19
3, XXV, 2022
Economics
10.15240/ ul/001/2022-3-002
CONSUMERS’ PERCEPTIONS OF HEALTH
AND FACTORS INFLUENCING FULFILMENT
OF THE NEED FOR HEALTHCARE IN EU
COUNTRIES
I ena An ošo á1, Naďa Hazucho á2, Jana S á ko á3
1 Mendel Uni e si y in B no, Facul y o Business and Economics, Depa men o Ma ke ing and T ade, Czech Republic,
ORCID: 0000-0002-4331-4187, [email p o ec ed];
2 Mendel Uni e si y in B no, Facul y o Business and Economics, Depa men o Ma ke ing and T ade, Czech Republic,
ORCID: 0000-0002-5693-9872, [email p o ec ed];
3 Mendel Uni e si y in B no, Facul y o Business and Economics, Depa men o Ma ke ing and T ade, Czech Republic,
ORCID: 0000-0002-0889-0218, [email p o ec ed].
Abs ac : The pape deals wi h subjec i e pe cep ions o heal h by indi iduals. The esea ch aimed
a unde s anding socioeconomic and demog aphic ac o s in luencing he ul ilmen o heal hca e
needs and a inding ou ca ego ies o ac o s ha lead o he highes chances o mee ing he need
in consume segmen s o med acco ding o pe cep ions o hei heal h s a us. The analyses we e
based on he EU-SILC da abase o p ima y da a on he income si ua ion and li ing condi ions
o households. In 2017, he da abase included ex a ques ions on heal h. The me hod o clus e
analysis was employed. As a esul , h ee clus e s o indi iduals ep esen ing EU coun ies o med
depending on he pe cei ed s a e o heal h – he au ho s named he clus e s ‘op imis ic’, ‘neu al’,
and ‘pessimis ic’. Fo each segmen , he bina y logis ic eg ession was applied o de e mine
ca ego ies o ac o s leading o he highes p obabili y o mee ing he heal hca e need. The g ea es
in luence o e he ul ilmen o he need o heal hca e has been con i med o he ac o “Sec o
o economic ac i i y”, ollowed by he ype o economic ac i i y. Some di e ences we e e ealed
be ween segmen s. Fo example in he hi d segmen , i.e., esponden s who a ed wo s hei
heal h, a s ong in luence o educa ion has been iden i ied. The highes chances o mee ing he
need o heal h ca e a e achie ed in he i s segmen by execu i es, bu in he second and he hi d
segmen by indi iduals ac i e in educa ion. On he o he hand, c a smen and wo ke s ha e he
lowes chances. In all segmen s, he in luence o household composi ion was con i med, wi h single
households and single-pa en households epo ing lowe chances o mee ing hei heal hca e
needs. Responden s who did no eel hei heal hca e need was me mos ly said i was due o
inancial easons, long wai ing imes, o ea o medical ea men .
Keywo ds: Heal h, need o heal hca e, consume beha iou , income, household.
JEL Classi ica ion: I31, P46.
APA S yle Ci a ion: An ošo á, I., Hazucho á, N., & S á ko á, J. (2022). Consume s’
Pe cep ions o Heal h and Fac o s In luencing Ful ilmen o he Need o Heal hca e in EU
Coun ies. E&M Economics and Managemen , 25(3), 19–34. h ps://doi.o g/10.15240/
ul/001/2022-3-002
In oduc ion
The heal h is de ined as a s a e o a pe son’s
physical, men al, and social well-being.
Responsibili y o heal h is de e mined no
only by he heal hca e sys em and gene ic
p edisposi ions o indi iduals bu also by
one’s li es yle and app oach o achie ing and
keeping a good s a e o heal h (Wo ld Heal h
EM_3_2022.indd 19 15.9.2022 13:48:54
20 2022, XXV, 3
Economics
O ganiza ion, 2006). Consume beha iou
conce ning heal hca e di e s om o he a eas,
abo e all because i is “a ques ion o li e and
dea h”. The e o e, his ype o decision-making
ends o ge signi ican ly a ec ed by emo ions
(Cazacu, 2015). Ano he signi ican di e ence
is ha consume s ge heal hca e p oduc s
and se ices h ough a hi d pa y, mos o en
a physician, who ecommends s eps o be
aken and makes he decisions (Radulescu e
al., 2012). Kenkel (1990) s a es ha physicians
can c ea e o educe demand o hei se ices.
Mee ing heal h ca e needs is no always
a ma e o consume choice, bu o he ac o s
also play a ole.
The main goal o he pape is o e eal
socioeconomic and demog aphic ac o s
in luencing he ul ilmen o EU consume s’
heal hca e needs and o ind ou ca ego ies
o ac o s ha lead o he highes p obabili y
o mee ing he need in consume segmen s
o med acco ding o pe cep ions o hei heal h
s a us. How an indi idual’s heal h is pe cei ed
and, mos impo an ly, whe he heal hca e
needs a e me when hey occu ha e been he
basic esea ch ques ions o he pape . To lea n
abou subjec i e iews on heal h, he au ho s
used he EU-SILC su ey. The su ey p o ides
da a on subjec i e pe cep ions o heal h as
such, as well as in o ma ion abou mee ing
he need o heal hca e and possible easons
o no mee ing his need. The esul s o he
analyses may ep esen a s ong a gumen o
implemen ing imp o emen s in he heal hca e
sys ems. This means in pa icula imp o ing
access o heal hca e se ices o he majo i y
o consume s.
1. Theo e ical Backg ound
People s i e o mee hei heal hca e needs
unde he condi ions se by he heal hca e
sys em and he inancial esou ces hey ha e
a ailable. The a i ude o a household o hei
heal h and he use o heal hca e se ices a ec s
he household’s li ing s anda d (Callande
e al., 2019). Khan and Ul Husnain (2019)
demons a ed ha heal hca e expendi u e and
income a e co-in eg a ed and, he e o e, he e
is a link be ween he s anda d o li ing, income
si ua ion, and heal h s anda d. Lenha (2019)
examined he e ec o income on he s a e o
heal h and ound ha highe income inc eased
he chances o excellen o e y good heal h
being epo ed by households’ heads. The
inc ease obse ed he e anged om 6.9 o
8.9 pe cen age poin s. Acco ding o Knaul e
al. (2012), low-income households li ing nea
he isk-o -po e y h eshold spen mo e on
heal hca e. I means hei heal hca e expenses
accoun ed o a highe pa o hei disposable
income. Howe e , in absolu e numbe s, hey
could a o d ewe heal hca e se ices han
households in highe -income ca ego ies.
Acco ding o Blumbe g e al. (2014), heal h-
ela ed expendi u es a e ising as e han
incomes, bo h a he na ional and household
le els. Sha es o households’ disposable
incomes spen on heal hca e a e inc easing.
In coun ies whe e pa s o he popula ion ha e
no heal h insu ance, he inancial demands o
heal hca e could lead o pe sonal bank up cies.
The subjec i e heal h in Cen al and
Eas e n Eu opean coun ies is in luenced by
a complex mix o de e minan s (Bo iso a,
2019). Di e ences in indi iduals’ socioeconomic
s a uses (s emming om di e en economic
ac i i ies, educa ion, o income ca ego ies) can
con ibu e o heal h inequali ies and o chances
o mee he heal hca e needs. Indi iduals wi h
highe economic s a uses a e mo e in luenced
by beha iou al and psychological ac o s in
hei app oach o heal h han hose wi h lowe
socioeconomic s a uses (A kinson & Ma lie ,
2010; Pe e i-Wa el e al., 2016). Socioeconomic
s a us a ec s heal h- ela ed quali y o li e
(Pucia o e al., 2020). Sel -pe cei ed heal h is
in luenced by income and labou s a us and by
demog aphic ac o s such as gende o age in
EU coun ies (Jind o á & Labudo á, 2020).
Chaupain-Guillo and Guillo (2015) s a e
ha demog aphic ac o s a e o he ac o s ha
in luence consume s’ access o heal hca e.
Gende is one o he ac o s a ec ing app oach
o heal h- ela ed ques ions (Socías e al., 2016;
Roy & Chaudhu i, 2008). The esul s o he
s udy by Roy and Chaudhu i (2008) showed ha
women ended o a e hei heal h wo se and
used ewe heal h se ices – epo edly because
o he lowe socioeconomic s a us o women.
Sonik e al. (2020) added ha disc imina ion
agains women in access o heal hca e was
no necessa ily he eason. Women and men
simply o en had di e en p e e ences as
a as medical ea men is conce ned. Nex
o gende , age is ano he signi ican ac o ,
wi h p e en i e and aes he ic mo i es o
medical ea men p e ailing a younge ages.
Ano he signi ican ac o co-de e mining he
EM_3_2022.indd 20 15.9.2022 13:48:55
21
3, XXV, 2022
Economics
a e age numbe o doc o ’s isi s is educa ion
(Hoeck e al., 2011). Pucia o e al. (2020) see
educa ion as impo an ac o a ec ing an
indi idual’s le el o pe cei ed heal h. Acco ding
o Czibe e e al. (2019) he le el o he highes
a ained educa ion in luence he heal h s a us
indi ec ly. They p o ed ha he educa ion ha e
signi ican impac only on he age when an
illness begun. Pucia o e al. (2020) alks abou
a ma i al s a us as a de e minan o heal h
condi ions. The ma i al s a us is closely ela ed
o he household composi ion. Radulescu e
al. (2012) explain ha amily membe s and
also iends can in luence an app oach o he
indi idual o he heal h. Gende , age and o he
demog aphic and socioeconomic ac o s also
a ec an app oach o indi iduals o heal h isky
beha iou (Mo ke ičius e al., 2020; Kim e al.,
2018). Kunzo á and H ubá (2013) poin ou
ha heal h is co ela ed wi h many ac o s and
also wi h li es yle. Failu e o main ain a heal hy
li es yle pu consume s u i ely a isk o ill
heal h (Mlčocho á & Papežo á, 2012).
The need o heal hca e o medical
ea men may no be me due o a a ie y
o easons. Kenkel (1990) explained he
ela ionship be ween heal hca e and
indi iduals’ le el o knowledge and a ailable
in o ma ion. Acco ding o his au ho , poo ly
in o med consume s ended o unde es ima e
he impo ance o heal hca e. Schmid (2015),
on he o he hand, ound ha in o ma ion had
a nega i e e ec on he use o heal hca e
se ices. This was supposedly ela ed o ea s
o being examined and diagnosed due o which
people did no seek necessa y medical ca e.
Acco ding o Fio illo (2020) and Popo ic e
al. (2017), he mos common easons o no
seeking medical ca e we e inancial and ime
cons ain s and he dis ance o heal h acili ies
(in connec ion wi h he ‘wai -and-see’ app oach
used by he s a o medical acili ies).
I mus be aken in o accoun ha mee ing
heal h- ela ed needs has a s ongly indi idual
dimension and is no a ma e o cou se o all
indi iduals. Sa is ac ion a es a e no he same
o e e yone unde iden ical condi ions (Ban hin
e al., 2008).
The e a e plen y o objec i e indica o s
ha ell us abou he a ailabili y and quali y
o heal hca e in indi idual coun ies. These
objec i e da a speak o he heal hca e sys em
as a whole in e ms o i s quali y, new me hods,
and achie ed esul s. Howe e , he objec i e
da a do no add ess how heal hca e se ices
p o ided a e pe cei ed by indi iduals, whe he
heal hca e is a ailable a he ime and quali y
needed, no wha a e he easons o any
ailu e o mee he need o heal hca e. The
in o ma ion on how indi iduals subjec i ely
pe cei e heal h and heal hca e se ices a e
o u mos impo ance o any esponsible
na ional heal hca e sys em – hence he alue
o subjec i e a iables in analyses in his
a ea (Schokkae e al., 2017). As explained
by Bo iso a (2019), bo h subjec i e and
objec i e indica o s o heal h should be used
whe e e possible because hey o en in e ac
wi h each o he . Heal h policies should adop
a mul idimensional app oach and de elop
incen i es o emo e ba ie s ha limi
consume access o heal h se ices (Popo ic
e al., 2017).
2. Resea ch Me hodology
To lea n abou he beha iou o indi iduals
in ela ion o hei s a e o heal h, he au ho s
used da a ob ained wi hin he EU-SILC su ey
(Eu opean Union – S a is ics on Income and
Li ing Condi ions), speci ically, he EU-SILC
2017. In addi ion, he ex ensi e EU-SILC
mic oda a se p o ided de ailed in o ma ion
on he income si ua ions o households and
indi iduals. The da a also allowed o he
iden i ica ion o households and indi iduals
in e ms o a ious demog aphic and
socioeconomic ac o s, as well as a desc ip ion
o households’ and indi iduals’ li ing condi ions
in di e en a eas o li e. The EU-SILC su ey
is manda o y in all EU coun ies and ollows
a uni o m me hodology published by Eu os a
(Eu os a , 2019). Eu os a also publishes
a uni o m me hodology o u he p ocessing o
he esul s. In 2017, EU-SILC was conduc ed
in a o al o 256,468 Eu opean households and
had a o al o 515,880 indi idual esponden s
( his is he numbe o cases analysed in his
pape ).
The EU-SILC mic oda a da abase o iginally
included 7 indica o s desc ibing subjec i e
pe cep ions o esponden s conce ning he
need and he a ailabili y o heal hca e se ices.
The da abase has been ex ended in 2017
by an ad-hoc module o ano he 7 indica o s
desc ibing he inancial demandingness o
heal hca e, as pe cei ed subjec i ely by
households. This means, o example, he cos
o medicines and den al ca e, o he numbe o
EM_3_2022.indd 21 15.9.2022 13:48:55
22 2022, XXV, 3
Economics
isi s o medical specialis s. The EU-SILC da a
con ain a con e sion ac o which is used as
a weigh in he con e sion o he sample da a o
he base popula ion (i.e., he whole popula ion
o he coun y and he whole EU). A i e-poin
scale (1 – e y good s a e o heal h; 2 – good;
3 – ai ; 4 – poo ; 5 – e y poo ) was used o
subjec i e s a e o heal h assessmen s.
The au ho s used clus e analysis o
iden i y segmen s o EU ci izens ha showed
simila i ies in subjec i e pe cep ions o he
s a e o heal h. Subjec i e assessmen o
heal h e alua ed by consume s is he a iable
applied in he clus e analysis. The clus e s
a e o med acco ding o he p opo ion o
indi iduals among esponden s in each coun y
who a e hei heal h as e y good, good, ai ,
poo and e y poo . The goal o clus e analysis
is o classi y objec s in o a ce ain numbe o
clus e s. Objec s wi hin a clus e a e simila
o he g ea es ex en possible and objec s
wi hin a clus e a e he leas possibly simila o
objec s om o he clus e s. Indi idual objec s
a e g adually g ouped in o smalle clus e s and
hese clus e s a e hen me ged o o m la ge
clus e s (Meloun & Mili ký, 2012). The au ho s
used he K-means algo i hm which iden i ies
homogeneous g oups o esea ch objec s
based on selec ed cha ac e is ics. Fo each
o he ini ial clus e s, he au ho s de e mined
he cen oid alue (cen oid is a ec o o he
a e age alues o each a iable). Objec s we e
assigned o clus e s based on he cen oid
o which he objec was closes . The op imal
numbe o clus e s is e i ied by applying
ANOVA analysis showing signi ican di e ence
be ween clus e s.
Acco ding o Hebák e al. (2015),
K-means algo i hm is an i e a i e p ocedu e
ha minimizes he unc ion o he ollowing
o mula (1):
, (1)
whe e he uih ∈ {0,1} elemen s indica e whe he
he i- h objec belongs ( alue 1) o does no
belong ( alue 0) o he h- h clus e and is
a ec o o a e age alues o he h- h clus e .
The condi ions o he ollowing o mula (2) mus
be me :
(2)
The chances o mee ing he need o
heal hca e wi h espec o di e en ca ego ies o
demog aphic and socioeconomic ac o s ha e
been assessed by logis ic eg ession analysis.
The explained a iable could ake wo alues:
unme need o heal hca e (0) and me need o
heal hca e (1). The ollowing ac o s we e used
as explana o y a iables: gende , educa ion,
economic s a us, sec o o economic ac i i y,
and household income g oup. The bina y
logis ic eg ession model can be exp essed
by he o mula (3) showing he ela ionship
be ween he p obabili y o a phenomenon P(x)
(Y = 1), i.e., mee ing he need o heal hca e,
unde condi ions gi en by he alues o he
independen a iables (x):
. (3)
The ln (P/(1 – P)) o mula (called he logi
o P), can be exp essed as a weigh ed sum o
he alues o he independen a iables. The
logi o P is he loga i hm o he p obabili y o
occu ence o he phenomenon unde s udy.
The model can be also exp essed by he
ollowing o mula (4):
, (4)
whe e he pa ame e es ima es βi a e ob ained
om he measu emen ma ix o x. I βi is equal
o ze o, hen he pa ame e has no e ec on
he obse ed phenomenon (Hendl, 2006).
The quali y o he bina y eg ession model
is assessed by he Nagelke ke R-squa ed
indica o , he signi icance o he model is
e i ied by he Hosme and Lemeshow es .
The VIF indica o is used o e i y a p esence
o mul icollinea i y in models. The VIF alues
highe han 10 indica es mul icollinea i y in he
model (Hebák e al., 2015).
The EU-SILC da a ha e been p ocessed
by he IBM SPSS S a is ics so wa e. The
algo i hm o clus e analysis and he bina y
logis ic eg ession ha e also been implemen ed
in he SPSS so wa e.
3. Resea ch Resul s
The au ho s ook he oppo uni y o analyse da a
om he EU-SILC su ey conduc ed in 2017. In
ha yea , he su ey was ex ended by an ad hoc
module aimed a heal hca e. The esponden s
commen ed on how hey subjec i ely pe cei ed
EM_3_2022.indd 22 15.9.2022 13:48:56
23
3, XXV, 2022
Economics
hei s a es o heal h and whe he hei heal hca e
needs we e me . I a esponden said hei need
o heal hca e was no me , hey we e asked o
gi e he easons. The esul s o he su ey p o ide
impo an in o ma ion on heal h- ela ed beha iou
o people and, gi en he ep esen a i eness o he
popula ion, a e e y use ul o he implemen a ion
o co ec i e measu es in he heal h sec o .
Gi en he size o he su ey sample (co e ing
27 coun ies, i.e., abou 515 housand EU
esponden s and dozens o con en ques ions),
his pape could no co e all he alues included
in he su ey, ins ead, he au ho s ocused on
ypical and ex eme alues only.
3.1 Indi idual Pe cep ions o he S a e
o Heal h
The esul s o he subjec i e assessmen s o he
s a e o heal h showed ha he e we e coun ies
whe e almos 50% o esponden s a ed hei
s a es o heal h as ‘ e y good’ ( o example
Cyp us and G eece). In mos coun ies, a majo
pa o esponden s e alua ed hei s a es o
heal h by he g ade o ‘2’, i.e., ‘good’ ( epo ed by
abou 50% o esponden s), o g ade ‘3’ – ‘ ai ’
(20–30% o esponden s). Howe e , he e we e
coun ies whe e some esponden s (up o 10%
o in he o de o ens o %) a ed hei s a es
o heal h as ‘ e y poo ’ o ‘poo ’. The highes
equencies o nega i e heal h e alua ions
we e ound in he ollowing coun ies (Tab. 1).
In coun ies wi h nega i e heal h a ings
(see Tab. 1), esponden s also mo e equen ly
epo ed issues ela ed o long- e m illnesses
ha limi ed hei e e yday ac i i ies. In o he
EU coun ies ( hose no lis ed in Tab. 1), 2%
o ewe esponden s assessed hei s a es o
heal h as e y poo .
To p o ide an o e all o e iew and summa i-
ze he subjec i e pe cep ion o he s a e o heal h
in all EU coun ies, he au ho s employed clus e
analysis and he K-means algo i hm. As a esul ,
h ee clus e s o indi iduals we e iden i ied
based on he pe cei ed s a uses o heal h. The
p opo ions o indi iduals e alua ing hei heal h
s a us as e y good, good, ai , poo and e y
poo enabled he o ma ion o h ee segmen s
and so ed coun ies in o segmen s acco ding
assessmen s by esiden s’ ep esen a i es
(Tab. 2).
C oa ia Po ugal Hunga y La ia Li huania Poland Bulga ia
Ve y poo s a e
o heal h 3.9% 3.6% 3.2% 3.1% 3.1% 2.7% 2.5%
Poo s a e
o heal h 14.0% 11.0% 9.0% 13.8% 13.0% 10.0% 8.0%
Sou ce: EU-SILC mic oda a (Eu os a , 2021), own using IBM SPSS S a is ics
Tab. 1: EU coun ies wi h he highes p opo ions o indi iduals pe cei ing nega i ely
hei s a es o heal h
Clus e 1
‘op imis ic’
Clus e 2
‘neu al’
Clus e 3
‘pessimis ic’
EU coun ies
in he clus e
Aus ia, Cyp us,
G eece, C oa ia,
I eland
Belgium, Bulga ia, Ge many,
Denma k, Spain, Finland,
F ance, I aly, Luxembou g,
Mal a, Ne he lands, Romania,
Sweden, Slo akia
Czech Republic,
Es onia, Hunga y,
Li huania, La ia,
Poland, Po ugal,
Slo enia
Ve y good SH 40% 23% 13%
Good SH 33% 48% 41%
Fai SH 18% 21% 32%
Poo SH 7% 6% 11%
Ve y poo SH 2% 2% 3%
Sou ce: EU-SILC mic oda a (Eu os a , 2021), own using IBM SPSS S a is ics
Tab. 2: Subjec i e assessmen s o heal h (SH) in EU coun ies
EM_3_2022.indd 23 15.9.2022 13:48:56

24 2022, XXV, 3
Economics
Subsequen ly, ANOVA analysis con i med
he co ec numbe o clus e s iden i ied (Tab. 3).
Signi icance alues a e below he signi icance
le el α = 0.05. Clus e s a e signi ican ly di e en .
Also, he e was a ze o change acco ding o he
i e a ion his o y a e h ee i e a ions du ing
K-means algo i hm p ocess. I mo e clus e
we e o med, he di e ence be ween clus e s
was no con i med (signi icance alues we e
abo e he signi icance le el).
The K-means algo i hm assigned
indi iduals om i e coun ies o he i s
clus e , o which almos h ee qua e s a ed
hei heal h as good and 40% as e y good.
Due o he posi i e heal h assessmen s, he
clus e has been named as ‘op imis ic’. In he
second g oup, abou hal o he esponden s
a ed hei heal h as good. The second segmen
included he la ges numbe o EU coun ies
compa ed o he o he segmen s. The hi d
g oup has been mo e pessimis ic abou hei
heal h, wi h a highe numbe o esponden s
a ing hei heal h as ai . On a e age, 14% o
esponden s in his g oup e alua ed hei heal h
as poo . Responden s om coun ies wi h mo e
nega i e a ings we e mo e likely o epo
p oblems ela ed o long- e m illness o heal h
limi a ions.
The a e age sha e o a coun y’s popula ion
epo ing limi a ions in e e yday ac i i ies due o
poo heal h ha e amoun ed o uni s o pe cen .
Ye , he e we e coun ies whe e people did no
pe cei e such limi a ions a all (Spain, I eland,
Mal a, and Sweden). Howe e , when d awing
hese conclusions, we need o ake in o accoun
whe he he condi ions c ea ed by he s a e a e
so sa is ac o y ha people can lead ac i e li es
wi hou limi a ions, o whe he he epo ed
opinions we e shaped by low awa eness o he
possibili ies o imp o ing li ing condi ions.
3.2 Pe cei edFul ilmen o Heal hca e
Needs
When asked whe he he medical assis ance
eques ed was ac ually ecei ed, he e we e
coun ies whe e almos 100% o esponden s
answe ed posi i ely. These we e, o example,
Spain, Aus ia, Mal a, and Luxembou g. In
some coun ies, on he o he hand, signi ican
amoun s o esponden s answe ed nega i ely,
i.e., ha hey did no ecei e he ea men
hey needed. In G eece, o example, 25%
o esponden s ga e nega i e answe s, in
Es onia, i was 13% o esponden s, in Poland
12%, and in La ia 10%. In o he EU coun ies,
unme heal hca e needs we e epo ed by up o
10% o esponden s.
The au ho s used bina y logis ic eg ession
o ind ou which ac o s in luenced he ul ilmen
o he need o heal hca e and which ca ego ies
o demog aphic and socioeconomic ac o s
inc eased he chances o he ul ilmen o he
need.
The explained a iable in he model has
been he ul ilmen o he need o heal hca e.
The a iable could ake wo alues: 0 indica ing
no sa is ac ion o he need ( ailu e o mee
he need o heal hca e); and 1 indica ing
sa is ac ion o he need. The explana o y
a iables en e ing he eg ession model we e
Gende , Educa ion, Household composi ion,
Economic ac i i y, Income quin ile based on he
household’s disposable income, and Sec o o
economic ac i i y based on ISCO (In e na ional
S anda d Classi ica ion o Occupa ions).
The au ho s ha e calcula ed he bina y
logis ic eg ession o all h ee segmen s
(c ea ed based on he subjec i e assessmen s
o heal h by he esponden s – see Tab. 2). This
allowed o explana ions o he esul s o bina y
logis ic eg essions in ela ion o op imis ic and
Clus e E o F Sig.
Mean squa e d Mean squa e d
Ve y good SH 1,124.076 236.397 24 30.884 0.000
Good SH 428.770 222.651 24 18.930 0.000
Fai SH 415.434 215.236 24 27.267 0.000
Poo SH 66.515 25.447 24 12.211 0.000
Ve y poo SH 3.953 20.626 24 6.315 0.006
Sou ce: EU-SILC mic oda a (Eu os a , 2021), own using IBM SPSS S a is ics
Tab. 3: ANOVA in he clus e analysis
EM_3_2022.indd 24 15.9.2022 13:48:57
25
3, XXV, 2022
Economics
pessimis ic assessmen s o heal h by indi idual
esponden s.
Be o e in e p e ing he model, he p esence
o mul icollinea i y in h ee models o all
segmen s was e i ied. The linea eg ession
p ocedu e wi h same p edic o s was used o
his pu pose and collinea i y diagnos ics we e
eques ed. All alues o VIF indica o s a e below
he alue 10 (Tab. 4). Mul icollinea i y is no
p esen in he models, as indica ed by he low
alues o he Condi ion indexes implemen ed in
he IBM SPSS S a is ics.
In all h ee logis ic eg essions, Hosme
and Lemeshow es s we e used o p o e
he signi icance o he models ( he esul ing
p- alues had o be lowe han 0.05). The
Nagelke ke R-squa ed indica o s p o ed he
quali y o he models as 83% o he a iabili y
in he dependen a iable was explained o
he i s segmen (Tab. 5), o he second model
(Tab. 6) i was 89%, and o he hi d model
(Tab. 7) i was 73% o he a iabili y o he
dependen a iable. Ca ego ies wi h he highes
chances o mee ing he heal h o heal hca e
ha e been highligh ed in bold in he ables.
The esul s o he bina y logis ic eg ession
o he i s segmen (Tab. 5) showed ha all he
explained a iables in luenced he ul ilmen o
he need o heal hca e. The s onges in luen-
ce has been iden i ied in he ‘Sec o ’ a iable,
Segmen 1
VIF
Segmen 2
VIF
Segmen 3
VIF
Gende 9.446 8.677 9.347
Household composi ion 4.994 4.733 5.349
Educa ion 9.654 9.847 9.822
Economic ac i i y 3.549 3.585 3.705
Quin iles 5.791 6.041 6.184
Sec o 4.496 5.096 5.435
Sou ce: EU-SILC mic oda a (Eu os a , 2021), own using IBM SPSS S a is ics
Tab. 4: Collinea i y s a is ics
Es ima e B S anda d
de ia ion Wald d Sig. Exp(B)
Gende (males) 0.074 0.003 863.334 1 0.000 1.077
Household composi ion
(o he )a 30,666.113 4 0.000
Household composi ion
(single) 0.024 0.003 48.946 1 0.000 1.024
Household composi ion
( wo adul s) 0.433 0.003 20,608.297 1 0.000 1.542
Household composi ion
(single pa en ) 0.228 0.009 692.013 1 0.000 1.256
Household composi ion
( wo adul s and child en) 0.378 0.004 10,978.219 1 0.000 1.460
Educa ion (uni e si y)a 138,956.161 20.000
Educa ion (basic) −0.382 0.004 11,072.918 1 0.000 0.682
Educa ion
(seconda y/high school) 0.681 0.003 52,078.088 1 0.000 1.976
Tab. 5: Chances o mee ing he need o heal hca e o segmen 1 – ‘op imis ic’ – Pa 1
EM_3_2022.indd 25 15.9.2022 13:48:57
26 2022, XXV, 3
Economics
whe e legisla o s and execu i es we e 11 imes
mo e likely o ha e hei heal hca e needs me
han c a smen and wo ke s. The second mos
impo an a iable in e ms o in luence signi i-
cance was Economic ac i i y, whe e employees
we e ound o ha e he highes chances o he
ul ilmen o hei need o heal hca e.
Simila ly, he esul s o he logis ic
eg ession o he second segmen da a (Tab. 6)
p o ed he signi icance o mos o he ac o
ca ego ies excep o he ca ego y o ag icul u e
in he Sec o a iable and he ca ego y o
single pa en s in he Household composi ion
a iable. The mos signi ican ac o in e ms
Es ima e B S anda d
de ia ion Wald d Sig. Exp(B)
Economic ac i i y (o he )a 152,401.000 4 0.000
Economic ac i i y
(employed) 1.225 0.004 115,136.806 1 0.000 3.403
Economic ac i i y
(sel -employed) 0.510 0.005 11,126.210 1 0.000 1.666
Economic ac i i y
(unemployed) −0.035 0.004 66.343 1 0.000 0.965
Economic ac i i y
(old-age pensione ) 0.669 0.003 36,924.689 1 0.000 1.952
Quin iles ( i h)a 27,200.371 4 0.000
Quin iles ( i s ) 0.467 0.004 16,704.470 1 0.000 1.595
Quin iles (second) 0.460 0.004 17,002.700 1 0.000 1.584
Quin iles ( hi d) 0.411 0.004 13,717.567 1 0.000 1.508
Quin iles ( ou h) 0.436 0.004 14,494.801 1 0.000 1.546
Sec o (c a smen
and wo ke s)a 358,990.333 10 0.000
Sec o (legisla o s
and execu i es) 2.430 0.006 158,988.056 1 0.000 11.354
Sec o (science
and echnology) 1.892 0.007 68,520.486 1 0.000 6.630
Sec o (heal hca e) 1.816 0.008 46,283.147 1 0.000 6.147
Sec o (educa ion
and aining) 2.118 0.008 67,896.034 1 0.000 8.311
Sec o (public
adminis a ion) 2.051 0.007 83,857.200 1 0.000 7.777
Sec o (in o ma ion
echnology) 1.840 0.012 23,631.131 1 0.000 6.293
Sec o (law, cul u e, spo ) 1.445 0.008 33,479.194 1 0.000 4.243
Sec o (o icials) 1.385 0.005 79,697.080 1 0.000 3.995
Sec o (se ices and sales) 0.986 0.003 95,813.287 1 0.000 2.681
Sec o (ag icul u e,
o es y, ishing) 0.488 0.003 20,737.801 1 0.000 1.630
Sou ce: EU-SILC mic oda a (Eu os a , 2021), own using IBM SPSS S a is ics
No e: a This pa ame e has been se o ze o because i is edundan .
Tab. 5: Chances o mee ing he need o heal hca e o segmen 1 – ‘op imis ic’ – Pa 2
EM_3_2022.indd 26 15.9.2022 13:48:58
27
3, XXV, 2022
Economics
o inc easing he likelihood o ul illing he need
o heal hca e has been he Sec o a iable
again, whe e he employees in he heal hca e
sec o , he educa ion and aining sec o had
he highes chances o ha ing hei heal hca e
needs me . The chances o bo h ca ego ies
we e almos 7 imes highe compa ed o
c a smen and wo ke s. The esul s ha e also
shown ha people wi h p ima y educa ion had
he highes chances o he ul ilmen o hei
heal hca e needs (e en h ee imes highe
compa ed o uni e si y g adua es). This
inding may be ela ed o he ac ha p ima y
educa ion (as he highes le el o educa ion
a ained) was epo ed la gely by elde ly
esponden s who we e no longe economically
ac i e and had su icien ime o heal hca e.
Ac ually, ime cons ain s we e one o he main
easons o no ul illing he need o heal hca e.
I is wo h no ing ha he lowes chances we e
iden i ied in he g oups o single mo he s and
single households ( he Household composi ion
a iable), in he i s segmen (Tab. 5).
Es ima e B S anda d
de ia ion Wald d Sig. Exp(B)
Gende (males) 0.452 0.001 215,965.723 1 0.000 1.571
Household composi ion
(o he )a 462,436.433 4 0.000
Household composi ion
(single) 0.132 0.001 11,210.770 1 0.000 1.141
Household composi ion
( wo adul s) 0.620 0.001 262,190.359 1 0.000 1.859
Household composi ion
(single pa en ) 0.003 0.003 1.207 1 0.272 1.003
Household composi ion
( wo adul s and child en) 0.690 0.001 255,848.438 1 0.000 1.994
Educa ion (uni e si y)a 868,163.355 20.000
Educa ion (basic) 1.194 0.002 542,042.619 1 0.000 3.299
Educa ion (seconda y/high
school) 0.929 0.001 755,609.824 1 0.000 2.531
Economic ac i i y (o he )a 1,174,238.320 4 0.000
Economic ac i i y
(employed) 1.204 0.001 1,026,477.356 1 0.000 3.333
Economic ac i i y
(sel -employed) 1.075 0.002 287,131.505 1 0.000 2.931
Economic ac i i y
(unemployed) 0.546 0.002 102,352.112 1 0.000 1.726
Economic ac i i y
(old-age pensione ) 1.009 0.001 588,650.365 1 0.000 2.744
Quin iles ( i h)a 320,935.300 4 0.000
Quin iles ( i s ) 0.056 0.001 1,740.229 1 0.000 1.057
Quin iles (second) 0.361 0.001 69,696.199 1 0.000 1.434
Quin iles ( hi d) 0.615 0.001 186,094.038 1 0.000 1.850
Quin iles ( ou h) 0.554 0.001 150,418.704 1 0.000 1.741
Tab. 6: Chances o mee ing he need o heal hca e o segmen 2 – ‘neu al’ – Pa 1
EM_3_2022.indd 27 15.9.2022 13:48:59
34 2022, XXV, 3
Economics
S a us o he Unemployed. E&M Economics
and Managemen , 23(3), 23–37. h ps://doi.
o g/10.15240/ ul/001/2020-3-002
Radulescu, V., Ce ina, I., & O zan, G.
(2012). Key Fac o s ha In luence Beha io
o Heal h Ca e Consume , he Basis o Heal h
Ca e S a egies. Con empo a y Readings in
Law, 4(2), 992–1001.
Roy, K., & Chaudhu i, A. (2008). In luence
o socioeconomic s a us, weal h and inancial
empowe men on gende di e ences in heal h
and heal h ca e u iliza ion in la e li e: e idence
om India. Social Science & Medicine,
66(9), 1591–1962. h ps://doi.o g/10.1016/j.
socscimed.2008.01.015
Schmid, C. (2015). Consume Heal h
In o ma ion and he Demand o Physician
Visi s. Heal h Economics, 24(12), 1619–1631.
Schokkae , E., S eel, J., & Van de Voo de,
C. (2017). Ou -o -Pocke Paymen s and
Subjec i e Unme Need o Heal hca e. Applied
Heal h Economics and Heal h Policy, 15(5),
545–555. h ps://doi.o g/10.1007/s40258-017-
0331-0
Socías, M. E., Koehoo n, M., & Sho elle , J.
(2016). Gende Inequali ies in Access o Heal h
Ca e among Adul s Li ingin B i ish Columbia,
Canada. Women’s Heal h Issues, 26(1), 74–79.
h ps://doi.o g/10.1016/j.whi.2015.08.001
Sonik, R. A., C eedon, T. B., P ogo ac, A.
M., Ca son, N., Delman, J., Delman, D., & Cook,
B. L. (2020). Dep ession ea men p e e ences
by ace/e hnici y and gende and associa ions
be ween pas heal hca e disc imina ion
expe iences and p esen p e e ences in
a na ionally ep esen a i e sample. Social
Science & Medicine, 253, 112939. h ps://doi.
o g/10.1016/j.socscimed.2020.112939
Wo ld Heal h O ganiza ion. (2006).
Cons i u ion o he Wo ld Heal h O ganiza ion.
Re ie ed Ma ch 16, 2022, om h ps://www.
who.in /go e nance/eb/who_cons i u ion_en.pd
EM_3_2022.indd 34 15.9.2022 13:49:01