2009; Hu chison 2016; Ku u illa and Li 2021; Jackson,
Ku u illa, and F ege 2013). Union ede a ions in Asia ha e, o
some ex en , u ilized hese ne wo ks o d i e be e union
coo dina ion, like, o example, being sp ingboa ds o bases o
speci ic bene i s and ba gaining ou comes (Buise 2021;
Ku u illa e al. 2002; F enkel and Ku u illa 2002), howe e ,
he e is s ill oom o imp o ed embeddedness o Asian unions
compa ed o hose in o he de eloped coun ies like in Eu ope
(Ku u illa e al. 2002).
On he o he hand, independen unions, being s and‐alone
en i ies a e also ad an ageous, as hei ba gaining s a egies and
posi ions a e mo e lexible, a ge ed and au onomous
(Hend icks and Kahn 1982), wi h li le in luence by ede a ions.
In libe al ma ke economies such as I eland, he lexibili y o
independen and i m‐le el unions is associa ed wi h expanded
union membe ship and en husiasm as well as secu ing mo e
angible bene i s (Paolucci and Roche 2024b). Fede a ions, as
en i ies, a e compa able o ‘second‐ ie ’unions (Bi onio 2012),
which means ha wo ke s a e subsumed unde wo laye s o
ep esen a ion— he local ede a ion chap e and he na ional
ede a ion leade ship. This addi ional laye migh be dis-
ad an ageous and migh nega i ely in luence he ba gaining
p ocess. One example is how ede a ions migh be emb oiled in
na ional con o e sies o ‘ ed‐ agged’as sympa he ic o ce ain
poli ical mino i ies (e.g., le ‐wing mili an g oups), in which
case hei local chap e s migh su e om ma ginaliza ion and
a weake ba gaining le e age. Indeed, Bi onio (2012) epo s
ha CBA co e age by independen s is highe han ha o ed-
e a ion a ilia es o chap e s.
In he Philippines, ede a ions a e sup a‐le el ne wo ks
composed o a leas 10 local chap e s o wo kplace‐le el
unions. Fede a ion chap e s accoun ed o 60% o all CBAs in
2010−11 (Bi onio 2012). No e ha in he Philippines, i is
possible ha a ede a ion chap e and an independen union
bo h ope a e wi hin he same es ablishmen (Bi onio 2012).
While ede a ions engage wi h ex e nal pa ies and pa ici-
pa e in social mo emen s, lobbying, p o es s and o he
poli ical demons a ions a hena ionalle el(Hi sch2008;
Oska sson 2003), he locus o ba gaining emains a he
wo kplace le el. This ‘ne wo k ad an age’migh aid he
ba gaining p ocess and expedi e he i s collec i e ag ee-
men . Buise (2021) no ed ha unions can be e coo dina e
when hey ha e access o ne wo ks o e ed by ede a ions o
ansna ional o ganiza ions (i.e., unions o sea a e s in he
Philippines bene i ing om ansna ional ne wo ks o sea-
a e s). Indeed, Kallas, Pa k, and Aleks (2023) ind empi ical
e idence in he Uni ed S a es ha coali ion building and
public allies a e associa ed wi h a highe likelihood o se -
ling i s con ac s wi hin plan s.
The e is no conclusi e e idence om he li e a u e on whe he
ede a ion chap e s o independen unions ha e an accele a ed
ansi ion o he signing o a i s con ac and he p eceding
heo e ical conside a ions may poin o ei he di ec ion. I he
‘ne wo k ad an age’o ede a ion chap e s ou weighs he
‘ lexibili y ad an age’o independen unions, one can expec he
o me as mo e likely o secu e a con ac . Along hese lines, we
hypo hesize ha ede a ion chap e s ha e access o mo e
in o ma ion and ne wo k esou ces ha can aid in he ba -
gaining p ocess such ha :
H2: (union ype). Local ede a ion chap e s ha e an ele a ed
ansi ion o he i s con ac ag eemen compa ed o
independen unions.
3.3 | C oss‐Sec o al Di e ences
We know om p io li e a u e ha unioniza ion and collec i e
ba gaining ou comes di e ac oss sec o s. Gi en he his o ical
embeddedness o labo mo emen s in he indus ializa ion o de-
eloped coun ies, i is unsu p ising ha unions a e mo e s ongly
p esen in he manu ac u ing sec o o his day (Ma in and
B ady 2007). Scheue (1997) no es ha he sec o s wi h ‘collec i is
legacies’like manu ac u ing, cons uc ion and anspo ha e
highe union membe ship and ba gaining co e age han he dis-
ibu i e ades o he se ices sec o . As such, manu ac u ing is
seen as he ‘ adi ional union s onghold’(Schnabel 2013,259).
Rela edly, he global decline in unioniza ion is hea ily a ibu ed
o s uc u al ans o ma ion— he shi in employmen composi-
ion om manu ac u ing o se ices, whe e unions g apple o
expand (Schnabel 2013;Hi sch2008).
Among de eloping coun ies, indus ializa ion is expec ed o
inc ease unioniza ion and ba gaining co e age a es i hei
de elopmen ajec o ies ollow ha o oday's indus ialized
and high‐income socie ies (Ma in and B ady 2007). Howe e ,
he economic expansion ha many de eloping coun ies ha e
been expe iencing, including he Philippines, has skipped he
massi e expansion in manu ac u ing in a o o di ec ed g ow h
in se ices (B ady 2007; Se ano 2018). In he Philippines,
manu ac u ing emains he adi ional s onghold o unions,
al hough i has also expe ienced declining membe ship a es
o e ime (Bi onio 2012; Se ano 2018). Ku u illa e al. (2002)
no ed ha s ong an i‐union beha io is p onounced in ce ain
sec o s, pa icula ly in he la ge expo ‐o ien ed elec onics
indus y whe e i ms deem unions as de e en s om in es o s.
In pa icula , bo h he Philippines and Indonesia ha e expe i-
enced s ong ins i u ional ba ie s ha inhibi union coo dina-
ion wi hin expo ‐o ien ed sec o s (e.g., ma ginaliza ion o
wo ke s engaging in s ike ac i i y) (Ca away and Fo d 2017;
Hu chison 2016; Viaja 2009; Ku u illa e al. 2002). E ickson
e al. (2003) no ed ha in he se ice sec o s (e.g., anspo ,
u ili ies and hospi ali y), i ms ha e demons a ed a g ea e
willingness o shu down ope a ions, whe he empo a ily o
pe manen ly. This undamen ally challenges and hampe s un-
ion success in he se ices sec o .
Ou heo e ical expec a ion is a he in ui i e. I union ac i i y
is mo e obus ly p esen in one sec o , he chances o ‘union
success’in e ms o secu ing a collec i e ag eemen migh also
193 o 275
be highe he e. Recognizing he his o ically‐o ien ed and
s ong p esence o unions in he manu ac u ing sec o which
may acili a e he ba gaining p ocess, we es he hypo he-
sis ha :
H3: (sec o ). Manu ac u ing sec o unions ha e an ele a ed
ansi ion o he i s con ac ag eemen ela i e o unions in
nonmanu ac u ing sec o s.
4 | Da a and Me hods
4.1 | Da a and Analy ical Sample
En e p ise‐based unions (o wo kplace‐le el unions) accoun
o a as majo i y o all unions in he Philippines. As o 2024,
he e a e a ound 18,000 p i a e‐sec o unions, accoun ing o
abou 90% o all unions, wi h he es being in he public sec o
(Bu eau o Labo Rela ions 2024). No e ha union egis a ion
in he Philippines does no expi e, hence, i is likely ha many
o hese unions may well be inac i e. O all unionized wo ke s,
only 18% a e cu en ly co e ed by a collec i e ag eemen . This
la ge gap be ween union membe ship and ba gaining co e age
indica es many unions do no o canno success ully o m
ba gaining ag eemen s a he en e p ise‐o es ablishmen ‐le el.
To u he p obe he ansi ion be ween union egis a ion and
collec i e ag eemen , we use wo no el egis e da a o new
union egis a ions and CBAs be ween 2016 and 2021 om he
BLR (n= 858). The union egis a ion da a se con ains basic
in o ma ion abou he union and he i m, he numbe o
membe s and employees, as well as gene al in o ma ion abou
he indus y and geog aphic a ea. Meanwhile, he CBA da a se
con ains ag eemen s egis e ed be ween 2016 and 2021,
ega dless o when hei union was egis e ed. The e o e, we
me ge hese da ase s using bo h i m and union names,
esul ing in a union‐le el da a se wi h an iden i ie o whe he
a CBA was e en ually signed. This unique da a se is a no el y
in he li e a u e as we a e able o es ima e ‘popula ion‐le el’
ansi ions a he han ansi ions among a selec sample o
unions.
Ou key a iables o in e es a e: (1) whe he o no a union has
secu ed a CBA; (2) i yes, how long his ansi ion ook
(in yea s); (3) union ype ( ede a ion chap e o independen
union); (4) densi y (high‐densi y s. low‐densi y); and (5) sec o
g oup (manu ac u ing s. nonmanu ac u ing). Fo addi ional
con ex , we de ine union densi y as he sha e o union membe s
ela i e o he o al numbe o employees. As a s a i ying a -
iable o he ansi ion a es, we de ine high densi y as unions
wi h membe ship densi ies highe han 50% and low densi y o
hose a and below 50%.
4.2 | E en His o y Analysis
As we a e in e es ed in he du a ion be ween wo dis inc s a es,
we u ilize empi ical ools om e en his o y analyses o es i-
ma e haza d a ios. This me hod accoun s o igh ‐censo ing in
ou da a se , whe e he e en has no ye occu ed a he ime o
obse a ion bu could occu in he u u e. In ou s udy, we only
co e he unions ha egis e ed be ween 2016 and 2021.
The e o e, a union ha egis e ed in 2020 migh no ha e ye
secu ed a CBA by he end o 2021, bu i doesn' mean ha i
can' secu e one a a la e poin . E en his o y models accoun
o his ‘ empo al in ensi y’. When igh ‐censo ing is no ad-
d essed, as will be he case when con en ional logis ic eg es-
sion models a e used, his may bias he es ima ion o he ‘ ue’
haza d (Keiding and Ande sen 2015).
4.2.1 | Su i al Cu es
As a i s ‐pass and isual es o di e ences in ansi ion a es,
we employ he s anda d nonpa ame ic es ima ion de ined by
Equa ion 1(Kaplan and Meie 1958), which gi es he su i al
p obabili ies o e ime.
≤
∼
∼
s
D
YX
()= 1−()
X
i
i
i
(1)
The obse a ion imes a e deno ed by X˜
i
,{i=1,…,n}, whe e each
X˜
i
cap u es he ime o su i al X
i
(wi h a CBA) o censo ship
(wi hou a CBA). Fo each obse a ion, D
i
is a dummy ha
indica es whe he he union has a CBA. Meanwhile, Y(X˜
i
)
cap u es all he censo ed unions a ime ha may s ill nego-
ia e a CBA (e.g., ‘a ‐ isk’popula ion) (Kaplan and Meie 1958;
Kleinbaum and Klein 2012). When compa ing su i al cu es,
hypo hesis es ing is conduc ed using log‐ ank es s (Pe o and
Pe o 1972; P en ice and Ma ek 1979).
4
We s a i y he su i al
cu es by union densi y, union ype and b oad sec o g oup, ou
h ee main a iables o in e es .
4.2.2 | Haza d Ra ios
As an ex ension o isually p esen ing he Kaplan−Meie su -
i al cu es, we employ a s anda d Cox p opo ional haza ds
(Cox PH) model ha con ols o addi ional con ounde s,
namely he geog aphic loca ion (island g oup) and he scope o
he ba gaining uni . Ou es imands o in e es a e haza d a ios,
he exponen ia ed coe icien s om he eg ession model, and
could be in e p e ed as he ‘haza d’(e.g., ha ing a CBA) o one
g oup ela i e o a e e ence g oup. Conside he model:
()
h h βX()= ()×exp ii0(2)
whe e he β
i
coe icien s a e ans o med o haza d a ios.
5
Haza d a ios will show he di e ence in ansi ion a es o
independen , high‐densi y and manu ac u ing unions, ela i e
o hei e e ence g oups. One impo an assump ion o he Cox
PH model is ha o p opo ional haza ds, which means ha he
ela i e haza d unc ions do no change o e ime (Keiding and
Ande sen 2015; G ambsch and The neau 1994). We demon-
s a e an empi ical es o his in he Appendix.
6
Al hough s emming om epidemiological li e a u e, he
Kaplan−Meie su i al unc ion and Cox p opo ional haza ds
model ha e been u ilized o unde s and he du a ion o ce ain
194 o 275 Indus ial Rela ions Jou nal, 2025
e en s in he social sciences (Qui oz Flo es 2022). These
models ha e been used o es ima e labo ma ke and indus ial
ela ions phenomena such as unemploymen du a ion
a e college g adua ion in Vie nam (T ang e al. 2024)and
he e ec s o ea ly con lic on he du a ion o ba gaining
ela ionships and he p obabili y o dissolu ion o ba gaining
ela ionships o p i a e sec o i ms in On a io (Weinbe g
2018). We build on his s and o he li e a u e ha uses e en
his o y models o analyze du a ional ou comes gi en he
p esence o censo ing.
5 | Resul s and Discussion
5.1 | Desc ip i e Resul s
We i s show he ends in union egis a ion and collec i e
ag eemen o e ime. Figu e 2shows ha indeed, a as
majo i y o new unions ha e been unable o secu e a collec i e
ag eemen . O he 858 unions in ou analy ical sample, only
16% (o 139) had managed o success ully egis e a CBA.
Table 1shows he composi ion o all union egis a ions, by ou
key explana o y a iables o in e es —union densi y, ype and
sec o . Due o missing alues, he o al numbe o obse a ions
o union densi y is sligh ly lowe . These a e u he dis-
agg ega ed by he CBA indica o (wi h CBA o no CBA), wi h a
co esponding ela i e pe cen age coun o he CBA indica o .
This is o desc ip i ely show how many e en s and censo ed
cases we ha e ac oss di e en le els o ou explana o y a i-
ables. To aid he in e p e a ion o his able, among all new
union egis a ions in ou da a se wi h in o ma ion abou un-
ion densi y, 59% a e classi ied as ‘low densi y’(50% and below
membe ship). Among low‐densi y unions, only 39% ha e
secu ed a CBA compa ed o 61% o high‐densi y unions.
Focusing mo e closely on union densi y, i is in e es ing o no e
ha he e is a di e ence in he dis ibu ion o hose wi h and
wi hou CBAs. Fo he en i e sample, we ind ha he median
union densi y is a ound 40%. S a i ying by hose wi h and
wi hou CBAs, Figu e 3shows ha hose wi hou CBAs ha e a
mo e igh ‐skewed dis ibu ion, indica ing ha mo e unions
wi hou CBAs ha e ela i ely lowe densi ies compa ed o hose
FIGURE 2 | Union and CBA egis a ions. CBA, collec i e ba gaining ag eemen .
TABLE 1 | Summa y s a is ics.
Explana o y a iable F equency % Wi h CBA Rel. % No CBA Rel. %
Union densi y 769 119 650
Low densi y 452 58.8 46 38.7 406 62.5
High densi y 317 41.2 73 61.3 244 37.5
Union ype 858 139 719
Fede a ion chap e 644 75.1 75 54.0 569 79.1
Independen union 214 24.9 64 46.0 150 20.9
Sec o 858 139 719
Nonmanu ac u ing 432 50.3 48 34.5 384 53.4
Manu ac u ing 426 49.7 91 65.5 335 46.6
Abb e ia ion: CBA, collec i e ba gaining ag eemen .
195 o 275
ha ha e one.
7
In e es ingly, unions ha se led a CBA had a
median union densi y o 56%, and hose wi hou we e a 36%,
p o iding he i s desc ip i e suppo o one o ou hypo heses
ha (high) densi y accele a es i s con ac se lemen .
5.2 | Su i al Cu es
Figu e 4shows he Kaplan−Meie su i al cu es o he en i e
da a se (Panel A) as well as disagg ega ed cu es by densi y,
ype and b oad sec o g oup (Panels B−D). Fo mal hypo heses
es ing using log‐ ank es s indica e signi ican di e ences in
su i al p obabili ies by union densi y, union ype and b oad
sec o g oup (see Table A1 in he Appendix). No e he cu es all
depic a consis en ly ‘low ansi ion a e’whe e none has
eached s( )<0.5. This is expec ed as we only obse e 139 CBAs
om 858 new union egis a ions. Consequen ly, he e is no
‘median’ ime‐ o‐e en .
Ou da a sugges s ha wi hin he i s 5 yea s a e union eg-
is a ion, mos unions (80%) do no ha e a CBA. E en high‐
densi y, independen and manu ac u ing sec o unions do no
FIGURE 3 | Union densi ies.
FIGURE 4 | Kaplan−Meie su i al cu es.
196 o 275 Indus ial Rela ions Jou nal, 2025
FIGURE A1 | P opo ional haza ds es ( ully‐speci ied model).
FIGURE A2 | In e ac ion models.
203 o 275
