Revised 31

st
May 2007
Preliminary draft: Comments and suggestions are welcome
Are Life Insurance Demand Determinants valid for
Selected Asian Economies and India?
Paper for Presentation at the 11
th
Annual Meeting of APRIA, NCCU Taipei,
July 22-25 2!
"u#i r "en $r% "% Ma&hes'aran
PhD Student
Centre for Economic Studies & Policy
Institute for "o(ial an& )(ono*i( Change
Nagarabhavi, Bangalore 7,
!arnata"a# $ND$%
sens&isec'ac'in
Ph' No'# ()*)+,-.,/)0
Professor
Centre for Economic Studies & Policy
Institute for "o(ial an& )(ono*i( Change
Nagarabhavi, Bangalore 7,
!arnata"a# $ND$%
madhes&isec'ac'in
Ph' No'# ()*+/0*-,.+ 12ffice# E3tn' */4
JEL Classification: G22, C33, C32,
Keywords: ife !nsurance, !nsurance "emand, Panel #stimation, $ime %eries &nalysis'
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
Abstract
"uring t)e *ost 1++0 *eriod, services sector in most of t)e &sian economies witnessed
growt) fueled ,y su,stantial c)anges in t)e financial sector of t)ese economies' $)e
insurance industry, in most of t)e &sian economies were *u,licly owned and remained
isolated from domestic *rivate or foreign *artici*ation' -ut, regulatory reforms and
*olicy c)anges in t)e &%#&. economies in t)e *ost financial crisis *eriod, li,erali/ation
*rocess in some of t)e %&&RC countries, C)ina0s accession to 1$2 and creation of
3ong 4ong %&R )ad led to *)enomenal c)anges in t)e growt) *attern of insurance
industry in t)ese economies' $)is study is focused on 5 %&&RC countries, 2 countries
from Greater C)ina Region and 6 &%#&. countries and is an attem*t to searc) for
crucial economic and ot)er socio7*olitical varia,les w)ic) can *lay a significant role in
e8*laining t)e life insurance consum*tion *attern in t)ese economies for t)e 11 year
*eriod 91++572005: studied' &n inde*endent e8ercise is underta;en for re7assessing
varia,les ,est e8*laining consum*tion *attern in !ndia for t)e *eriod 1+6< to 2005' 1e
find some of t)e varia,les ca*a,le of strongly determining life insurance demand ,ut
contradictions to earlier studies arise as some of t)e determinants are o,served to ,e
eit)er wea; or due to t)e met)odology em*loyed, turn out to ,e incom*etent in
e8*laining demand'
2
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
Introduction
$)e growt) of t)e services sector in t)e &sian economies led to su,stantial c)anges in t)e
financial sector' $)e &sian =inancial Crisis, affecting t)e &%#&. economies in
*articular resorted to more regulatory measures to en)ance delivery of *roducts wit)
minimal ris;s and failures' $)e countries surrounding t)e &%#&. economies also went
t)roug) a *)ase of economic7restructuring, t)e most nota,le event ,eing t)e im*act of
C)ina0s accession to 1$2' $)e insurance industry in most of t)e &sian economies were
*u,licly owned and o*erated' Government mono*oly ;e*t t)is segment of t)e financial
mar;et isolated from domestic *rivate or foreign *artici*ation' -arring few e8ce*tions,
t)e insurance mar;et on an average remained underdevelo*ed in terms of insurance
density and *enetration' Regulatory c)anges since mid eig)ties and o*ening of t)ese
mar;ets to *rivate and foreign entry )ave ,een luring glo,al )eavyweig)t insurers to
enter t)ese economies' &s more su**liers enter t)ese mar;ets, t)e issue is to re7e8amine
t)e factors t)at can *ro,a,ly elevate demand for insurance *roducts' $)is study on 5
%&&RC countries, 2 countries from Greater C)ina Region and 6 &%#&. countries, deals
wit) t)is *articular issue' &fter reviewing e8isting t)eoretical as well as em*irical
literature, we identify varia,les determining demand and categori/e t)em as economic,
demogra*)ic, legal and socio7*olitical varia,les' >sing a time series framewor; we t)en
ta;e into account a single economy, !ndia, and reassess t)e issues and )y*ot)esis dealt in
t)e *anel analysis'
$)e *a*er is organi/ed as follows' $)e *a*er ,egins wit) t)e t)eoretical as well as
em*irical review of literature' -ased on t)e t)eoretical review as well as t)e results
o,tained from earlier researc) studies, we frame our set of researc)a,le ?uestions and
issues' 1e *resent our identified *referred set of determinants and a *riori e8*ected
relations)i* wit) t)e demand *ro8y' $)e met)odology *ortion is divided into e8*laining
,riefly t)e 12 selected &sian economies and !ndia followed ,y data sources' .e8t, we
s*ecify t)e two se*arate models to ,e estimated via use of two different econometric
tec)ni?ues vi/' *anel data analysis and time series analysis' Proceeding to t)e results and
inter*retation, we *resent t)e estimation results relating to t)e *anel and t)e time series
data' astly, we summarise t)e findings and draw conclusions'
3
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
Review of Literature
(a) Theoretical Studies
%tudies on life insurance consum*tion dates ,ac; to 3eu,ner 91+52: w)o *ostulated t)at
)uman life value )as certain ?ualitative as*ects t)at gives rise to its economic value' -ut
)is idea was normative in nature as it suggested @)ow muc)0 of life insurance to ,e
*urc)ased and not @w)at0 will ,e *urc)ased' $)ere were no guidelines regarding t)e ;ind
of life *olicies to ,e selected de*ending u*on t)e consumers ca*acity and t)e amount of
ris; to ,e carried in t)e *roduct'
#conomic value Audgments are made on ,ot) t)e normative as well as *ositive issues'
ater studies
1
on insurance gradually incor*orated t)ese issues via assimilating
develo*ments in t)e field of ris; and uncertainty following wor;s ,y von .eumann and
Morgenstern 91+57:, &rrow 91+<3:, "e,reu 91+<3: and ot)ers' $)e economics on
insurance demand ,ecame more focused on evaluating t)e amount of ris; to ,e s)ared
,etween t)e insured and t)e insurer rat)er t)an evaluation of life or *ro*erty values' $)is
emerged ,ecause it was ris; associated wit) individual life or *ro*erty t)at called for an
economic valuation of t)e cost of *roviding insurance'
ife insurance is essentially a form of saving, com*eting wit) ot)er forms of saving in
t)e mar;et' $)e t)eory of life insurance demand t)us develo*ed t)roug) t)e life7cycle
model9s: of saving' et a *erson0s income rate and )is consum*tion *lan are re*resented
,y a continuous function of time ( ) y t
and ( ) c t
res*ectively' $)us, net saving 9*ositive
or negative: at time t is given ,y
2
( ) ( ) ( ) { }
0
t
t t
s t e e y s c s ds

=

777 91:
w)ere, is t)e rate of interest'
1
Baari 91+63:, Mossin 91+6C:, 3a;ansson 91+6+:, =is)er 91+73:, -orc) 91+77:, Pissarides 91+C0:,
Cam*,ell 91+C0:, 4arni and Dilc)a 91+C<, 1+C6:, ewis 91+C+:, -ern)eim 91++1: to mention few'
2
$)is section is ,ased on C)a*ter 5 of -orc)91++0:
5
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
#ven t)oug) all individuals want to consume as muc) *ossi,le, t)is does not )a**en in
reality and t)e a,ove e8*ression of saving is constrained ,y num,er of *ossi,ilities' =or
e8am*le, t)e *erson )as no de,t E ( ) 0 s t
F, solvent at time of )is deat) at time T E
( ) 0 s T
F and leaves a ,e?uest of amount B E ( ) s T B
F'
>ncertainty associated wit) time of deat) is re*resented ,y a random varia,le wit)
density ( ) ( ) f t t =
, w)ere ( ) t
is t)e *ro,a,ility t)at a *erson at age x s)all still ,e
alive after a time t' $a;ing e8*ectation of 91: and discontinuing, we get
( ) ( ) ( ) ( ) { } ( )
0 0 0
t
t s
e s t t dt y s c s e ds t dt





=



777 92:
( ) ( ) ( ) { }
0
s
s e y s c s ds

777 93:
!f t)e individual ta;es pure endowment insurance, single *remium ,eing
( ) ( )
0
t
t
t x x s
E t e E ds

+

= = +

,
$)us, net saving e?uals ( ) ( ) ( )
0 0
t
x s
E ds y t c t dt

+

+




'
$)is s)ows t)at saving t)roug) life insurance ta;es *lace at a )ig)er rate of interest t)an
conventional saving' !n t)e determination of o*timal insurance consum*tion, t)e
conventional utilitarian t)eory is ado*ted w)ic) reflects individuals *references over
different consum*tion *atterns' et us consider t)e utility function of t)e form
( ) ( )
0
T
t
v c e u c t dt

=

777 95:
w)ere, is t)e @im*atience0 to consume'
&ssuming t)e *erson )ad no de,t at time T i'e', ( ) 0 s T =
, t)e *ro,lem is to ma8imi/e its
utility 95:' $)e solution gives us, ( )
( ) t
v c t ke

=

' 1it)in t)is framewor;, various
forms of life insurance are introducedG t)e *ro,a,ility associated wit) deat) or num,er of
ma8imum life years is consideredG and t)en t)e e8*ected utility of consum*tion is
<
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
ma8imi/ed su,Aect to any one of t)e t)ree restrictions on t)e net savings as descri,ed
a,ove'
$)e role of insurance in t)e a,ove model )as ,een *redominantly to smoot)en out
consum*tion over time, ma;e ,e?uests, and re*ay de,ts or to insure a constant income
stream after retirement' $)e ongoing discussion also reveals t)at individuals0 current
income and future antici*ated consum*tion e8*enditure *lays a crucial role in
determining t)e amount of insurance *urc)ased 9we are, for a w)ile ignoring t)e form in
w)ic) insurance is *urc)ased:' $)e im*ortance of rate of interest 9: or t)e im*atience
factor 9: is also wort) considering' Preferences over different consum*tion *attern vary
from *erson to *erson and t)ere are @?ualitative0 factors w)ic) affects suc) *references'
>sing t)e e8*ected utility framewor; in a continuous time model, Baari 91+6<: studied
t)e *ro,lem of uncertain lifetime and life insurance' !ncluding t)e ris; of dying in t)e life
cycle model, )e s)owed conce*tually t)at an individual increases e8*ected lifetime utility
,y *urc)asing fair
3
life insurance and fair annuities' %im*le models of insurance demand
were *ro*osed ,y Pratt 91+65:, Mossin 91+6+:, %mit) 91+6C: and ot)ersG considering a
ris; averse decision ma;er wit) an initial wealt) W' $)e results indicate t)at demand for
life insurance varies inversely wit) t)e wealt) of t)e individuals' 3a;ansson 91+6+: used
a discrete7time model of demand for financial assets and life insurance *urc)ase in
*articular to e8amine ,e?uest motive in considera,le detail' Pissarides
5
91+C0: furt)er
e8tending Baari0s wor; *roved t)at life insurance was t)eoretically ca*a,le of a,sor,ing
all fluctuations in lifetime income' ewis 91+C+: found out t)at t)e num,er of de*endents
as an influence on t)e demand for life insurance'
$o sum u*, t)e t)eoretical review yields macroeconomic varia,les li;e income, rate of
interest, and accumulated savings in wealt) formG along wit) a set of demogra*)ic or
social varia,les )aving *otential im*act on an individuals0 decision to o*t for or not to
demand life insurance' ife insurance consum*tion increases wit) t)e ,readwinner0s
3
@=air0 actuarially means t)e *remium or *rice of insurance t)at e8actly e?uals e8*ected value of *ayments
in case of claims or loss ,y t)e insurer, wit)out c)arges for e8*enses or *rofits'
5
Refer %;i**er 92003: *age 20'
6
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
*ro,a,ility of deat), t)e *resent level of family0s consum*tion and t)e degree of ris;
aversion'
(b) Empirical Studies
!n t)is section, we e8*lore some of t)e *ioneering em*irical e8ercises on determinants of
life insurance' Most of t)ese studies )ad focused on ,ot) t)e demand side factors and t)e
su**ly side factors'
3eaden and ee 91+75: studied t)e effects of s)ortrun financial mar;et ,e)aviour and
consumer e8*ectations on *urc)ase of ordinary life insurance and develo*ed structural
determinants of life insurance demand' $)ey considered t)ree different sets of varia,les:
first, varia,les stimulating demand as a result of insurer efforts 9e'g' industry advertising
e8*enditure, si/e of t)e sales force, new *roducts and *olicies, etc':G second, varia,les
affecting )ouse)old saving decision 9e'g' dis*osa,le, *ermanent and transitory income,
e8*enditure e8*ectation, num,er of ,irt)s, marriages, etc': and lastly, varia,les
determining a,ility to *ay and si/e of *otential mar;ets 9e'g' net savings ,y )ouse)olds,
financial assets, and consumer e8*ectation regarding future economic condition:' $)ey
concluded t)at life insurance demand is inelastic and *ositively affected ,y c)ange in
consumer sentimentsG interest rates *laying a role in t)e s)ortrun as well as in t)e
longrun'
>sing an international dataset 912 countries over a *eriod of 12 years: to e8amine t)e
relations)i* ,etween *ro*erty lia,ility insurance *remiums and income, -eenstoc; et al'
91+CC: found out t)at marginal *ro*ensity to insure i'e', increase in insurance s*ending
w)en income rises ,y 1H, differs from country to country and *remiums vary directly
wit) real rates of interest' &ssuming a two *eriod sim*le modelG t)ey considered t)e case
w)en wealt) W is reduced ,y G following a loss and no insurance *urc)ased in t)e first
*eriod' !f t)ere )ad ,een some insurance *urc)ased t)an wealt) in t)e first *eriod e?uals
9W7 premium paid: and assuming loss, end *eriod wealt) is 9G I sum insured:' $)us,
7
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
again t)e decision of consumer and )isJ)er initial wealt) status too are significant factors
w)en s)ortrun or longrun consum*tion of insurance is considered'
$)e study ,y $ruett et al' 91++0: discussed t)e growt) *attern of life insurance
consum*tion in Me8ico and >nited %tates in a com*arative framewor;, during t)e *eriod
1+65 to 1+C5' $)ey assumed t)at at an a,stract level demand de*ends u*on t)e price of
insurance, income level of individual, availa,ility of su,stitute and ot)er individual and
environment s*ecific c)aracteristics' =urt)er, t)ey e8*erimented wit) demogra*)ic
varia,les li;e age of individual insured9s: and *o*ulation wit)in t)e age grou* 2< to 65
and also considered education level to )ave some ,earing on insurance consum*tion
decision' $)ey concluded t)e e8istence of )ig)er income inelasticity of demand for life
insurance in Me8ico wit) low income levels' &ge, education and income were significant
factors affecting demand for life insurance in ,ot) countries'
%tarting wit) a ,rief review of ewis0s t)eoretical study and an assum*tion t)at
in)a,itants of a country are )omogeneous relative to t)ose of ot)er countries, t)e study
,y -rowne et al' 91++3: e8*anded t)e discussion on life insurance demand ,y adding
newer varia,les namely, average life e8*ectancy and enrollment ratio of t)ird level
education' $)e study ,ased on 5< countries for two se*arate time *eriods 91+C0 and
1+C7: concluded t)at income and social security e8*enditures are significant
determinants of insurance demand, )owever, inflation )as a negative correlation'
"e*endency ratio, education and life e8*ectancy were not significant ,ut incor*oration of
religion
<
, a dummy varia,le, indicates t)at Muslim countries )ave negative affinity
towards life insurance'
-ased on a cross7sectional analysis of 5< develo*ing countries, 2utreville 91++5:
analysed t)e demand for life insurance for t)e *eriod 1+C6' !n )is study )e considered
varia,les suc) as agricultural status of t)e country re*resented ,y t)e *ercentage of
agricultural la,our forceG )ealt) status of t)e country in terms of amenities li;e
<
KReligion can *rovide weig)ts into individuals L and life insurance consum*tion is less in *redominantly
!slamic countriesM studies ,y "ouglas et al' 91+C2:, 3enderson et al' 91+C7:, Deli/er 91+7+: and 1arsaw
91+C6:: cited in -rowne et al 91++3: *age 621'
C
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
*ercentage of *o*ulation wit) access to safe drin;ing waterG *ercentage of la,our force
wit) )ig)er education and t)e level of financial develo*ment as factors e8*laining
insurance demand ot)er t)an t)e varia,les we )ave discussed a,ove' $wo dummy
varia,les were used to reflect t)e e8tent of com*etition in t)e domestic mar;et and
foreign *artici*ation in t)e countries considered' $)e analysis s)ows t)at *ersonal
dis*osa,le income and level of financial develo*ment significantly relates to insurance
develo*ment' %ince t)e *olitical *)iloso*)y regarding mar;et o*enness varies from
country to country, mar;et structures dummy a**eared to ,e significant'
$a;ing into account t)e e8*ansion of t)e service sector during t)e early nineties and
growt) of insurance services in *articular, -rowne et al' 92000:, tried to e8*lain t)e
differences in *ro*erty lia,ility insurance consum*tion across countries' $)ey considered
individual0s income and wealt), degree of ris; aversion, loss *ro,a,ility and *rice of
insurance as varia,les affecting *ro*erty7lia,ility insurance demand, similar to t)ose used
for life insurance demand analysis' $)e analysis was focused on t)e 2#C" countries and
concluded t)at in general, insurance *urc)ase is influenced ,y various economic and
demogra*)ic conditions' &not)er study ,ased on nine 2#C" countries e8amined t)e
s)ort run and long run relations)i* e8)i,ited ,etween economic growt) and growt) in t)e
insurance industry' $)is study ,y 1ard et al' 92000: is a co7integration analysis using
annual data for real G"P and total real *remiums for t)e *eriod 1+61 to 1++6' Results
give an indication t)at country s*ecific factors
6
influence t)e causal relations)i* ,etween
economic growt) and insurance mar;et develo*ment'
&llowing income elasticity to vary as G"P grows for an economy, #n/ 92000: *ro*osed
t)e S7curve relation ,etween *er7ca*ita income and insurance *enetration' >sing t)is one
factor model one can generate longrun forecast for life insurance demand' 2,serving t)e
outlier countries or countries distant from t)e %7curve *lot, it is *ossi,le to identify
structural factors li;e insurance environment, ta8ation structures, etc' resulting in suc)
deviations'
6
Country s*ecific factors referred )ere to are attitudes towards ris; and ris; management, regulatory
factors, legal environment, ot)er modes or availa,ility of financial intermediation, etc'
+
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
$)ere are two detailed studies on t)e determinants of life insurance demand, one ta;ing
into consideration only t)e &sian countries and t)e ot)er ,ased on 6C economies' $)e
former study ,y 1ard et al' 92003: and t)e later ,y -ec; et al' 92003: evolves around t)e
issue of finding t)e cause ,e)ind variations in life insurance consum*tion across
countries' &fter almost t)ree decades of em*irical wor; in t)is direction, it is still )ard to
e8*lain t)e anomalous ,e)aviour of &sian countries wit) )ig)er savings rate, large and
growing *o*ulation, relatively low *rovision for *ensions or ot)er security and a sound
ca*ital mar;et ,ut com*aratively low *er7ca*ita consum*tion of insurance' #8ce*t Na*an,
most of t)e &sian countries )ave low density and *enetration figures'
$)e two main services *rovided ,y life insurance: income re*lacement for *remature
deat) and long7term savings instruments, are t)e starting *oint for -ec; et al' 92003:'
$)ey considered t)ree demogra*)ic varia,les 9young de*endency ratio, old de*endency
ratio and life e8*ectancy:, )ig)er levels of education and greater ur,ani/ation as
inde*endent factors in e8*laining insurance demand' #conomic varia,les li;e Gini inde8
and 3uman develo*ment inde8 are new additions along wit) institutional varia,les
reflecting *olitical sta,ility, access to legal ,enefits and an inde8 of institutional
develo*ment were used' $)e analysis considering t)e time *eriod 1+61 to 2000 s)ows
t)at countries0 wit) develo*ed ,an;ing system, )ig) income and lower inflation )ave
)ig)er life insurance consum*tion' $)e association of insurance demand wit)
demogra*)ic is not statistically strong )owever older t)e *o*ulation, )ig)er tends to ,e
insurance consum*tion' $)e lu8ury good nature of insurance did not reflect t)roug) its
association wit) income distri,ution'
!n contrast to -ec; et al' 92003: results, t)e study ,y 1ard et al 92003: is indicative of t)e
fact t)at im*roved civil rig)ts and *olitical sta,ility leads to an increase in t)e
consum*tion of life insurance in t)e &sian region as well in t)e 2#C" region' =ollowing
a*orta et al 91++7, 1++C, and 2000: wor;s relating to su**ortive as*ect of legal
environment for finance, t)ey too considered t)e same in determining insurance demand'
&naly/ing t)e data from 1+C7 t)roug) 1++C for 2#C" and &sian countries, t)ey
10
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
o,served t)at income elasticity ,etween develo*ed economies and emerging economies
are consistent wit) KS7cureM insurance growt) findings ,y #n/ 92000:'
Issues and uestions
$)e maAor issue to start wit) would ,e to re7e8amine t)e significant varia,les t)at can
,est fit as determinants of life insurance demand' Recently, t)ere are num,er of studies
on single economies: 3wang and Gao 92003: study were on t)e C)inese economyG im
and 3a,erman 92005: focuses on MalayasiaG and 3wang and Greenford 9200<: on
Mainland C)ina, 3ong 4ong and $aiwan' enten and Rulli 92006: e8*lored t)e time
series *ro*erties of t)e demand for life insurance in &ustralia using a novel statistical
*rocedure t)at allows uno,serva,le com*onents to ,e e8tracted'
Diet/ 92003: and 3ussels et al' 9200<: )as reviewed t)e efforts of researc)ers to e8*lain
consumer ,e)aviour concerning t)e *urc)ase of life insurance for almost <0 years' $)e
review of earlier studies concludes t)at ,ul; of t)e em*irical studies underta;en finds a
*ositive association ,etween increase in savings ,e)aviour, financial services industry
and demand for life insurance' $a;ing t)is forward, our first issue is to see w)et)er or not
*er ca*ita gross domestic savings and financial de*t) influences life insurance
consum*tion' G"P and Per7ca*ita G"P are often )ig)ly correlated wit) t)e *ro8y
varia,les measuring insurance demand7 density and *enetration' 1e t)erefore ignore
t)ese two varia,les and assume t)at as income grows, it will add to insurance demand via
rise in t)e savings com*onent i'e', G"%'
$)e demogra*)ic factors de*endency ratios, adult literate *o*ulation, life e8*ectancy and
adult literate *o*ulation are considered and in line wit) earlier studies we e8*ect t)ese to
,e significant in e8*laining life insurance demand' !t is e8*ected t)at young de*endency
will ,e negatively related and t)e rest t)ree demogra*)ic varia,les are e8*ected to )ave a
*ositive relation' 1e )ave assumed first t)at consumer *rice inde8 9CP!: will ,e t)e ,est
*ro8y for inflation alt)oug) we )ave considered log difference of CP! as an alternative
measure for inflation' astly we )ave t)e real interest rate, w)ic) is in our case t)e
de*osit interest rate minus inflation' !n $a,le 1, we *resent our identified @*otential0
11
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
determinants of life insurance consum*tion wit) t)e e8*ected signs for t)e cross7
sectional analysis'
Ta#le 1+ $eter*inants, Resear(h ,uestions an& )-pe(te& Relationship
$eter*inants ,uestions )-pe(te& "ign
GDP per-capita $s income a significant factor in e35laining life insurance
consum5tion6
(
GDS per-capita $s level of saving a significant factor in e35laining life insurance
consum5tion6
(
Financial Depth Does the level of financial sector develo5ment have any s5ill7over
effect on enhancing insurance consum5tion6
(
Urbanization 8hat relation e3ists bet9een the rate of urbani:ation and insurance
consum5tion6
(
Total Dependency
Ratio
;he im5act of rising <de5endent 5o5ulation= on insurance
consum5tion
(
Young
Dependency
Ratio
;he im5act of rising 5o5ulation in the /7*, years age grou5 on
insurance consum5tion
>
Old Dependency
Ratio
;he im5act of rising 5o5ulation above .- years of age on insurance
consum5tion
(
Adult Literate
Population
8hether or not level of education, as an indicator of a9areness,
affect insurance consum5tion6
(
Lie !"pectancy
at #irth
;he effect of life e35ectancy on insurance consum5tion (
$rude Death Rate ;he effect on number of deaths on insurance consum5tion (
%nlation Do 5rice fluctuations affect insurance consum5tion6 >
Real %ntere&t Rate 8hat is the effect of changes in interest rate on insurance
consum5tion6
>
Price Does insurance 5rice determine the demand for insurance6 >
=or t)e time series analysis on !ndia, we )ave included varia,les G"P *er7ca*ita, total
de*endency ratio, C"R, and ur,an *o*ulation growt) rate w)ic) are su**osed to
*ositively stimulate growt) ,ut our *rice varia,le is e8*ected to ,e inversely related'
!ethodolo"#
The $% Asian Economies& $)e study focuses on 12 selected economies wit) 5 from t)e 7
%&&RC
7
countries 9!ndia, -anglades), Pa;istan and %ri an;a:G 2 from t)e Greater
C)ina region 9C)ina and 3ong 4ong: and 6 from t)e 10 nations &%#&.
C
9!ndonesia,
Malaysia, P)ili**ines, %inga*ore, $)ailand, and Oietnam:' $)ese 12 nations a*art from
,eing t)e most *romising emerging and growing economies, most of t)em are also
struggling to overcome *overty' #8ce*t 3ong 4ong and %inga*ore, most of t)e countries
7
%out) &sia &ssociation for Regional Coo*erationG t)e rest 3 countries are Maldives, .e*al and -)utan
C
$)e &ssociation of %out)east &sian .ations and t)e ot)er mem,ers are -runei "arussalam, Cam,odia,
aos and Myanmar'
12
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
)ave less t)an >%" <000 G"P *er ca*ita' %i8 of t)e selected economies in our sam*le
)ave G"P *er ca*ita of less t)an >%" 1000'
Ta#le 2+ .ife Insuran(e Pre*iu*s an& Penetration /igures for the 12 Asian
)(ono*ies
$''( $''' %))(
*remiums+ Densit#
,
*remiums+ Densit#
,
*remiums+ Densit#
,
-an"ladesh /'/*
*
/'///* */0'7* /'+* 0*',+ *'..
.hina /7',* *'7* */7-'*7 +'-+ 0/-+,'0* 0'./
/on" 0on" 2536.44 422.88 4973.26 752.73 14643.54 2122.25
India 0.-'7/ 0')) .0/.'-/ .'0* *,7*.'0. *0'.0
Indonesia .-'/+ 0'0/ 7+/'7) 0'+0 *,).'*) .'++
!ala#sia *'0) ./'+* *+)'*, +0'0 07)+'/* *-'-0
*a1istan **'// *'77 *+.',) *'0+ 0/'*, *'-*
*hilippines 07-'/+ -'.* -0+')) 7'/ 7)-'7+ )'7-
Sin"apore 1476.57 431.62 2837.70 718.04 6437.32 1532.70
Sri Lan1a /'/*
*
/'///. 70'+/ ,'/- **-',. -')-
Thailand *0-'/ *'/ *--/',+ -'7, )0.',) ,.'*/
2ietnam /'/*
*
/'/// 0-'. /',. ,,',. -'0+
Source& Sigma 9various issues:, %wiss Re'
Pin Millions of >%" at constant 2000 *rices
Q ife Premiums as *ercentage of total mid7year *o*ulation
1
=igures are e8tra*olated
$)e overall s*ending on insurance and related *roducts in t)ese economies is very low
w)en com*ared to t)e 2#C" economies' 1it) *o*ulation growt) rate a,ove or near 1
*ercent and a com*aratively low level of insurance consum*tion Imeasured ,y *er ca*ita
*remium, indicates t)e e8isting economies of sco*e and t)e *otential for growt)'
&lt)oug), some of t)ese economies were gri*ed ,y t)e &sian =inancial Crisis, t)e issue
is to create stronger infrastructure for financial services and t)ere,y ,oost t)e
underdevelo*ed insurance mar;et'
$)ese economies dis*lay immense diversity in et)nicity, culture and religion' $)e legal
system too varies a lot wit) most of t)em following t)e #uro*ean legal system due to
t)eir colonial influence' =ollowing t)e General &greement on $rade and %ervices, most
of t)ese economies agreed to o*en u* t)ere financial sector in general and insurance in
*articular, to eit)er domestic *rivate com*anies or foreign com*anies or to ,ot)' $a,le <
s)ows t)at most of t)ese countries were dominated ,y large num,er of insurers ,ut t)ere
insolvencies called for nationali/ation' !t was in t)e ,eginning of +0s t)at regulatory
c)anges were initiated and im*lemented w)ic) once again o*ened t)e insurance industry
13
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
,ut wit) more cus)ions in terms of regulation and su*ervision' $)e 11 years *eriod
considered for t)is study from 1++5 to 2005, witnessed im*ortant regulatory c)anges'
"uring t)is *eriod, in most of t)e economies e8*ansion of life segment of t)e insurance
mar;et *articular gained momentum' $)is )el*ed to reAuvenate insurance density' $)e
only e8ce*tion from t)is increasing trend was Pa;istan, w)ere actually density )as
dro**ed'
Insuran(e penetration in 12 Asian )(ono*ies
/

,
.
+
*)), *))) //,
0ear
P
r
e
*
i
u
*
s

a
s

1

o
f

2
$
P
Bangladesh China ?ong !ong $ndia
$ndonesia @alaysia Pa"istan Philli5ines
Singa5ore SriAan"a ;hailand Bietnam
%ource: %wiss Re, Sigma 9various !ssues:
Ta#le 3+ Correlations
*er .apita 3D* 3DS Life *remium *enetration Densit#
*er .apita 3D* *'/// 7/'/+/ /'/- /'+**CC /'+7)CC
3DS *'/// /'+*7CC /'/-/ 7/'/-7
Life *remium *'/// /',-0CC /'0/-
*enetration *'/// /')*/CC
Densit# *'///
PPCorrelation is significant at t)e 0'01 level 927tailed:'
Time Series Anal#sis on India& 1e e8amine t)e determinants e8*laining t)e demand for
insurance in general and life in *articular for t)e *eriod starting 1+6< to 2005' & num,er
of c)anges )ave ta;en *lace during t)is selected time )ori/on' !n 1+<6, t)e life segment
was nationali/ed: we witnessed a war and a famine w)ic) )ad adversely affected t)e
economyG in late C0s t)e *rocess of li,eralisation was initiated wit) t)e financial sector
reforms in 1++1G and lastly 1+++ mar;ed t)e re7o*ening of t)e insurance mar;et and
esta,lis)ment of !R"&' $)e de*endent varia,les considered are insurance *enetration,
insurance density and net *remiums'
15
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
Data Source& $)e *anel for 12 selected &sian countries studied for 11 years starting
1++5 to 2005 is constructed using annual aggregate data from different secondary source'
$)e !nsurance *remium figures are collected from various issues of Sigma, a *u,lication
from %wiss Re' $)e #conomic as well as demogra*)ic varia,les used are collected from
t)e International inancial Statistics !""# and t)e World $evelopment Indicators !""#'
$)e e8*lanatory varia,les in t)e model are t)e economic and demogra*)ic varia,les
9refer %ppendix: 1:' $)e time series analysis uses insurance ,usiness data availa,le for
t)e *eriod 1+6< to 2005 from t)e ife !nsurance Cor*oration of !ndia and !nsurance
Regulatory and "evelo*ment &ut)ority0s &nnual Re*orts' "ata on economic and
demogra*)ic varia,les are collected from W$I !""#, availa,le in local currency units at
current *rices' &ll t)e economic varia,les are transformed to constant 2000 *rices'
Research Desi"n &
.ross4.ountr# Anal#sis&
1e estimate two different *anel data regression models' $)e models are different since
we are trying to re*resent life insurance demand ,y two different de*endent varia,les:
insurance *enetration and insurance density' $o address t)e *ro,lems of estimation and
inference associated wit) *anel data, we estimate ,ot) t)e fi8ed effects model 9=#M: and
t)e random effect model 9R#M:' $)e decision regarding t)e ,est model ,etween =#M
and R#M is arrived at using t)e 3ausman test'
$)e s*ecification of t)e fi8ed7effects model as well as t)e random effect model is as
follows:
1 2
log9 : log9 : log9 :
it i it it
&enetration G$S&' I($ = + +
3 5 < 6 7
log9 : log9 : log9 : log9 :
it it it it it
)*B +$* ,$* %$-& T-E.* + + + + +
C +
91J :
it it it
*I* '&I + + +
&nd t)e second regression e?uation estimated is,
1 2
log9 : log9 : log9 :
it i it it
$ensity G$S&' I($ = + +
3 5 < 6 7
log9 : log9 : log9 : log9 :
it it it it it
)*B +$* ,$* %$-& T-E.* + + + + +

C +
91J :
it it it
*I* '&I + + +
3ere, su,scri*t i denotes t)e country, t)e su,scri*t t re*resents time,
i
are cross7
sectional interce*t terms,0s are t)e slo*e *arameters and
it
are random error term' $)e
1<
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
model argues t)at variations in t)e inde*endent varia,les in t)e 12 selected economies
may contri,ute to insurance *enetration and insurance density'
India& Time Series Anal#sis&
=or any time7series analysis t)e first ste* is to deal wit) t)e availa,le data series, c)ec;
for seasonality and t)en transformations if needed de*ending u*on t)e modelJ)y*ot)esis
to ,e tested' & transformation is made to varia,les at level7valueG )owever, varia,les of
rate value are not transformed ,ecause t)ey are already in a *referred form as t)ey are a
measure of c)ange' -ased on t)is rationale, t)e varia,les of rate value form, i'e', P#.,
"#., !.=R, R!R, $"R, 2"R, B"R, C"R, $#RR and >R-GR not su,Aect to any
transformation' $)e varia,les of t)e level value form are su,Aect to a transformation ,y
ta;ing t)e natural logarit)m of t)eir level values' $)e varia,les transformed are G"P *er7
ca*ita, G"% *er7ca*ita, =!.", PR!C#, and PR#M'
&ll time series varia,les, w)et)er transformed or not transformed are nor su,Aect to test
for unit roots in order to investigate t)eir stationarity *ro*erties' 1e )ave considered two
different tests: t)e &ugmented "ic;ey =uller 9&"=: $est and t)e 4P%% $est'
Au"mented Dic1e# 5uller (AD5) Test: $)is test is ,ased on t)e model of t)e form,
1
P
1
1
p
t t / t / i
/
y y y u

=
= + +
L L 9i:
and t)e *air of )y*ot)esis isG
0 1
: 0 : 0 0 versus 0 = <
$)e t7statistics is calculated for t)e co7efficient S from an 2% estimation of 9i: 9"ic;ey
and =uller, 1+7+:' $)e null )y*ot)esis is reAected if t)e calculated t7statistics is less t)an
t)e critical value and im*lies t)at t)e series is stationary and vice versa'
0wiat1ows1i4*hillips4Schmidt4Shin (0*SS) Test: $)e integration *ro*erties of a series
y
t,
may also ,e tested against a unit root 9t)e null )y*ot)esis t)at t)e series is stationary:
( ) ( )
0 1
: T 0 : T 1
t t
0 y I against 0 y I
4wiat;owis;i et al', 91++2: )as derived a test for t)is *air of )y*ot)esis' !f t)ere are no
linear trend term t)an, t)ey start from a
t t t
y x 1 = +
G w)ere x
t
is a random wal; and
1 t t t
x x v

= +
, w)ere ) (
2
T 0,
t v
v iid
and 1
t
is a stationary *rocess' $)e new *air of
)y*ot)esis is now:
2 2
0 1
: 0 : 0
v v
0 against 0 = > '
16
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
$)e test statistics *ro*osed is as under:

=
T
t
t
S
T
2&SS
1
2 2
2
U J
1

w)ere,
=
=
t
/
/ t
S
1
U
G
y y
t t
= U
and
2
U

is an estimator of t)e longrun variance

of t)e *rocess 1
t
' $)e null )y*ot)esis of stationarity is reAected for larger values of t)e test
statistics 4P%% and vice versa'
-ased on t)e two test results, all t)e series are again transformed in order to ma;e t)em
stationary' $)e de*endent varia,les are insurance *enetration and insurance density' !n
general t)e model9s: estimated is of t)e form:
[
] Ratio "e*endency , Rate, "eat) Crude
, #8*ectancy ife Price, !nterest, !nflation, %aving, !ncome, t, "evelo*men =inancial f "emand
on )r3ani1ati
=
L' L' 12
Results and Interpretation
.ross .ountr# Anal#sis&
1e )ave dro**ed G"P *er7ca*ita to reduce t)e )ig) level of correlation in t)e fi8ed
effects model and we )ave considered t)e inverse of CP!' $)e discussion in t)is section is
,ased on t)e final results of t)e models and t)e summary results are re*roduced in $a,le
5'
2ur results ,ased on t)e ,alanced *anel, allows us to e8*loit ,ot) cross7country and time7
series variation in t)e data' $)e fi8ed effect model 9Model 1: s)ows t)at variation in
insurance *enetration is e8*lained ,y G"% *er7ca*ita, financial de*t), ur,ani/ation, old
de*endency ratio, adult literacy, life e8*ectancy and measure of inflation' -ut, in t)e
random effect model 9Model 2:, financial de*t), old de*endency ratio and inflation turn
out to ,e significant' =rom t)e fi8ed effects regression, t)e statistics for testing t)e Aoint
significance of t)e country effect is 13'1+ V E11,10<FW' $)is is greater t)an t)e critical
value and t)e evidence is strongly in favour of country effect in t)e data' &lso, t)e
3ausman test for t)e fi8ed and random effects regressions gives t)e test statistics 101'C+
wit) + degrees of freedom' $)e critical value of c)i7s?uare wit) + degrees of freedom is
17
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
16'+2 at +< *ercent level of significance' $)us, we can reAect t)e fact t)at t)e individual
effects are uncorrelated wit) ot)er regressors in t)e model' $)is furt)er suggests t)at of
t)e two alternatives we )ave considered, t)e fi8ed effects model 9Model 1: is a ,etter
c)oice'
Ta#le 4+ Panel $ata )sti*ation
Model 1 Model 2 Model 3 Model 4
Dependent Variable Insuran(e Penetration Insuran(e $ensity
Independent Variables Di3ed Effect Eandom
Effect
Di3ed Effect Eandom
Effect
Constant 7*.*'))0C
10)'/)*4
7-)'*-C
17',)*4
7*.*'+-,C
10)'07+4
7..'-0C
1+'-)-4
FDS 5er7ca5ita '))7C
1*'/-,/4
/'+.+
1/'+--4
0'*)/C
1*'/.4
*'07,
1/'+7-4
Dinancial De5th ,'.+C
1*'0-4
,').C
1*'//4
,'0)C
1*'0.4
,',)-C
1*'/)4
Grbani:ation 7/'/++C
1/'/,04
7/'/*
1/'/*.4
7/'/+7C
1/'/,04
7/'//
1/'/*74
Houng De5endency Eatio 7.'.)7
1-').+4
7*'.*7
10')/*4
7.'+/,
1.'/*4
7*'7,.C
1,'/,4
2ld De5endency Eatio 7+',,+C
10'.0)4
7+'+.)C
10'0++4
7+'*-C
10'...4
7+'+*.
10',,74
%dult Aiterate Po5ulation 7**'*/0C
10'*4
/'7-
1/',)-4
7*/'77-C
10'0.4
/'.,
1/'-04
;otal Aife E35ectancy at Birth **)'7,)C
1/',)4
/'7*
1*0'),/4
**+',,+C
1'07,4
0'7*.
1*,'--.4
$nflation 7+'-//C
1,'./4
7*/',+-C
10')**4
7+',*,C
1,'.-,4
7*/'*,)C
10')0-4
Eeal $nterest Eate /'**)
1/',)4
7/'*-*
1/'+,4
/'*/)
1/'-*4
7/'*00
1/'7)4
2bservations *. *. *. *.
D ;est Statistics I

*0'*) */*'-7 *0'*- *0,').

E

8ithin /'.,.7 /'-++ /'.-)+ /'-.0)

Bet9een /'/+) /'--*, /'0.- /'7/+.
2verall /'/.+ /',.7 /'-++ /'.0.
?ausman ;est Statistics */*'+) )*'/0
CStatistically significantJ Digures in 5arentheses are standard errors
%imilarly, w)en we use insurance density as t)e de*endent varia,le to see t)e im*act of
variations of t)e inde*endent varia,les, fi8ed effects model 9Model 3: turns out to ,e a
,etter c)oice' Model 3 also suggest t)at t)ere is a significant relation ,etween G"% *er7
ca*ita, financial de*t), ur,ani/ation, old de*endency ratio, adult literacy, life e8*ectancy
and inflation'
$)e regression results from Models 1 and 3 s)ows t)at G"% *er7ca*ita )as a *ositive
relations)i* wit) insurance density and *enetration' $)is suggests t)at, as savings activity
1C
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
increases, it will *us) u* *er ca*ita insurance e8*enditure and t)ere ,y en)ance insurance
reac)' $)e significant *ositive relation ,etween density and *enetration wit) financial
de*t) suggest t)at wit) strengt) in t)e ot)er segments of t)e financial sector, insurance
sector too will ,enefit' &lt)oug) ur,ani/ation is significant, its sign is contradictory to
earlier studies and our e8*ectation t)at t)e level of ur,ani/ation in an economy raises
insurance consum*tion' !n our case, it suggests t)at ur,ani/ation will decrease insurance
consum*tion' $)is may ,e as a result of rural7ur,an migration and a rising s)are of *oor
*o*ulation in t)e ur,an centres in most of t)e selected economies'
Boung de*endency ratio )as t)e e8*ected sign ,ut it0s insignificant in all t)e 57models'
2ld de*endency ratio is significant ,ut again its sign does not corro,orate wit) earlier
studies' !t suggests t)at t)e growt) of t)e old de*endency ratio will decrease insurance
density as well as insurance *enetration' !nteresting, t)e education varia,le in t)e model,
adult literate *o*ulation is significant ,ut not in line wit) e8*ectation' $)is mig)t
*ro,a,ly give rise to t)e issue w)et)er )ig)er education or a certain minimum,
guarantees awareness of insurance ,enefits' $)e last of t)e demogra*)ic varia,le, t)e life
e8*ectation of t)e total *o*ulation at ,irt) is )ig)ly significant and suggests t)at as living
conditions and longevity of life im*roves, demand for insurance *roducts would go u*'
&s e8*ected, inverse of inflation is inversely related wit) density and *enetration and is
significant ,ut real interest rate is not' $)e relation ,etween inflation wit) t)e demand
*ro8ies, do not corro,orate wit) earlier studies and )ence we conclude t)at current
interest rate or *rice situation does not affect insurance consum*tion decisions'
Time Series Anal#sis
$)e results o,tained are more sur*rising and re7enforces one to *ro,e furt)er w)y most
of t)e determinants successful in determining t)e demand for insurance fails to e8*lain in
t)e !ndian scenario' 1e )ave considered two de*endent varia,les 9i'e', insurance
*enetration and density: and tried to e8*lain t)e variation in t)ese ,y set of economic
varia,les 9income, savings, *rices of insurance *roduct, inflation and interest rates: and
1+
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
demogra*)ic varia,les 9de*endency ratio, life e8*ectancy at ,irt), crude deat) rate and
ur,ani/ation:' $)e results o,tained are *resented in $a,le 7'
Ta#le !+ Results of Ti*e "eries Analysis
Dependent Variable Insuran(e Penetration Insuran(e $ensity
Independent Variables $ $$ $$$ $B
Constant 7/'*,-7)-
1/'*.-*4
7/'/7)-0
1/'*.,-4
/'//*+-,
1/'//0-4
/'//0/+
1/'//0)4
FDP 5er7ca5ita 7/'--0C
1/'7.)4
7/'--,)C
1/'7)74
/'-*/*0C
1/')/+4
/'-*/7++C
1/')**4
FDS 5er7ca5ita
Dinancial De5th /'-+,-+C
1/'0/./4
/',))//*
1/'0/.,4
/'-)0*,C
1/'0/+4
/'-.))7*C
1/'0*7.4
Price I Cost of $nsurance /',)..7C
1/'/)+,4
/',0+.*C
1/'*//04
/',/..0C
1/'/))04
/',+00C
1/'*//-4
$nflation /'//-,-.
1/'.4
7/'/)7-.+
1/'.*74
7/'0/,+.
1/'.,+74
7/'00,)0
1/'.,/+4
Eeal $nterest Eate 7/'*7+/7C
1/'//0*4
/'//-,-7C
1/'//0*4
/'//./0C
1/'//04
/'//.//-C
1/'//04
Grbani:ation /'/-)C
1/'/,/)4
/'/-0.-C
1/'/)/4
/'/,+,).C
1/'/7-4
/'/-0/-*C
1/'/++4
Houng De5endency Eatio ,')/)0+7
1.'/7*04
-'./-+
1-').7)4
2ld De5endency Eatio 7++'..0*
1+/'-+/4
77-'-77.0
17+'77/04
Crude Death Eate 7/'//.00,
1/'//+-4
7/'//.*
1/'//+-4
;otal Aife E35ectancy at Birth /'/0.-)7
1/'/+/4
/'//0)
1/'/,/74
2bservations 0+ 0+ 0+ 0+
D ;est Statistics .'., .',- .'* .'**
E

/'.,7/ /'.,/, /'.+* /'.7.

%dK7E

/'-,). /'-,* /'--- /'-,)

D78 Statistics '7--. '70-/ '7/* '7)
P %tatistically %ignificant at some X level of significanceG values in *arent)eses are standard errors'
$)e analysis s)ows t)at income 9G"P *er7ca*ita:, financial de*t), *er *olicy *rice of
insurance *roducts and real interest rates are significant varia,les' &mong t)e
demogra*)ic varia,les, ur,ani/ation i'e', t)e growt) rate of ur,an *o*ulation as a
*ercentage of total *o*ulation is significant' !n terms of t)e sign, t)e results are
contradicting our e8*ected signs' !ncome is inversely related to t)e demand insurance
*enetration ,ut in line wit) e8*ectation is *ositively related to insurance density'
=inancial de*t) )as a *ositive relation wit) ,ot) *enetration and density suggesting t)at
as t)e financial system as a w)ole growsG it will )ave a *ositive im*act in raising demand
for insurance *roducts' !n ot)er words, we can say t)at as t)e traditional ,oundaries ,rea;
,etween s*eciali/ed financial institutions, t)e overall involvement in financial sector in
20
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
different ,usiness would lead to ,etter su**ly of insurance *roduct and t)us raise demand
furt)er' Price is e8*ected to ,e inversely related to demand' -ut, )ere it s)ows t)at in all
t)e four different models, it )as a *ositive association and suggests t)at @a )ig)er
insurance cost encourages *urc)ase of insurance0' $)e last of t)e significant varia,le,
ur,ani/ation )as a *ositive relation and a*tly corro,orates wit) t)e Austification t)at as
more and more ur,an centres are created, t)e demand or consum*tion of insurance
*roducts would increase'
.onclusions
2verall, our cross7country analysis confirm t)at if we e8ogenously consider income to ,e
a crucial factor in e8*laining insurance consum*tion, economic varia,les of im*ortance
would ,e gross domestic savings, level of financial sector develo*ment and inflation' &s
s*eciali/ed financial institutions turns to financial conglomerates, one im*ortant *olicy
im*lication can ,e strengt) and wea;nesses of ,an;ing and ot)er non7,an;ing
institutions w)ic) mig)t )ave a *ositive or negative s*illover effect on t)e insurance
industry' &s more ,an;s line u* for insurance service *rovision, t)e entry of t)ese
institutions will also *us) u* demand' 2ur analysis suggests t)at as t)e savings increases
it raises insurance consum*tion' -ut insurance as suc) is not *urely savings and )ence its
*urc)ase may smoot)en t)e income or wealt) over time' !f savings *lus life ris;
insurance *roducts are sold, it mig)t ,oost insurance consum*tion' &lt)oug) real interest
rate was not significant in our cross country analysis, it turned out to ,e significant in our
time series analysis' %ome variant of it may also *lay an im*ortant role in e8*laining
individuals0 c)oice ,etween insurance and ot)er savings instruments'
2ur results ,ased on *anel of 12 economies over 11 years su**orts t)e fact t)at
demogra*)ic varia,les li;e life e8*ectancy, young and old de*endency ratio, adult
literacy rate and rate of ur,ani/ation are significant determinants of life insurance
demand' 3owever, t)e same cannot ,e concluded from t)e time series analysis on !ndia
as only ur,ani/ation )as some significant relation to t)e demand *ro8ies' $)e
insignificant de*endency ratios once again *roved conclusion of Rosen/weig0s 91+CC:
study t)at im*licit insurance *rovided ,y networ;s of family and friends in t)e case of
21
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
!ndia *rovided some ;ind of alternative arrangement to ta;e care of life related ris;s
+
'
$)erefore, if we consider t)at t)e older generation in most of t)e low7income and
develo*ing countries considered consum*tion of insurance as an unnecessary e8*ense,
t)e *olicy recommendation to en)ance awareness among t)e current generation is via
*ro*er information dissemination'
&ccording to iedt;e 92007: insurance s)ould ,e considered a ;ey com*onent of
economic develo*ment and t)e ,est mec)anism to ta;e care of multidimensional ris;s in
modern economies' !t is necessary to clear t)e confusion regarding considering life
insurance as a su*erior or lu8ury good among *otential consumers in develo*ing
countries wit) com*aratively low *er7ca*ita income' -ut as *er7ca*ita income is steadily
cree*ing u* in t)e selected economies
10
wit) c)anges in t)e standard of living, t)e
su**liers mig)t stimulate demand and increase t)e availa,ility of insurance *roducts'
$)is would reduce t)e scarce and costly outloo; of life insurance *roducts' $)e study can
,e e8tended considering more varia,les and dummies to loo; for t)e country and time
s*ecific factors affecting demand' $)e results are of im*ortance to t)e *olicy ma;ers if
t)ey are as*iring to elevate insurance density and *enetration in t)e economies' Most of
t)e selected economies )ave undergone c)anges, *articularly in terms of regulatory
reforms recently' !t would ,e useful to ta;e a muc) ,igger sam*le in terms of countries
and *eriods considered, to understand w)y some of t)e varia,les ,e)aved so differently
t)an e8*ected'
Appendi6 $&
T#pe of 2ariable 2ariable 7ame Description
ECONOMIC Fross Domestic Product
1FDP4
$t is the sum of value added by all resident 5roducers 5lus any
5roduct ta3es 1less subsidies4 not included in the valuation of
out5ut' Fro9th is calculated from constant 5rice FDP data in local
currency unit
FDP 5er7ca5ita $t is gross domestic 5roduct divided by midyear 5o5ulation'
FDS 5er7ca5ita ;he gross domestic savings divided by midyear 5o5ulation'
Consumer Price $nde3
1CP$4
$t reflects changes in the cost to the average consumer of acLuiring
a bas"et of goods and services that may be fi3ed or may change at
s5ecified intervals, such as yearly'
+
$)e information was collected from t)e study ,y $ownsend91++<: w)ic) dealt wit) some of t)e issues
relating to evaluation of ris; ,earing systems in low7income economies'
10
%inga*ore and 3ong 4ong are )owever two countries in t)e )ig)7income grou*'
22
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
AiLuid Aiabilities as a M of
FDP 1for Dinancial De5th4
1D$ND4
;hey include ban" de5osits of generally less than one year 5lus
currency' $t is the sum of currency and de5osits in the central ban",
5lus transferable de5osits, electronic currency, 5lus savings in time
and savings de5osits, foreign currency transferable de5osits,
certificate of de5osit and securities re5urchase agreements etc'
$nflation Eate 1$NDE4 Aog difference of CP$
Eeal $nterest Eate 1E$E4 De5osit interest minus inflation
DEMOGRAPIC
Po5ulation ;otal mid7year 5o5ulation
;otal De5endency Eatio
1;DE4
;he ratio of de5endent young Po5ulation and de5endent old
Po5ulation to the 9or"ing age 5o5ulation7those ages *- to .,
Houng De5endency
Eatio 1HDE4
;he ratio of de5endents75eo5le younger than *- years of age to
the 9or"ing age 5o5ulation7those ages *- to .,
2ld De5endency Eatio
12DE4
;he ratio of de5endents75eo5le older than ., years of age to the
9or"ing age 5o5ulation7those ages *- to .,
%dult Aiteracy Eate
1%DAP4
;he 5ercentage of 5eo5le ages *- and older 9ho can, 9ith
understanding, both read and 9rite a short, sim5le statement about
their everyday life'
Aife E35ectancy at Birth
1;AENE4
;he number of years a ne9born infant 9ould live if 5revailing
5atterns of mortality at the time of its birth 9ere to stay the same
throughout its life'
Crude Death Eate 1CDE4 No' of Deaths 5er O/// 5o5ulation
Grban Po5
n
1GEB4 ;he 5o5ulation of the urban agglomeration, a contiguous inhabited
territory 9ithout regard to administrative boundaries'
Grban Po5
n
Fro9th Eate
GEBFE
Percentage change in Grban Po5
n
as a 5ercentage of ;otal Po5
n
IN!"RANCE
#"!INE!!
Premiums 1PEE@4 ;otal Premium generated 1net4
Price 1PE$CE4 Eatio of ;otal *
st
Hear Premium $ncome to Sum of No' of ne9
individual 5olicies sold and the no' of lives covered under grou5
schemes'
Premium Density 1DEN4 ;otal Premiums 5er7ca5ita
Premium Penetration
1PEN4
;otal Premium as a 5ercentage of Fross Domestic Product
1PS%B4 ;otal Premium $ncome as 5ercentage of Fross Domestic Savings
i !9on, 8' Pean 1//*4 <;o9ard Dree ;rade in Services# ;he %SE%N $nsurance @ar"et=, %%F
Occa&ional Paper, No' 0J $nternational $nsurance Doundation, 8ashington, DC'
!9on, 8' Pean 1//4 <;he $nsurance @ar"ets of South %sia=, %%F Occa&ional Paper, No' ,J
$nternational $nsurance Doundation, 8ashington, DC'
!9on, 8' Pean 1//4 <;he 8;2 and $nsurance in Freater China# ;he Peo5les Ee5ublic, ?ong
!ong, @acau, and ;ai9an=, %%F Occa&ional Paper, No' .J $nternational $nsurance Doundation,
8ashington, DC'
23
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
Economies
Ta#le + 5 Development of ife Ins!ran"e Ind!str#
$
Re"ulator# Authorit# *re4Re"ulator# Re"ime Re"ulator# .han"e 5orei"n 8wnership 7o9 of
Insurers
-an"ladesh ;he De5t' of $nsurance,
headed by the $nsurance
Directorate of the @inistry of
Commerce'
%fter se5aration from Pa"istan in
*)7, state7o9ned mono5oly life
insurer Piban Bima Cor5' 9as
established in *)70
*)+,
%llo9ed -/M 5rivate
under9riting in *))/
Not %llo9ed *+
* Public
.hina *9R9 China $nsurance Eegulatory
Commission 1CE$C4 formed in
*))+
P$CCQs
*
mono5oly as insurer ended
in *)+-J - ne9 5ublic insurer
createdJ , regional insurer e3isted
*))J $nterim @gmt'
Eegulation for foreign
$nsurance $nstitutions allo9ed
foreign entry
%llo9edJ //
Eegulations on %dmn' of
Doreign7invested $ns' Co'
,7
/on" 0on" 2ffice of the Commissioner of
$nsurance 12C$4J *))/
Com5aratively the most com5etitive
insurance mar"et 9ith over /,
insurers
*))7after it became PECQs
s5ecial administrative region
%llo9edJ */, ca5tive
insurers in o5erating from
- countries
*,/
India $nsurance Eegulatory and
Develo5ment %uthority 1$ED%4J
*)))
;he Aife $nsurance Cor5' %ct of *)-.
merged -. life insurers to form A$C

*//M 5rivate 5artici5ation

allo9ed since *)))
.M eLuity share in local
com5anies
*.
* Public
Indonesia ;he Directorate Feneral for
Dinancial $nstitutions along 9ith
@inistry of Dinance
777 777 %llo9ed .
!ala#sia @inistry of Dinance de facto
regulator 9ith administration by
Ban" Negara @alaysia 1BN@4
$n *)+, ;a"aful $nsurance %ct
5ermitted insurer o5erations based
on $slamic 5rinci5les'
%l9ays o5en to Private
Aa9 and Eegulatory changes
in *)).'
%llo9ed 7
*a1istan Securities and E3change
Commission of Pa"istan
1SECP4J *)))
$n *)7, 0, out of -/ e3isting
insurers 9ere merged to form State
Aife $nsurance Cor5'

*))/
*//M 5rivate entry %llo9ed
since *))
%llo9ed -
* Public
*hilippines ;he $nsurance Commission
12omiyon ng Seguro4
Private sector 9as al9ays
em5hasi:ed
;he *epu3lic %ct (o4 5678 of
*)). 5ermits *//M foreign
o9nershi5
Ne9 foreign insurer is not
allo9ed to hold a
com5osite license
0,
Sin"apore ;he $nsurance De5artment of
the @onetary %uthority of
Singa5ore 1@%S4
Gntil *)./s the mar"et 9as loosely
regulated
///
;he mar"et 9as liberali:ed to
direct insurers
$n ///, ,)M restriction
on foreign o9nershi5 in
local com5anies 9as lifted
+
Sri Lan1a ;he $nsurance Board of Sri
Aan"aJ //*
$nsurance Cor5' %ct in *).*
nationali:ed the life insurance
$ndustry to form $CS
,
//*,
*//M 5rivate o5eration
allo9ed since *)+.
G5 to )/M foreign
investment in local
com5anies
*/
* Public
Thailand ;he De5artment of $nsurance
1of the @inistry of Commerce4
inde5endent from //
@ar"et 9as bloc"ed for decades till
late *))/s
*))
$nsurance business is
relatively liberal
-M of foreign o9nershi5
in domestic insurers
-
/
2ietnam @inistry of Dinance Baoviet, a state o9ned life insurer
started o5eration in *)). but 9as
initially created for non7life insurance
*)))
///, a55roved & enacted
ne9 set of la9s
%llo9ed foreign
com5anies in *)))
follo9ed by in ///
*/
0 Public
Source# C Eefer to Endnote i
*
Peo5leQs $nsurance Com5any of ChinaJ

Aife $nsurance Cor5oration of $ndia under the control of Fovt' of $ndiaJ

0
;he Cor5oration established under %rticle ** of
the Aife $nsurance 1nationali:ation4 2rder of *)7 J
,
$nsurance Cor5oration of Sri Aan"a, renamed the Sri Aan"a $nsurance Cor5oration Atd' in *))0J
-
to be lifted to
,)MJ
2<
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
Ta#le 5+ "ele(te& )(ono*i(, $e*ographi( 6aria#les an& Regulatory Restri(tions
:ear
%))(
*op
n
3rowth
Rate
,
3D* *er ;
capita
<
Reli"ion&
!a=orit#
*op
n
Life -usiness Re"ulator# Restrictions
*rice Re"ulation Investment Re"ulation Solvenc# 7orms
-an"ladesh *'++ ,/'/7 @uslim Not !no9n Stringent 8ea"ly im5osed
.hina *9R9 /'./ *00'*, Confucianism %55roval from C$EC Strict restrictions, confining the
areas of insurer investment to
ban" de5osits, Fovt' bonds,
financial bonds, etc'
;he solvency margin is a
function of insurerQs si:e of
business measured by
5remiums, claims, or both'
/on" 0on" *'*. 7,,.'0 Confucianism Self7regulatory No s5ecific guidelines' ,M of mathematical reserves
and /'0M of ca5ital at ris" or
R?! million
India *',0 -0+'0* ?indu %55roval from $ED% Stringent
Not less than -/M in a55roved
Securities
E3cess of the value of insurer
over its liability amount value
Indonesia *'0- )/.'*) @uslim Not !no9n ;otal invest' e3cluding
mortgage loans should be
eLual to technical reserves
*M of Premium Eeserves
!ala#sia *'+. ,+)'7 @uslim *)). $nsurance %ct &
the $nsurance
Eegulation has
5rovisions regarding
5ricing
0/M in #u'iputra Shares,
o9nershi5 share by ethnic
@alay
%ctual valuation of liability,
aggregate $nsurance
coverage, etc'
*a1istan ',* -..'/0 @uslim ;o follo9 SECP
Fuidelines
Statutory fund for each 5roduct
sold
Prescribes asset & liability
valuation methods
*hilippines *'7) */+,') Eoman
Catholic
%55roval from the
commission of the
5remium rates
PartialI varying %ggregate insurance
Coverage in force, subKect to
min' amount
Sin"apore *'0* ,*.0')* Buddhist %55roval from 5rinci5al
2fficers and directors
from $nsurance De5t'&
certificate from
a55ointed actuary
PartialI varying Dund solvency margin &
com5any solvency margin
Sri Lan1a /'+. ).*'.* Buddhist %55roval reLuired 0/M reserve funds in Fovt'
securities
;otal %ssets sufficiently S
;otal Aiabilities
Thailand /'+7 0--')) Buddhist %55roval reLuired $nvestments to be 9ithin
;hailand
M of reserve fund subKect to
min' amount
2ietnam *'/, -/*')) Buddhist %55roval reLuired Se5arate for foreign and
domestic com5anies
E3ist
Q &nnual rate of growt) in *ercentageG Y !n >%" at constant 2000 *ricesG %ource: Refer #ndnote !G World Bank 92006: 1"! C"7R2M 2006'
26
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
References
&rrow, 4' N' 91+<3:' @e role des valeurs ,oursiZres *our la re*artition la meilleure des
ris?uesM' Econom[trie'$ranslated as Kt)e Role of securities in t)e 2*timal
&llocation of Ris;7-earingM, Review of Economic Studies, Ool' 31, 1+65, ** +17
+6'
&rrow, 4' N' 91+6<:' @!nsurance, Ris; and Resource &llocation0 in 95oundations of
InsuranceEconomics:, G' "ionne and %' #' 3arrington 9eds':, 4luwer &cademic
Pu,lis)ers'
&svatana;ul, N' 92006: K$)ailand0s !nsurance !ndustryM, >ournal of Ris1 !ana"ement
and Insurance, Ool' 11G ** 1712'
-a,,el, "' =' 91+C<: K$)e Price #lasticity of "emand for 1)ole ife !nsuranceM,
>ournal of 5inance, Ool' 50G ** 22<723+'
-ec;, $' and !' 1e,, 92003: K#conomic, "emogra*)ic, and !nstitutional "eterminants of
ife !nsurance Consum*tion &cross CountriesM, ?orld -an1 Economic Review,
Ool' 17G ** <17CC'
-eenstoc;, M', G' "ic;inson and %' 4)aAuria 91+C6: K$)e "eterminants of ife
Premiums: &n !nternational Cross7%ection &nalysisM, Insurance& !athematics
and Economics, Ool' <G ** 2617270
-eenstoc;, M', G' "ic;inson and %' 4)aAuria 91+CC: K$)e Relations)i* ,etween
Pro*erty7ia,ility !nsurance Premiums and !ncome: &n !nternational &nalysisM,
>ournal of Ris1 and Insurance, Ool' <<G ** 2<+7272'
-ern)eim, -' "' 91++1: K3ow %trong are -e?uest Motives( #vidence ,ased on #stimates
of t)e "emand for ife !nsurance and &nnuitiesM, >ournal of *olitical Econom#,
Ool' ++G ** C++7+27'
-ern)eim, -' "', ' =orni, N' Go;)ale and ' N' 4otli;off 92003: K$)e Mismatc) ,etween
ife !nsurance 3oldings and =inancial Oulnera,ilities: #vidence from t)e 3ealt)
and Retirement %tudyM, American Economic Review, Ool' +3G ** 3<5736<'
-lac;, 4' Nr' and 3' "' %;i**er Nr' 92003:' @-ife and 0ealt; Insurance0' Pearson
#ducation, !ndia'
-lanc)et, "' 92007: KC)anging ife Cycles and t)eir !m*act on !nsuranceM, 3eneva
papers, Ool' 32G ** 1710'
-rowne, M' N' and 4' 4im 91++3: K&n !nternational &nalysis of ife !nsurance
"emandM, >ournal of Ris1 and Insurance, Ool' 60G ** 6167635'
-rowne, M' N', N71' C)ung and #' 1' =rees 92000: K!nternational Pro*erty7ia,ility
!nsurance Consum*tionM, >ournal of Ris1 and Insurance, Ool' 67G ** 737+0'
Cam*,ell, R' &' 91+C0: K$)e "emand for ife !nsurance: &n &**lication of t)e
#conomics of >ncertaintyM, >ournal of 5inance, Ool' 3<G ** 11<<71172'
Cargill, $' =' and $' #' $ro8el 91+7+: KModelling ife !nsurance %avings: %ome
Met)odological !ssuesM, >ournal of Ris1 and Insurance\ Ool' 56G ** 3+17510'
C)ang, B7M' and !' #rlic) 91+C<: K!nsurance, Protection from Ris;, and Ris; -earing,M
.anadian >ournal of Economics, Ool' 1CG ** <757<C6'
Cleeton, "' ' and -' -' Dellner 91++3: K!nsurance Ris; &version and t)e "emand for
!nsurance,M Southern Economic >ournal, Ool' 60G ** 15671<6'
"e,reu, G' 91+<3:' @>ne [conomie de ]certain0' Mim[o, ^lectricit[ de =rance' Cited in
@&n !ntroduction to !nsurance #conomics< in 9oundations of
27
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
InsuranceEconomics:, G' "ionne and %' #' 3arrington 9eds':, 4luwer &cademic
Pu,lis)ers'
"ic;ey, "' &' and 1' &' =uller 91+7+: K#stimators of &utoregressive $ime %eries wit) a
>nit RootM, >ournal of the American Statistical Association, Ool' 75, ** 5277
531'
#n/, R' 92000: K$)e %7curve Relations)i* -etween Per7Ca*ita !ncome and !nsurance
PenetrationM, 3eneva *apers on Ris1 and Insurance, Ool' 2<G ** 3+67506'
=is)er, %' 91+73: K& ife Cycle Model of ife !nsurance Purc)asesM, International
Economic Review, Ool' 15G ** 1327<2'
Greene, 1' 3' 92002: Econometric Anal#sis, 5
t)
#dition, Pearson #ducation &sia'
3a;ansson, .' 3' 91+6+: K2*timal !nvestment and Consum*tion %trategies >nder Ris;,
&n >ncertain ifetime, and !nsuranceM, International Economic Review, Ool' 10G
** 5537566'
3ammond, N' "', -' "' 3ouston and R' #' Melander 91+67: K"eterminants of 3ouse)old
ife !nsurance Premium #8*enditures: &n #m*irical !nvestigationM, >ournal of
Ris1 and !nsurance, Ool' 35G ** 3+7750C'
3ate;ar, .' and &' %ing) 92005: K"eterminants of ife !nsurance Consum*tion in !ndia:
& Cross7%tate &nalysisM, -ima@uest\Ool' 5G ** 3721'
3eaden, R' %' and N' =' ee 91+75: Kife !nsurance "emand and 3ouse)old Portfolio
-e)aviourM, >ournal of Ris1 and Insurance, Ool' 51G ** 6C<76+C'
3ussels, %', "' 1ard and R' Dur,ruegg 9200<: K%timulating t)e "emand for !nsuranceM,
Ris1 !ana"ement and Insurance Review, Ool' CG ** 2<7727C'
3wang, $' and -' Greenford 9200<: K& Cross %ection &nalysis of t)e "eterminants of
ife !nsurance Consum*tion in Mainland C)ina, 3ong 4ong, and $aiwanM, Ris1
!ana"ement and Insurance Review, Ool' CG ** 103712<'
3wang, $' and %' Gao 92003: K$)e "eterminants of "emand for ife !nsurance in an
#merging #conomy7$)e case of C)inaM, !ana"erial 5inance, Ool' 2+G ** C27+6'
4arni, #' and !' Dilc)a 91+C<: K>ncertain ifetime, Ris; &version and ife !nsuranceM,
Scandinavian Actuarial >ournal, ** 10+7123'
4wiatt;ows;i, "', P' C' -' P)illi*s, P' %c)midt and B' %)in 91++2: K$esting t)e .ull of
%tationarity against t)e &lternative of a >nit Root: 3ow sure are we t)at t)e
#conomic $ime %eries )ave a >nit Root(M >ournal of Econometrics, Ool' <5G **
1<+717C'
4won, 37%' and -' ' Nones 92006: K$)e !m*act of t)e "eterminants of Mortality on ife
!nsurance and &nnuitiesM, Insurance& !athematics and Economics, Ool' 3CG **
27172CC'
4won, 1' Nean 92001: K$oward =ree $rade in %ervices: $)e &%#&. !nsurance Mar;etM,
II5 8ccasional *aper, .o' 3G !nternational !nsurance =oundation, 1as)ington,
"C'
4won, 1' Nean 92002: K$)e !nsurance Mar;ets of %out) &siaM, II5 8ccasional *aper,
.o' 5G !nternational !nsurance =oundation, 1as)ington, "C'
4won, 1' Nean 92002: K$)e 1$2 and !nsurance in Greater C)ina: $)e Peo*les
Re*u,lic, 3ong 4ong, Macau, and $aiwanM, II5 8ccasional *aper, .o' 6G
!nternational !nsurance =oundation, 1as)ington, "C'
2C
&re ife !nsurance "emand "eterminants valid for %elected &sian #conomies and !ndia(
enten, im N' &' and "' .' Rulli 92006: K& $ime7%eries &nalysis of t)e "emand for
ife !nsurance Com*anies in &ustralia: &n >no,served Com*onents &**roac)M,
Australian >ournal of !ana"ement, Ool' 31G ** 51766'
ewis, ='"' 91+C+: K"e*endents and t)e "emand for ife !nsuranceM, American
Economic Review, Ool' 7+G ** 5<27566'
iedt;e, P'M' 92007: K1)at0s !nsurance to a Modern #conomy(M, 3eneva *apers, Ool'
32G ** 2117221'
im, C' C' and %' 3a,erman 92005: KModeling ife !nsurance "emand from a
Macroeconomic Pers*ective: $)e Malaysian CaseM, Researc) Pa*er: The A
th
International .on"ress on InsuranceB !athematics and Economics, Rome'
Mantis, G' and R' .' =armer 91+6C: K"emand for ife !nsurance,M >ournal of Ris1 and
Insurance, Ool' 3<G ** 25772<6'
Mossin, N' 91+6C: K&s*ects of Rational !nsurance Purc)asing,M >ournal of *olitical
Econom#, Ool' 7+G ** <<37<6C'
2utreville, N' =' 91++0: K$)e #conomic %ignificance of !nsurance Mar;ets in "evelo*ing
Countries,M >ournal of Ris1 and Insurance, Ool' 62G ** 5C775+C'
2utreville, N' =' 91++6: Kife !nsurance Mar;ets in "evelo*ing CountriesM, >ournal of
Ris1 and Insurance, Ool' 63G ** 263727C'
Pissarides, C' &' 91+C0: K$)e 1ealt)7&ge Relation wit) ife !nsuranceM, Economica,
Ool' 57G ** 5<175<7'
Rosenweig, M' 91+CC: KRis;, !m*licit Contracts and t)e =amily in Rural &reas of ow7
!ncome CountriesM, Economic >ournal, Ool' +CG ** 115C71170'
$ownsend, R' M' 91++<: KConsum*tion !nsurance: &n #valuation of Ris; -earing
%ystems in ow !ncome #conomiesM, >ournal of Economic *erspectives, Ool' +G
** C37102'
$ruett, "' -' and ' N' $ruett 91++0: K$)e "emand for ife !nsurance in Me8ico and t)e
>nited %tates: & Com*arative %tudyM, >ournal of Ris1 and Insurance\Ool' <7G
** 321732C'
1ard, "' and R' Dur,ruegg 92000: K"oes !nsurance Promote #conomic Growt)(
#vidence =rom 2#C" CountriesM, >ournal of Ris1 and Insurance, Ool' 67G **
5C+7<06'
1ard, "' and R' Dur,ruegg 92002: Kaw, Politics and ife !nsurance Consum*tion in
&siaM, 3eneva *apers on Ris1 and Insurance, Ool' 27G ** 3+<7512'
Baari, M' #' 91+65: K2n Consumer0s ifetime &llocation ProcessM, International
Economic Review, Ool' <G ** 3057317'
Baari, M' #' 91+6<: K>ncertain ifetime, ife !nsurance, and t)e $)eory of t)e
ConsumerM, Review of Economic Studies, Ool' 32G ** 13771<0'
D)u, M' 92002: K&n &nalysis of "emand for !ndividual ife !nsurance in C)inaM,
>ournal of Ris1 !ana"ement and Insurance, Ool' 7G ** 2C751'
Diet/, #' .' 92003: K&n #8amination of t)e "emand for ife !nsuranceM, Ris1
!ana"ement and Insurance Review, Ool' 6G ** 1<+71+1'
2+