S udies Managemen and Finance Economics, o Jou nal
0504-2644 (online): ISSN 0490,-2644 (p in ): ISSN
5202 Oc obe 10 Issue 80 Volume
8.317 Fac o : Impac ,05-i10-10.47191/je ms/ 8 DOI: A icle
3266 -6618 No: Page
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6618
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion
in Macedonia
Boban Tomeski
Gene al Di ec o –Execu i e Membe o he Boa d o Di ec o s C oa ia Insu ance AD Skopje – Li e
ABSTRACT: The pu pose o his s udy is o in es iga e he long- un de e minan s o economic g ow h by examining he ela ionship
be ween selec ed mac oeconomic a iables using ime se ies da a o he pe iod 2003–2024. Pa icula emphasis is placed on
e alua ing how s uc u al and demog aphic ac o s in luence g ow h dynamics in he con ex o a small open economy. This analysis
con ibu es o he empi ical li e a u e by p o iding upda ed e idence o he pos - ansi ion pe iod and o e ing insigh s ele an o
policy design. To achie e his objec i e, he s udy employs coin eg a ion echniques o e i y he p esence o long- un equilib ium
ela ionships among he a iables. Once coin eg a ion is con i med, he Fully Modi ied O dina y Leas Squa es (FMOLS) and Dynamic
O dina y Leas Squa es (DOLS) es ima o s a e applied o ob ain e icien and unbiased es ima es o he long- un coe icien s. The
empi ical esul s sugges ha in es men , ade openness, and human capi al exe a posi i e and s a is ically signi ican e ec on
long- un economic g ow h, while high public deb and in la ion exe a nega i e impac . Demog aphic ac o s such as dependency
a ios also show a measu able in luence, wi h young dependency exe ing a g ow h-dampening e ec . The consis ency o he esul s
ac oss FMOLS and DOLS ein o ces he obus ness o he indings. Based on hese esul s, he s udy ecommends policies aimed a
os e ing in es men in physical and human capi al, main aining mac oeconomic s abili y h ough p uden iscal managemen , and
implemen ing s uc u al e o ms o enhance p oduc i i y. S eng hening demog aphic policies o mi iga e he bu den o dependency
a ios and p omo ing inclusi e g ow h s a egies a e also emphasized as c ucial o ensu ing sus ainable long- un g ow h.
KEYWORDS: Economic ac o s, Demog aphic ac o s, Ins i u ional d i e s, Li e insu ance consump ion, Macedonia
INTRODUCTION
Li e insu ance plays a signi ican ole in he inancial sec o . Li e insu ance policies ep esen inancial p oduc s ha p o ide
wo main se ices: income eplacemen in he e en o p ema u e dea h and a mechanism o long- e m sa ings. The e a e se e al
ypes o li e insu ance, each wi h i s own cha ac e is ics, bu all can gene ally be classi ied in o h ee ca ego ies: e m li e insu ance,
which p o ides paymen o he ag eed sum in he e en o he insu ed’s dea h; endowmen insu ance, whe e he insu e is obliged o
pay he sum only i he insu ed su i es he ag eed pe iod; and whole/mixed li e insu ance, which ep esen s a combina ion o he
p e ious wo ypes, wi h he ag eed sum being paid ei he upon su i al o immedia ely upon he dea h o he insu ed i i occu s
be o e he end o he e m (Jo ano ski, 2005).
A he global le el, li e insu ance ma ke s in ecen decades ha e aced signi ican s uc u al changes. In addi ion o
con inuous changes in egula o y amewo ks, many coun ies ha e libe alized and de egula ed hei insu ance ma ke s in o de o
s imula e compe i ion (Boonyasai e al., 2002; Camino-Mog o e al., 2019; Chand apal, 2019). These p ocesses ha e been
accompanied by a wa e o me ge s and acquisi ions, pa icula ly in Eu ope and he Uni ed S a es (Bo ges e al., 2008; Cummins e al.,
2015; Da u yan & Klumpes, 2008; Klumpes, 2022; Zanghie i, 2009), as well as an inc ease in c oss-bo de ade in insu ance se ices
(Cummins e al., 2017). A he same ime, li e insu e s a e acing p essu es o inno a e and imp o e e iciency, d i en by global
economic shi s, compe i ion om banks, mu ual unds, and in es men ad iso y i ms, low in e es a es (Eling & Schape , 2017),
aging popula ions, educ ions in public pensions (Swiss Re Ins i u e, 2019), and echnological ad ances in sales, unde w i ing, claims
p ocessing, and cus ome se ice (Cummins e al., 1999).
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6619
Mo e ecen li e a u e examining he li e insu ance ma ke in he egion, and pa icula ly in Macedonia, highligh s he
in luence o mac oeconomic and ins i u ional ac o s on consump ion. Gocko and Kamenja ska (2021) show ha he p o i abili y o
insu ance companies in No h Macedonia depends on ac o s such as ma ke sha e, in la ion, inancial in es men s, and economic
g ow h. Addi ionally, egional s udies on he Wes e n Balkans con i m ha pe capi a income, wages, and economic g ow h posi i ely
a ec li e insu ance consump ion, while unemploymen and in e es a es ha e a nega i e impac (S ojano ić e al., 2021). F om an
ins i u ional pe spec i e, s udies on ansi ion economies emphasize ha he quali y o egula ions, he ule o law, and us in
ins i u ions a e impo an de e minan s o consump ion (Hoxha & Bak açi, 2017). Pa icula ly impo an is he con ibu ion o
Kjose ski (2012), who, o Cen al and Sou heas e n Eu ope, including Macedonia, shows ha income, in la ion, he sa ings a e, and
inancial sec o de elopmen a e he main ac o s de e mining he demand o li e insu ance.
Gi en he impo ance o li e insu ance o economic and inancial de elopmen , he ollowing ques ion a ises: which
demog aphic, economic, and ins i u ional ac o s de e mine li e insu ance consump ion in he Republic o Macedonia? Add essing
his ques ion will p o ide insigh s in o how li e insu ance consump ion is shaped and will o e guidance o policymake s in designing
be e s a egies o ma ke de elopmen . The ollowing sec ions will p esen he heo e ical ounda ions o he economic and
ins i u ional de e minan s o li e insu ance consump ion, a e iew o he ele an li e a u e, an analysis o he cu en s a e o he li e
insu ance ma ke in Macedonia, he choice o a iables and esea ch me hodology, ollowed by he p esen a ion o esul s and
co esponding ecommenda ions.
The emainde o he pape is s uc u ed as ollows: Sec ion 2 e iews he ele an li e a u e; Sec ion 3 in oduces he da a
and p esen s s ylised ac s; Sec ion 4 ou lines he econome ic amewo k; Sec ion 5 discusses he empi ical esul s; and Sec ion 6
concludes wi h policy implica ions.
2. LITERATURE REVIEW
2.1. Theo y o he De e minan s o Li e Insu ance Consump ion
The heo e ical amewo k o explaining li e insu ance consump ion ini ially began o de elop in he wo ks o Yaa i (1965)
and Hakansson (1969). Wi hin his amewo k, he analysis ocuses on he demand side o li e insu ance. The demand o li e insu ance
is based on he insu ed’s desi e o p o ide inancial suppo o dependen s (spouse and child en) in he e en o hei dea h, as well
as o secu e income o e i emen .
Acco ding o his amewo k, he demand o li e insu ance depends on weal h, he expec ed li e ime income o he insu ed,
in e es a es, he cos s associa ed wi h li e insu ance policies (adminis a i e expenses), and he assumed subjec i e discoun a e,
e lec ing he p e e ence o cu en consump ion o e u u e consump ion.
This ounda ional model was u he ex ended by Lewis (1989), who explici ly inco po a ed he p e e ences o he insu ed’s
dependen s (spouse and child en), who a e he ac ual bene icia ies o he insu ance in he e en o he insu ed’s dea h. Lewis de i es
he demand o li e insu ance as a u ili y-maximiza ion p oblem o he bene icia ies, assuming an isoelas ic u ili y unc ion, and
p esen s i as ollows:
(1-lp)F = max{[(1-lp) / l(1-p)]1/δTC – W,0} (1)
Whe e l ep esen s he a io o li e insu ance cos s o ac ua ial alue, p is he p obabili y o dea h o he household’s main
income ea ne ( he insu ed), F is he nominal alue o all li e insu ance policies held by he insu ed o e hei li e ime, δ is a measu e
o he ela i e isk a e sion o he insu ance bene icia ies, TCis he p esen alue o consump ion o each bene icia y o e hei
expec ed emaining li e ime, and W ep esen s he ne weal h o he insu ed’s amily.
Acco ding o his equa ion, he demand o li e insu ance inc eases wi h a highe p obabili y o dea h o he household’s main
income ea ne , a highe p esen alue o consump ion o he bene icia ies, and g ea e isk a e sion. Con e sely, he demand
dec eases as he a io o li e insu ance cos s o ac ua ial alue ises and as he ne weal h o he insu ed’s amily inc eases.
Beck and Webb (2003) no e ha li e insu ance consump ion is de e mined no only by demand-side ac o s bu also by supply-side
ac o s ha in luence cos s and, consequen ly, he p ice o li e insu ance. Key supply-side ac o s a ec ing li e insu ance cos s include
human and in o ma ional esou ces. Insu ance companies equi e hese esou ces o accu a ely measu e cos s and ese e needs, as
well as o decide how o in es he unds collec ed om li e insu ance sales in inancial ma ke ins umen s. These supply-side ac o s
a ec li e insu ance cos s and can be inco po a ed in o he a io o li e insu ance cos s o ac ua ial alue in equa ion (1).
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6620
Bo ch (1990) and Sen (2008) explain ha he heo y o li e insu ance consump ion is oo ed in he li e-cycle consump ion
hypo hesis, he Keynesian consump ion hypo hesis, and he pe manen income hypo hesis. Acco ding o hese au ho s, li e insu ance
consump ion is iewed as a o m o sa ings, which can be exp essed by he ollowing equa ion:
s( ) = eδ ∫e-δ {y( ) – c( )}ds (2)
Acco dingly, as de i ed om he a o emen ioned heo ies, sa ings s( ) ep esen he po ion o income ha emains a e
mee ing consump ion needs. Speci ically, i we conside an indi idual wi h income y( ) and consump ion c( ), bo h ep esen ed as
con inuous unc ions o e ime, hen ne sa ings o e he ime in e al om 0 o (whe e ep esen s he e i emen pe iod) a e
de ined as he di e ence be ween income y( ) and he maximized consump ion c( ) a gi en in e es a es δ o each in e al.
Li e insu ance consump ion is in luenced no only by consume demand bu also by signi ican supply-side ac o s ha a ec
bo h a ailabili y and p icing. Insu ance companies ely on skilled pe sonnel and access o in o ma ion o accu a ely assess p icing,
ese e equi emen s, and po en ial in es men oppo uni ies in inancial ma ke s. Addi ionally, he p o ec ion o p ope y igh s and
e ec i e con ac en o cemen suppo insu e s’ in es men ac i i ies. These supply-side elemen s a e likely o impac he cos o li e
insu ance p oduc s and can be ep esen ed by he policy loading ac o in he Lewis model.
P ice emains a key de e minan o li e insu ance consump ion, and omi ing i could lead o bias in empi ical analysis. To
add ess his, we conside p ice as a unc ion o se e al supply- ela ed ac o s. Va ia ions in u baniza ion, mone a y s abili y,
ins i u ional quali y, poli ical s abili y, and banking sec o de elopmen all in luence insu e s’ capaci y o o e cos -e icien insu ance
p oduc s.
2.2. Empi ical s udies
Nume ous s udies ha e empi ically examined he impac o economic de e minan s on li e insu ance consump ion. Mos o
hese wo ks analyze bo h demand- and supply-side ac o s join ly. Fo une (1973), o example, applied a demand-side model and
concluded ha li e insu ance demand is d i en p ima ily by income, weal h, and he discoun a e. Headen and Lee (1974) conside ed
h ee g oups o explana o y a iables: (i) ac o s encou aging consump ion h ough insu e s’ e o s, such as ad e ising expendi u e
and he in oduc ion o new policies and p oduc s; (ii) household-le el sa ing de e minan s, including disposable income, expec a ions
ega ding u u e expenses, e ili y, and ma iage a es; and (iii) ac o s e lec ing he policyholde ’s abili y o pay, such as ne
household sa ings, inancial ins umen s, and expec ed u u e consump ion. Thei indings sugges ha demand o li e insu ance is
ela i ely un esponsi e o he i s se o a iables, while he second and hi d se s— ep esen ing consume beha io and inancial
capaci y—exe a posi i e e ec .
Li e al. (2007), applied he Gene alized Me hod o Momen s (GMM) o 30 OECD coun ies, inding s ong posi i e income
elas ici y o demand along wi h signi ican posi i e e ec s om inancial de elopmen , compe i ion, low in la ion, and low in e es
a es. Simila ly, Ibiwoye e al. (2010) analyzed Nige ia om 1970 o 2005 using Johansen coin eg a ion, showing ha eal GDP had a
s ong posi i e e ec on li e insu ance demand, while in e es a es had a nega i e impac ; o he a iables we e no s a is ically
signi ican .
Roman (2011), ocusing on E hiopia, epo ed ha income posi i ely in luences demand in he sho un bu shows an in e se
ela ionship in he long un. Insu ance p ice, eal in e es a e, and g oss domes ic sa ing pe capi a we e ound o ha e signi ican
nega i e e ec s.
Kjose ski (2012), in a s udy o Cen al and Sou heas Eu ope, highligh ed GDP pe capi a, in la ion, heal h expendi u e,
educa ion, and ule o law as obus p edic o s, whe eas in e es a es, dependency a ios, co up ion, and go e nmen e ec i eness
we e no consis en ly linked o li e insu ance demand.
Am o (2014), also analyzing E hiopia, ound GDP pe capi a, educa ion, li e expec ancy, and eal in e es a es o be posi i ely
associa ed wi h insu ance demand, while in la ion had a nega i e e ec .
In Ghana, E ic and Had a (2015) applied logis ic eg ession o household-le el da a and showed ha income, educa ion,
numbe o dependen s, and posi i e pe cep ions o insu e s inc eased he likelihood o pu chasing li e insu ance, whe eas age had a
nega i e e ec .
Elisa e al. (2015), using I alian household da a, ound ha in la ion and in e es a es educed demand, while income and he old-age
dependency a io boos ed i . They u he emphasized he ole o inancial ma ke pa icipa ion in enhancing insu ance demand.
Abeneze (2017) conduc ed su ey-based esea ch in E hiopia and ound income, age, educa ion, and heal h s a us o be signi ican
de e minan s, wi h income exe ing he s onges e ec .
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6621
Ganesh (2018), ocusing on India, applied ANOVA, co ela ion, and eg ession echniques o show ha gende , ma i al s a us,
occupa ion, educa ion, awa eness, amily size, and income signi ican ly a ec insu ance pu chases, while eligion and age did no .
Tiges (2018), also on E hiopia, using OLS eg ession o e a 30-yea pe iod, ound ha educa ion, income pe capi a, and inancial
deepening signi ican ly inc eased demand, while li e expec ancy, p emiums, and in la ion nega i ely in luenced i ; eligion and age
had no measu able e ec .
Mo e ecen s udies ha e con inued o expand he empi ical e idence on he de e minan s o li e insu ance demand,
inco po a ing bo h adi ional mac oeconomic ac o s and new beha io al dimensions. D ago a e al. (2023), analyzing 29 OECD
coun ies be ween 2005 and 2017, ound ha GDP pe capi a, in es men eedom, and o eign di ec in es men signi ican ly and
posi i ely in luenced li e insu ance demand. Thei esul s unde sco e he g owing impo ance o ins i u ional and inancial openness
in shaping insu ance consump ion pa e ns ac oss ad anced economies. Simila ly, S ini asan and Mi a (2024), using da a om 30
OECD coun ies o e he pe iod 1996–2020, epo ed ha highe income, u baniza ion, and educa ion le els we e obus p edic o s
o inc eased li e insu ance pene a ion, sugges ing ha socio-economic mode niza ion consis en ly d i es ma ke de elopmen .
In eme ging economies, indings a e mo e mixed. A panel s udy on BRICS coun ies e ealed ha income and inancial
eedom posi i ely in luence insu ance demand, while unemploymen exe s a nega i e e ec ; addi ionally, in la ion and in e es a es
we e shown o signi ican ly shape li e insu ance pene a ion (Ou e ille, 2023). Hasan e al. (2025), ocusing on Bangladesh, p o ided
a con as ing pe spec i e, demons a ing ha GDP pe capi a has a signi ican nega i e impac on bo h insu ance densi y and
pene a ion. By con as , in la ion and heal hca e expendi u e we e ound o encou age li e insu ance consump ion, while educa ion
spending did no exe a measu able e ec . These esul s highligh how s uc u al and ins i u ional di e ences ac oss coun ies can
p oduce di e gen income-demand ela ionships.
Beyond adi ional economic de e minan s, beha io al pe spec i es ha e also gained ac ion in ecen li e a u e. A su ey-
based s udy in Uganda by Nami embe e al. (2023) employed s uc u al equa ion modeling o show ha heu is ic and p ospec biases
s ongly in luence household insu ance decisions, illus a ing ha demand is no always guided by a ional economic easoning.
Complemen a y e idence om expe imen al se ings u he con i ms his end: a game-based analysis by Liu (2023) demons a ed
ha loss aming signi ican ly shapes li e insu ance pu chasing decisions, pa icula ly among younge popula ions, al hough indi idual
isk a i udes hemsel es we e no di ec ly associa ed wi h demand. In a ela ed heo e ical con ibu ion, Xu and Zhang (2024)
explo ed he concep o money illusion, showing ha young adul s’ ocus on nominal a he han eal alues ends o inc ease li e
insu ance up ake, while simul aneously educing annui y demand among e i ees.
Finally, echnological and demog aphic shi s a e inc easingly being ecognized as c i ical ac o s in shaping demand. In China,
whe e apid popula ion ageing coincides wi h ad ances in digi al insu ance pla o ms and a i icial in elligence, ecen e idence
sugges s ha demog aphic p essu es combined wi h pe sonalized p oduc o e ings a e os e ing a new wa e o li e insu ance
demand, despi e challenges posed by low in e es a e en i onmen s (Financial Times, 2025). Taken oge he , hese ecen s udies
con i m he endu ing ele ance o economic undamen als such as income, in la ion, and inancial de elopmen , while also
highligh ing he g owing oles o beha io al biases, ins i u ional quali y, and echnological change in explaining c oss-coun y a ia ion
in li e insu ance consump ion.
3. DATA SOURCES AND VARIABLES
The ollowing sec ions ou line a iables associa ed wi h he demand unc ion as desc ibed by Lewis (1989), alongside supply-
side ac o s ha may se e as p oxies o he policy loading ac o . While he Lewis amewo k p ima ily emphasizes he mo ali y isk
aspec o li e insu ance, we ex end he analysis o include he sa ings and annui y componen s. Fu he mo e, conside ing he po olio
aspec o li e insu ance as pa o p i a e sa ings adds an addi ional laye o he discussion.
In ou s udy, we used a ime se ies da ase o Macedonia co e ing he pe iod 2003–2024. The da a we e ob ained om he
Agency o Supe ision o and he Wo ld Bank. The da ase is unbalanced, which, acco ding o Rinaldi and Sanchis-A ellano (2006),
allows o he inclusion o mo e obse a ions and educes dependence on a pa icula subpe iod, he eby imp o ing he obus ness
o he esul s. The selec ion o a iables was guided bo h by he a ailabili y o da a and by he de e minan s mos equen ly employed
in he li e a u e on li e insu ance demand, ensu ing consis ency wi h es ablished empi ical app oaches
Acco ding o he s udies (Beck, 2002) demand o li e insu ance can be measu ed in se e al ways: “li e insu ance pene a ion”
( a io o insu ance p emiums olume o GDP), “li e insu ance densi y” (insu ance p emiums pe capi a in cons an dolla s), a io o
“li e insu ance in o ce o GDP” and “li e insu ance in p i a e sa ings” ( ela ing insu ance p emiums o p i a e sa ings ins ead o
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6622
income). Li e insu ance densi y is mo e applicable o c oss-coun y analysis as he e adjus ing o income le el o he economy is no
needed. Lin and G ace (2006) also sugges aking he “ne amoun a isk” – di e ence be ween he policy ese e and he ace amoun
( he sum ha company pays o he bene icia y when he insu ed dies), as a p oxy o he demand o li e insu ance. The dependen
a iable in my model li e insu ance pene a ion. Pene a ion is commonly used as a key indica o o li e insu ance de elopmen
because i cap u es he ela i e impo ance o li e insu ance in an economy, measu ing o al li e p emiums as a sha e o GDP (Beck &
Webb, 2003). I e lec s bo h he accessibili y o insu ance p oduc s and households’ willingness o abili y o alloca e esou ces owa d
long- e m isk co e age (Li, Moshi ian, Nguyen, & Wee, 2007). Highe pene a ion indica es ha li e insu ance plays a mo e signi ican
ole in inancial in e media ion, long- e m sa ings, and isk managemen , which a e c i ical componen s o economic and social
de elopmen (Alhassan & Biekpe, 2016; Feyen, Les e , & Rocha, 2011).
Based on heo e ical and empi ical s udies, we iden i y he ollowing ac o s.
Demog aphic De e minan s
A highe young dependency a io— he p opo ion o young dependen s ela i e o he wo king-age popula ion—is expec ed
o inc ease he demand o mo ali y co e age while educing demand o sa ings h ough li e insu ance and annui ies. When a la ge
sha e o he popula ion consis s o dependen s, he o al p esen alue o consump ion o he bene icia ies ises, which inc eases he
need o li e insu ance ha p o ides paymen s in he e en o he p ema u e dea h o he p ima y income ea ne ( e lec ed as a highe
TC in equa ion 1). A he same ime, a high young dependency a io implies ha many indi iduals a e oo young o sa e o e i emen ,
esul ing in lowe demand o li e insu ance as a sa ings ehicle. Empi ical s udies by Beens ock e al. (1986), B owne and Kim (1993),
and T ue and T ue (1990) indica e a posi i e co ela ion be ween he young dependency a io and li e insu ance pene a ion.
Howe e , gi en he opposing e ec s on mo ali y and sa ings componen s, he o e all ela ionship is ambiguous.
Simila ly, a highe old dependency a io— he p opo ion o elde ly dependen s ela i e o he wo king-age popula ion—is
expec ed o inc ease he demand o he sa ings and annui y componen s o li e insu ance while educing demand o mo ali y
co e age. In coun ies wi h a la ge e i ed popula ion, li e insu ance p oduc s ha p o ide e i emen sa ings and p o ec agains
ou li ing one’s income become mo e impo an , whe eas p o ec ion agains he dea h o he p ima y wage ea ne becomes less
c i ical. The e o e, he ne e ec o he old dependency a io on li e insu ance demand is also ambiguous.
Longe li e expec ancy is likely associa ed wi h lowe mo ali y co e age cos s and a educed pe cei ed need o dea h
p o ec ion bu a highe demand o sa ings and annui y componen s o li e insu ance ( e lec ed by a lowe ppp in equa ion 1).
Consequen ly, li e expec ancy may ha e an ambiguous impac on o e all li e insu ance demand. P e ious s udies ha e gene ally ound
a posi i e co ela ion be ween li e expec ancy and li e insu ance pene a ion (Beens ock e al., 1986; Ou e ille, 1996).
A highe concen a ion o popula ion acili a es bo h he dis ibu ion o li e insu ance p oduc s and he dissemina ion o
in o ma ion abou hem o po en ial consume s, as i lowe s cos s o companies. Consequen ly, coun ies wi h a la ge sha e o u ban
popula ion a e gene ally expec ed o exhibi highe demand o li e insu ance. Al hough amily ies and solida i y end o be s onge
in u al a eas, mos li e insu ance con ac s a e concluded in u ban cen e s. This is la gely due o he highe concen a ion o
consume s, which simpli ies he wo k o sales ep esen a i es and educes hei ope a ing expenses (Schlag, 2003). Lowe ope a ional
cos s and p ices posi i ely in luence he supply side o he li e insu ance ma ke . Howe e , empi ical s udies ha e e ealed ha he
e ec o u baniza ion on li e insu ance demand is ambiguous.
Economic De e minan s
Li e insu ance consump ion esponds posi i ely o inc eases in disposable income (Beck and Webb, 2003). Speci ically, hese
au ho s explain ha i li e insu ance is iewed as co e age in he e en o he policyholde ’s dea h, hen wi h highe disposable income,
he insu ed indi idual and hei income-dependen amily membe s a e able o spend mo e, which inc eases he demand o li e
insu ance. This demand a ises om he need o p o ec income po en ial and expec ed consump ion o dependen s in case o he
insu ed’s dea h. Fu he mo e, when li e insu ance is conside ed as a sa ings ins umen , indi iduals wi h highe ea nings ha e g ea e
capaci y o pu chase li e insu ance policies as a means o sa ing and secu ing addi ional income du ing e i emen . In his s udy, eal
g oss domes ic p oduc (GDP) will be used as a p oxy o disposable income. Mos empi ical s udies employ GDP pe capi a as a a iable
ep esen ing disposable income (Beens ock e al., 1986; T ue and T ue , 1990; B owne and Kim, 1993; Ou e ille, 1996; Beck and
Webb, 2003). Real GDP has also been used as a a iable in he empi ical li e a u e, o example in Ibiwoye e al. (2010). In his pape
we used loga i hm o GDPPC in (cu en US$)
Theo y sugges s ha in la ion has a nega i e e ec on li e insu ance consump ion. The nega i e ela ionship be ween
in la ion and li e insu ance consump ion a ises om he ac ha sa ings in he o m o li e insu ance p o ide mone a y bene i s in
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6623
he long un. When in la ion is p esen in he economy, he alue o li e insu ance dec eases, since highe in la ion educes he e u ns
ha policyholde s ecei e om hei in es men s in li e insu ance companies (Ca gill and T oxel, 1979). The measu e o in la ion used
in his s udy is based on he annual pe cen age inc ease in he Consume P ice Index (CPI).
The de elopmen o he inancial sys em enables indi iduals o con e hei a ailable cash in o inancial ins umen s, such
as deposi s and secu i ies, in o de o secu e u u e income (Li e al., 2007). The heo y does no p o ide a clea answe ega ding he
impac o inancial sec o de elopmen on li e insu ance consump ion. Speci ically, li e insu ance companies o e p oduc s ha a e
pa o he inancial ma ke , and i hese p oduc s a e conside ed om a sa ings pe spec i e, inancial sec o de elopmen is expec ed
o inc ease li e insu ance consump ion.
Mo eo e , a well-de eloped inancial and banking sec o con ibu es o lowe ope a ing cos s o li e insu ance companies,
which can lead o lowe p oduc p ices and highe consump ion. Fo example, a de eloped banking sec o allows insu ance companies
o conduc ansac ions a lowe cos s, enabling hem o o e mo e compe i i ely p iced p oduc s.
On he o he hand, since indi iduals ha e limi ed income, i a signi ican po ion is in es ed, o example, in deposi s, inancial
sec o de elopmen may no necessa ily inc ease li e insu ance consump ion. Some empi ical s udies, such as Li e al. (2007) and
Ou e ille (1996), use he M2 money agg ega e ela i e o g oss domes ic p oduc as a measu e o inancial sec o de elopmen . Beck
and Webb (2003), howe e , ocus only on banking sec o de elopmen as he dominan pa o he inancial sec o , measu ed by
claims o he non- inancial sec o on deposi banks.
In his s udy, we will ollow Kjose ski (2012) and as a measu e o inancial sec o de elopmen we used Domes ic c edi o
p i a e sec o by banks (% o GDP)
Simila o he p e ious a iable, heo y does no p o ide a clea answe ega ding he impac o weal h on li e insu ance
consump ion. Acco ding o Beens ock e al. (1986), weal h may ha e a nega i e e ec on li e insu ance consump ion, as weal hy
indi iduals who own houses, s ocks, bonds, gold, and simila asse s a e less likely o need li e insu ance, since hei u u e income is
ela i ely secu ed in he e en o hei dea h. In con as , poo e indi iduals ely p ima ily on hei income and he e o e ha e a g ea e
need o li e insu ance, as hey mus pu chase policies o p o ec hei dependen s in case o hei dea h. The explana ion o a posi i e
impac o weal h on li e insu ance consump ion is simila o ha o income discussed abo e. As a p oxy measu e o weal h, his s udy
will use he eal es a e p ice index, since eal es a e is pa o weal h, and ising p ope y p ices inc ease household weal h. Beens ock
e al. (1986) and Beck and Webb (2003) use he Gini index, which e lec s he inequali y o income dis ibu ion among indi iduals in a
coun y, as a measu e o weal h.
Ins i ucional a iables (Poli ical S abili y, Rule o Law, and Co up ion Con ol)
The ins i u ional and poli ical en i onmen plays a c ucial ole in shaping he de elopmen o he li e insu ance ma ke . As
Wa d and Zu b uegg (2002) emphasize, poli ical and legal s abili y a e essen ial ac o s ha should be inco po a ed in o empi ical
models o insu ance demand. Since li e insu ance in ol es long- e m con ac ual ela ionships be ween consume s and insu e s, he
s abili y o he legal and poli ical sys em enhances he willingness o bo h pa ies o engage in such commi men s. In his espec ,
indica o s such as Poli ical S abili y (cap u ing he poli ical dimension), Rule o Law, and Con ol o Co up ion ( e lec ing he legal and
ins i u ional dimension), as p o ided by he Wo ld Bank, a e pa icula ly ele an .
A well- unc ioning ins i u ional amewo k suppo s he g ow h o he li e insu ance sec o in se e al ways. Fi s , i audulen
p ac ices in claims epo ing a e widesp ead, insu ance becomes p ohibi i ely cos ly o la ge segmen s o he popula ion. Second,
when b eaches o insu ance con ac s canno be e ec i ely appealed h ough he legal sys em, he pe cei ed alue o hese con ac s
o consume s declines, discou aging long- e m inancial commi men s. Thi d, weak p o ec ion o p ope y igh s and ine ec i e
en o cemen o con ac s cons ain insu e s’ abili y o in es e icien ly and inc ease he cos o p o iding insu ance. Finally, poli ical
ins abili y sho ens he economic ho izon o bo h po en ial policyholde s and insu e s, he eby hinde ing he de elopmen o a
sus ainable li e insu ance ma ke .
To cap u e hese ins i u ional and poli ical dimensions, h ee indica o s a e commonly employed. Rule o Law e lec s he
ex en o which ci izens a e able o ely on he legal sys em o media e dispu es and en o ce con ac s. Poli ical S abili y cap u es he
likelihood o go e nmen des abiliza ion h ough e en s such as e olu ions and coups. Con ol o Co up ion measu es he deg ee o
which public powe is exe cised o p i a e gain, which di ec ly a ec s he c edibili y and unc ioning o ins i u ions. Toge he , hese
indica o s p o ide a comp ehensi e measu e o he ins i u ional en i onmen ele an o he li e insu ance sec o .
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6624
Table 1. De ini ion o a iables
Symbol
Va iable
Measu emen
Sou ce
PEN
Li e insu ance pene a ion
Ra io o li e insu ance p emiums o GDP (%)
Agency o supe ision o insu ance in
Macedonia
CDR
Young dependency a io
Popula ion aged 0–14 ela i e o wo king-age
popula ion (15–64), %
Wo ld Bank (Wo ld De elopmen Indica o s
– WDI)
ODR
Old-age dependency a io
Popula ion aged 65+ ela i e o wo king-age
popula ion (15–64), %
Wo ld Bank (WDI)
LEXP
Li e expec ancy
A e age numbe o yea s a newbo n is expec ed
o li e
Wo ld Bank (WDI)
URB
U baniza ion
Sha e o u ban popula ion in o al popula ion (%)
Wo ld Bank (WDI)
GDPPC
GDP pe capi a
GDP pe capi a (cu en US$)
Wo ld Bank (WDI)
GINI
Income inequali y (Gini
index)
Gini index (0 = pe ec equali y; 100 = pe ec
inequali y)
Wo ld Bank (WDI); Wo ld Income Inequali y
Da abase (WIID)
INF
In la ion
Consume p ice index (annual %, yea -on-yea )
Wo ld Bank (WDI)
DCPS
Domes ic c edi o p i a e
sec o
C edi o p i a e sec o by banks as % o GDP
Wo ld Bank (WDI, Global Financial
De elopmen Da abase)
LAW
Rule o law
Go e nance indica o (−2.5 = weak; +2.5 = s ong)
Wo ld Go e nance Indica o s (Wo ld Bank)
PS
Poli ical s abili y
Go e nance indica o (−2.5 = weak; +2.5 = s ong)
Wo ld Go e nance Indica o s (Wo ld Bank)
COR
Con ol o co up ion
Go e nance indica o (−2.5 = weak; +2.5 = s ong)
Wo ld Go e nance Indica o s (Wo ld Bank)
We also p esen desc ip i e s a is ics o all coun ies and discuss he main ends in he e olu ion o he selec ed a iables
o e ime. The summa y s a is ics p esen ed in Table 2 e eal some no ewo hy a ia ions among he coun ies in ou sample.
Table 2. Desc ip i e s a is ics
S a is ic
PEN
CDR
ODR
LEXP
URB
GDPPC
INF
GINI
DCPS
LAW
PS
COR
Mean
0.16
27.72
19.85
74.85
57.84
5580.80
2.72
37.98
43.15
-0.27
-0.34
-0.31
Median
0.16
26.81
18.43
74.95
57.55
5507.68
1.48
38.25
47.75
-0.29
-0.30
-0.33
Maximum
0.30
32.75
27.66
76.65
59.87
9310.03
14.20
46.10
53.68
-0.06
0.26
-0.03
Minimum
0.04
25.68
15.27
73.24
57.09
2445.36
-0.74
33.00
17.56
-0.54
-1.16
-0.65
S d. De .
0.10
2.22
4.13
1.03
0.82
1790.35
3.64
3.72
11.01
0.13
0.39
0.18
Obse a ions
22
22
22
22
22
22
22
16
22
21
21
21
Sou ce: Own calcula ions
Li e insu ance pene a ion (PEN) shows a mean o 0.16% o GDP, wi h alues anging be ween 0.04% and 0.30%. This low
a e age sugges s ha he li e insu ance sec o emains unde de eloped, wi h limi ed con ibu ion o inancial deepening.
The demog aphic indica o s e eal expec ed pa e ns. The young dependency a io (CDR) a e ages 27.72, while he old-age
dependency a io (ODR) s ands a 19.85, e lec ing a ela i ely la ge sha e o young dependen s compa ed o elde ly dependen s.
The s anda d de ia ions indica e mode a e a iabili y, pa icula ly o ODR (4.13), which sugges s g adual bu no able aging dynamics
in he popula ion s uc u e.
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6625
Li e expec ancy (LEXP) is s able ac oss he sample, wi h a mean o 74.85 yea s and a low s anda d de ia ion (1.03), con i ming
imp o emen s in heal h ou comes bu wi h limi ed a ia ion o e ime. U baniza ion (URB) is also ela i ely s able, a e aging 57.84%
o he popula ion, wi h only a na ow ange be ween 57.09% and 59.87%.
Economic indica o s show mo e a iabili y. GDP pe capi a (GDPPC) a e ages USD 5,580, anging widely om USD 2,445 o
USD 9,310, e lec ing income g ow h and s uc u al ans o ma ions. In la ion (INF) displays g ea e dispe sion, wi h a mean o 2.72%,
bu a wide ange om –0.74% (de la ion) o 14.20%, and a ela i ely high s anda d de ia ion (3.64). This indica es pe iods o signi ican
p ice ins abili y.
Inequali y, measu ed by he Gini index, a e ages 37.98, wi h mode a e a ia ion ac oss he sample (SD 3.72), sugges ing
ela i ely pe sis en bu luc ua ing le els o income inequali y. Domes ic c edi o he p i a e sec o (DCPS) a e ages 43.15% o GDP,
wi h a no able dispe sion (SD 11.01), implying ha inancial deepening has been une en.
Ins i u ional quali y indica o s— ule o law (LAW), poli ical s abili y (PS), and con ol o co up ion (COR)—all ha e nega i e
mean alues (–0.27, –0.34, and –0.31, espec i ely), sugges ing ins i u ional weaknesses ela i e o global s anda ds. Thei ela i ely
low s anda d de ia ions imply ha ins i u ional quali y has emained consis en ly weak, hough wi h mino luc ua ions.
O e all, he desc ip i e s a is ics highligh h ee key pa e ns:
1. Unde de elopmen in li e insu ance ma ke s and ins i u ional amewo ks.
2. Mac oeconomic ola ili y, especially in in la ion and inancial deepening.
3. G adual demog aphic and s uc u al ans o ma ions, wi h s able li e expec ancy and u baniza ion bu ising dependency
p essu es.
4. METHODOLOGY
Since ime se ies a e used o he pu poses o he analysis in his s udy, i is i s necessa y o de e mine hei in eg a ion
p ope ies, which in ol es assessing he s a iona i y o non-s a iona i y o he selec ed a iables. Acco ding o Guja a i (2003), a se ies
can be conside ed s a iona y i i s mean and a iance emain cons an o e ime, and he co a iance be ween wo ime pe iods
depends only on he dis ance, o lag, be ween he pe iods a he han on he ac ual ime a which he a iance is calcula ed.
To iden i y hese cha ac e is ics o he ime se ies, we e applied he Augmen ed Dickey-Fulle (ADF) es . This es examine
he null hypo hesis ha he ime se ies has a uni oo , meaning ha i is non-s a iona y. To add ess non-s a iona i y, he ime se ies
mus be di e enced d imes, p oducing an in eg a ed se ies deno ed as I(d), whe e d ep esen s he o de o in eg a ion.
A 5% signi icance le el is used as he c i ical h eshold o de e mining whe he a ime se ies is s a iona y in his s udy. The
esul s om his es s o s a iona i y/non-s a iona i y o he ime se ies, p esen ed in Table 3, will se e as a basis o cons uc ing
he eg essions in he subsequen pa o he s udy.
Table 3. In eg a ion P ope ies o he Se ies
De e minan
A Le el -s a is ic
C i ical Value (5%)
Fi s Di e ence -s a is ic
C i ical Value (5%)
PEN
0.072
-3.01236
-3.628
-3.02069
CDR
-0.389
-3.01236
-2.089
-3.02069
ODR
1.504
-3.01236
-2.153
-3.02069
LEXP
0.604
-3.01236
-3.731
-3.02069
URB
0.144
-3.01236
-2.516
-3.02069
LGDPPC
2.556
-3.01236
-2.100
-3.02069
GINI
0.851
-3.01236
-3.292
-3.02069
INF
-1.635
-3.01236
-4.104
-3.02069
DCPS
0.742
-3.01236
-1.783
-3.02069
LAW
-0.799
-3.01236
-5.168
-3.02069
PS
-0.654
-3.01236
-2.035
-3.02069
Economic, Demog aphic, and Ins i u ional D i e s o Li e Insu ance Consump ion in Macedonia
JEFMS, Volume 08 Issue 10 Oc obe 2025 www.ije m.co.in Page 6626
De e minan
A Le el -s a is ic
C i ical Value (5%)
Fi s Di e ence -s a is ic
C i ical Value (5%)
COR
-1.171
-3.01236
-2.098
-3.02069
Sou ce: Own calcula ions
The esul s o he Augmen ed Dickey-Fulle (ADF) es indica e ha all analyzed a iables a e non-s a iona y a le els,
meaning ha we canno ejec he null hypo hesis o a uni oo . Howe e , a e i s di e encing, all se ies become s a iona y, which
is con i med by he -s a is ics being lowe han he c i ical alues a he 5% signi icance le el. Consequen ly, all a iables a e
in eg a ed o o de one, I(1), which cons i u es a undamen al p e equisi e o applying he ARDL model and es ing o coin eg a ion.
F om a me hodological pe spec i e in econome ic analysis, i is c ucial o selec a esea ch app oach ha co esponds o he
cha ac e is ics and speci ici ies o he da ase o ob ain economically meaning ul insigh s and esul s. In ou case, he indings om
he p e ious sec ion indica e ha he selec ed de e minan s become s a iona y a e i s di e encing, I(1). When some o he
de e minan s in he model a e non-s a iona y a le els (i.e., possess uni oo s), i is necessa y o es o he p esence o po en ial
coin eg a ion (Rinaldi and Sanchis-A ellano, 2006). The concep o coin eg a ion is ela ed o long- un equilib ium among wo o mo e
a iables. Speci ically, coin eg a ion implies he exis ence o a s a iona y linea combina ion among non-s a iona y a iables, meaning
ha despi e s ochas ic mo emen s, a s able long- e m ela ionship can be obse ed among hem (Engle and G ange , 1987).
In he economic li e a u e, he mos commonly used p ocedu es o assessing coin eg a ion among de e minan s include:
he esidual-based Engle and G ange (1987) es , he Johansen (1988; 1991) es , he Johansen and Juselius (1990) es , he Fully-
Modi ied OLS p ocedu e (FM-OLS) de eloped by Phillips and Hansen (1990), he Au o eg essi e Dis ibu ed Lag (ARDL) modeling
app oach de eloped by Pesa an and Shin (1995, e ised 1997), Pesa an and Pesa an (1997), Pesa an, Shin, and Smi h (1999), and he
Dynamic OLS (DOLS) p ocedu e de eloped by S ock and Wa son (1993).
Fo his pu pose, we employ he Fully Modi ied O dina y Leas Squa es (FMOLS) es ima o p oposed by Phillips and Hansen
(1990) and he Dynamic O dina y Leas Squa es (DOLS) es ima o de eloped by S ock and Wa son (1993). Bo h me hods a e designed
o p o ide consis en and e icien es ima es in he p esence o coin eg a ion, o e coming he limi a ions o con en ional OLS, which
is biased and inconsis en in such se ings due o endogenei y and se ial co ela ion.
The FMOLS es ima o modi ies he adi ional OLS by applying a semipa ame ic co ec ion o accoun o se ial co ela ion
in he e o e ms and he po en ial endogenei y o eg esso s (Phillips & Hansen, 1990). This is achie ed h ough adjus men s based
on he long- un co a iance ma ix o he inno a ions, which ensu es ha he es ima o is asymp o ically unbiased and no mally
dis ibu ed. The FMOLS es ima o is pa icula ly use ul when he sample size is limi ed, as i imp o es he small-sample p ope ies o
he es ima es (Phillips & Hansen, 1990; Hansen & Phillips, 1991).
The DOLS es ima o , in con as , augmen s he coin eg a ing eg ession wi h leads and lags o he di e enced explana o y
a iables, he eby explici ly con olling o endogenei y and se ial co ela ion in a pa ame ic way (S ock & Wa son, 1993). This
augmen a ion p o ides mo e lexibili y in add essing eedback e ec s be ween eg esso s and he e o e m, o en esul ing in
supe io ini e-sample pe o mance compa ed o FMOLS. Empi ical e idence sugges s ha DOLS ends o p oduce less biased and
mo e p ecise es ima es, especially in small o medium samples (Masih & Masih, 1996).
Fo mally, he DOLS eg ession can be exp essed as:
whe e 𝑦𝑡 is he dependen a iable, x x is he coin eg a ed eg esso , ∆𝑥𝑡−𝑘 a e he leads and lags o he i s -di e enced eg esso s,
and 𝑢𝑡is he e o e m. The inclusion o ∆𝑥𝑡−𝑘 ensu es ha he es ima o is obus o endogenei y, while in o ma ion c i e ia (such
as AIC o BIC) a e ypically employed o de e mine he op imal numbe o leads and lags.
In summa y, bo h FMOLS and DOLS es ima o s a e well sui ed o es ima ing coin eg a ed ime se ies models. While FMOLS
elies on a non-pa ame ic co ec ion o achie e asymp o ic e iciency, DOLS inco po a es dynamic adjus men s h ough leads and
lags, o e ing mo e lexibili y and s onge ini e-sample p ope ies. Gi en hese ad an ages, many empi ical applica ions—such as
