Catastrophic and impoverishing effects of health expenditure:new evidence from the Western Balkans
Caryn Bredenkamp,1*Mariapia Mendola 2and Michele Gragnolati 3
1
World Bank,Washington,DC,USA,2Department of Economics,University of Milan Bicocca,Milan,Italy and 3World Bank,Brasilia,Brazil
*Corresponding author.World Bank –HDNHE,1818H Street NW,Washington DC 20433,USA.E-mail:cbredenkamp@worldbank
Accepted 5August 2010
This paper investigates the effect of health-related expenditure on houhold welfare in Albania,Bosnia and Herzegovina,Montenegro,Serbia and Kosovo,all of which have undertaken major health ctor reform.Two methodologies are ud:(i)the incidence and intensity of ‘catastrophic’health care expenditure,and (ii)the effect of out-of-pocket payments on poverty headcount and poverty gap measures.Data are drawn from the most recent Living Standards and M
easurement Surveys,2000–05.While our analys are not without their limitations,and the lack of comparability across instruments precludes a direct comparison across countries,there is no doubt that health expenditure contributes substantially to the impoverishment of houholds—increasing the incidence of poverty and pushing poor houholds into deeper poverty—in each country.Both the catastrophic and the impoverishing effects of health expenditures are particularly vere in Albania and Kosovo.Transportation expenditure accounts for a large share of total health expenditures,especially in Albania and Serbia.Informal payments are substantial in all countries,and are particularly high in Albania.As countries in the sub-region continue the process of health system reform,an important policy question should be how to protect vulnerable groups from the catastrophic and impoverishing effects of health care expenditure.
Keywords Equity,financial protection,health expenditure,catastrophic,poverty
KEY MESSAGES
Health expenditure contributes substantially to houhold impoverishment in Albania,Bosnia and Herzegovina,Montenegro,Serbia and Kosovo,increasing the incidence of poverty and pushing poor houholds into deeper poverty. Both the catastrophic and the impoverishing effects of health expen
ditures are particularly vere in Albania and Kosovo. Expenditures on transportation and on informal payments account for a large share of total health expenditures,especially in Albania.
As the countries continue the process of health system reform,they would do well to consider how to protect vulnerable groups from the impoverishing effects of health care expenditure.
Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine ßThe Author 2010;all rights rerved.Advance Access publication 25October 2010Health Policy and Planning 2011;26:349–356
doi:10.1093/heapol/czq070
349
by guest on January 18, 2017
Introduction
The fundamental goal of a health care system is to ensure that its population has access to high quality care of the appropriate type in order to maintain and improve health status.At the same time,health systems should ek to ensure that,in eking care,houholds are protected from incurring health care expenditure that is so high that it adverly affects houhold economic wellbeing.This is often referred to as the ‘financial protection’goal of health systems.In countries where out-of-pocket expenditure is the most important source of health care financing,as is the ca in most developing countries (Roberts et al .2004),the effect of health expenditure on houhold economic status can be vere,particularly among the poor.
There is a growing interest in the effects of health expend-iture on houhold wellbeing in developing countries on all continents.One of the earliest studies looked at six Latin American countries,namely Argentina,Chile,Columbia,Ecuador,Honduras and Mexico (Baeza and Packard 2005).A number of empirical studies have also explored the catastrophic and/or impoverishing effect of health expenditure in East,Central and South Asia,including China (Lindelow and Wagstaff 2005),Thailand (Limwattananon 2007),India (Berman et al.forthcoming),Vietnam (Wagstaff and Van Doorslaer 2003),as well as Bangladesh,Nepal,Sri Lanka,Malaysia and the Kyrgyz Republic (Van Doorslaer et al.2007).Among the few African countries for which detailed studies are available are Z
ambia (Ekman 2007)and Uganda (Xu et al.2006).A recent article,using survey data from 89countries,found that 3%of houholds in low-income countries,1.8%of houholds in middle-income countries and 0.6%of houholds in high-income countries incur ‘catastrophic’health expend-itures (Xu et al.2007).1It is difficult to compare the findings of the studies becau of,inter alia ,variation in the comprehen-siveness of the types of health expenditure covered by surveys and the different methodologies employed to measure financial protection.Yet,they do leave the reader in no doubt that out-of-pocket health expenditures have important welfare implications.
We add to this literature by providing empirical evidence of the effect of out-of-pocket health expenditure on houhold welfare,and impoverishment,in five countries of South Eastern Europe,namely Albania,Bosnia and Herzegovina,Montenegro,Serbia and Kosovo.
The paper is organized as follows:we first provide a description of country health systems;we then describe the methodological approach,data and the measurement of key variables;in the next ction,we prent and discuss our findings on inequalities in health expenditure and the effects of health care expenditure on houhold welfare;and in the final ction,conclusions are drawn.
Description of country health systems
The countries under analysis have undergone significant tran-sitions in the past decade or two,which have been complicated by a ries of dramatic regional conflicts.After an initial pha focud on macro-economic stabilization and reconstruction,
reforms are now focusing on enhancing economic growth,employment generation and encouraging the containment and efficiency of public spending.The countries’shared aspiration to join the European Union (EU)exerts an important influence on policy decisions.In the health ctor,a number of major health reforms have been undertaken in recent years aimed at improving access to comprehensive quality rvices,protecting vulnerable groups from the impoverishing effects of ill health and ensuring the systems’fiscal sustainability (e Bredenkamp and Gragnolati 2008for an overview).
Bosnia and Herzegovina,Montenegro,Serbia and Kosovo,(together with Macedonia,Croatia and Slovenia)were part of the former Socialist Federal Republic of Yugoslavia (SFRY).Its health system,referred to as the Stampar model,was unique in Eastern Europe becau it was funded from compulsory social insurance contributions rather than the state budget.This financing mode persists in most of the new states and social health insurance is the dominant form of health financing in Serbia,Montenegro,and Bosnia and Herzegovina.In Kosovo,all health expenditure is currently financ
ed from the general budget and out-of-pocket expenditures,with some additional off-budget donor support,but a health insurance law is currently before Parliament.Albania’s health system,by con-trast,is bad on the former Soviet Semashko model and was historically funded directly from the central government budget,with central health allocations for different health inputs and for each health care institution made according to population-bad norms.Health insurance was only introduced in 1995and does not play as prominent a role in health financing as in the other countries.
With respect to health care delivery in the sub-region,the major reform elements have been the introduction of a new primary health care model that emphasizes family medicine,as well as hospital restructuring strategies to re-orient the delivery system towards preventive and primary care and increa the efficiency of the hospital network.The reforms have been accompanied by the rationalization of benefit packages.In the SFRY,access to health care was a constitutional entitlement of all citizens and benefit packages were exceptionally compre-hensive,often including a range of non-esntial rvices and even non-medical benefits (such as maternity leave and funeral expens).By contrast,Albania does not have the same legacy of generous health insurance-related entitlements and the benefit package is more limited.In general,the private ctor remains a relatively minor player in health care delivery,and also in health care financing.
A consideration of the effect of out-of-pocket health expend-iture on houhold wellbeing and poverty in the countries is timely and appropriate,not only due to the ongoing process of health system reform,but also due to the large share of total health expenditure that is in the form of out-of-pocket payments.The larger the share of health expenditure that is financed through out-of-pocket expenditures,the greater is the risk of impoverishment.Data from 2008show that,in all five countries,houhold out-of-pocket expenditure on health care constitutes a large component of total health expenditure.In Albania,out-of-pocket expenditure accounts for more than half of the total health expenditure,56%(WHO NHA databa,www.who.int/nha/en/,accesd May 2010).The burden is
350HEALTH POLICY AND PLANNING
by guest on January 18, 2017
similar in Kosovo at 50%(World Bank 2008).Figures are 42%,22%and 22%in Bosnia and Herzegovina,Montenegro and Serbia,respectively (WHO NHA databa,www.who.int/nha/en/,accesd May 2010).While the percentages consti-tute a mu
ch smaller share of total health expenditure than in many countries in Africa and Asia,the figures are far larger than in most of the 25countries of the European Union.
Methods
Methodological approach
In order to asss the effect of out-of-pocket health expenditure on houhold welfare,two methodologies are ud:(i)the incidence and intensity of ‘catastrophic’health care expend-iture,and (ii)the effect of out-of-pocket payments on poverty headcount and poverty gap measures.In short,the analysis of ‘catastrophic’expenditure involves measuring the extent to which health costs incurred exceed different fractions of pre-payment houhold income.The cond approach looks at the effect of health care expenditure on the incidence and depth of whether,and by how much,out-of-pocket payments push houholds below the poverty line.
The two approaches capture two different aspects of finan-cial protection.The first places the emphasis on the extent to which houholds are able to insure themlves against income loss due to health expenditures.Houholds that are relatively well-off may incur catastrophic health expenditures,but not become impoverished as a result of the expenditures.By contrast,the cond
methodology focus purely on impover-ishment.Houholds living on the brink of poverty may easily become impoverished by small health payments,even payments that may not be defined as catastrophic in terms of their share of total expenditure.The methodologies and their underlying assumptions and limitations are described in detail by Wagstaff (2008)and O’Donnell et al .(2008).
Data
Data are drawn from recent houhold surveys,either official Living Standards and Measurement Surveys (LSMS)or surveys that are considered equivalents.Data for Albania are from 2005,for Bosnia and Herzegovina from 2004,for Montenegro from 2004,for Serbia from 2003,and for Kosovo from 2000.Sample sizes,for the sample for which there are obrvations on all variables,are 15434individuals in Albania,2325in Bosnia and Herzegovina,8205in Montenegro,7871in Serbia and 16013in Kosovo (Table 1).Throughout the analysis,sample weights are ud to produce population estimates at the country level.Data
家庭聚会开场白can be downloaded from the World Bank website (World Bank 2009).
Measurement
The health modules of the different surveys vary somewhat,potentially introducing some measurement error into the analysis.There is heterogeneity in the categories of health expenditures included,with the most detailed information available for Albania and the least detailed for Montenegro.Data have been recoded to homogenize the categories of expenditure as much as possible.The term ‘general expend-iture’includes official treatment fees,expenditure on medicines and laboratory expens.Data on health-related transportation expenditure are available for four out of the five countries in the analysis.Bosnia and Herzegovina is the exception.Data on informal payments (in cash and in-kind)are available for Albania,Serbia and Kosovo,but not for Montenegro and Bosnia and Herzegovina.The measurement difficulties sur-rounding informal payments are well-known,including that patients may be misled into thinking that informal payments are part of the official cost of care and that patients may be reluctant to disclo to interviewers the payment of informal charges (Lewis 2006).
Another potential source of bias is that the surveys were conducted at different times of year,meaning that data are not strictly comparable since health expenditure may vary ason-ally.There is also heterogeneity in the recall period.Most survey questions refer to health-related events in the past four weeks,except in Bosnia and Herzegovina where a recall period of 14months was ud.Figures for Bosnia and Herzegovina were adjusted to reflect a 4-week period.
Changes in total per capita expenditure are ud to capture the extent of impoverishment.To obtain this measure,hou-holds were ranked by real total expenditure (including food,non-food,utilities and education expenditure,as well as the u value of durable goods owned by the houhold),adjusted for houhold size.Health expenditure is not included in the construction of quintile measures becau health expenditure is considered to be non-discretionary.The concepts ‘poor’and ‘non-poor’refer to houholds below and above the respective National Poverty Lines clo to the time of the surveys,calculated in local currency units (LCU)by the World Bank Poverty Asssment team.The national poverty lines ud are 5145.33new Lek per capita per month in Albania,2223.146KM per year in Bosnia and Herzegovina,90.34Euro per capita per month in Montenegro,4111.31dinars per capita per month in Serbia,and 106.689DM per capita per month in Kosovo.
Table 1Description of the data Country Data source
Year Sample size Albania
Living Standards Measurement Survey (LSMS)200515434Bosnia and Herzegovina Living in Bosnia and Herzegovina Survey
20042325Montenegro Institute for Strategic Studies and Prognos (ISSP)Houhold Survey 20048205Serbia Living Standards Measurement Survey (LSMS)20037871Kosovo
Living Standards Measurement Survey (LSMS)
2000
16013
CATASTROPHIC AND IMPOVERISHING EFFECTS IN THE BALKANS
351
by guest on January 18, 2017
Findings
Inequalities in health expenditure
The composition of health expenditure,in some countries,varies by socio-economic quintile.Table 2shows that most of the health expenditure incurred by tho who ek care can be categorized as treatment fees,expenditure on medicines and laboratory what we will refer to as ‘general expenditure’.Transportation costs and informal payments also reprent a relatively large share of total health expenditure,and often (but not always)especially among the poor.Examining this question on a country-by-country basis,we e that in Albania (which has the highest mean share of informal payments among the five countries),houholds at the poorest end of the income distribution pay,on average,8%of their total health expenditures in the form of informal payments compared with 4%in the richest quintile.In Serbia,on the other hand,the rich pay a slightly greater share of their health expenditure in informal payments than the poor do,but the share of health-related expenditure that the poor allocate to transportation expenditure is twice that which the rich do.Kosovo is the only place where the expenditure shares are more or less the same across quintiles.
Out-of-pocket health expenditure can account for a consid-erable share of total expenditure.In Table 3,we prent health expenditure as a percentage of total gross expenditure,by quintile.On average,houholds in the bottom fifth of the distribution spend less in absolute terms,but more in pe
rcent-age terms,on total health care (including transportation costs and informal payments)than houholds in the richest quintiles.In Albania the poorest spend about half the amount that the richest spend on health care,but this expenditure reprents twice the share of total expenditure.In Kosovo,as well,the highest burden of health expenditure is borne by the poorest quintile of the population:the poor spend about the same as the rich on health care,but this expen reprents 13%of their total expenditure compared with 4%for the richest.By contrast,in Bosnia and Herzegovina,Serbia and Montenegro,the poor spend much less than the rich for health care,but the share of total houhold expenditure devoted to health care is more similar across quintiles.
The incidence and intensity of ‘catastrophic’health care expenditure
Table 4prents the incidence (headcount )and the intensity (mean gap )of catastrophic out-of-pocket payments.The head-count is the percentage of individuals who health care costs,expresd as a proportion of income,exceeds a given discre-tionary fraction of their income,z .The mean gap is the average amount by which payments,as a proportion of income,exceed the threshold z .The incidence and intensity of the occurrence are related through the mean positive gap (MPG)which is defined as the gap over the headcount.3The nsitivity of the analys to different threshold levels is tested.
The table shows that in Albania,for instance,as much as 5%of the sample recorded out-of-pocket payments (as proportion of income)which exceeded 25%of non-health expenditure.The related mean gap measure is 0.5%,which means that,on average,health expenditure is 0.5%higher than the 25%threshold.Decreasing the threshold level to 10%rais the proportion of the population with catastrophic payments to almost 21%,while the mean gap ris to 2%.At the same threshold,in Kosovo the percentage of people spending more than the threshold for health care is around 26%,in Serbia 12%,in Bosnia and Herzegovina 3%and in Montenegro around 1%of the population.It is critical to realize,though,that the differences in the measurement of health expenditures across country surveys means that one cannot interpret the results as a ranking of the effectiveness of each of the country health systems in providing financial protection.The survey from Montenegro,for example,does not have information on inpatient care,which may explain the relatively high degree of financial protection obrved there.
Both the incidence and intensity is higher at lower thresholds and,in all cas,as thresholds increa,the MPG increas.Most of the increa in the MPG is due to a modest decline in the mean gap relative to the headcount as the threshold is raid.The interpretation is that the ‘catastrophic’effect of health costs manifests itlf more as an increa in poverty incidence than a deepening of poverty among tho who are already poor.
Table 2Composition of out-of-pocket health expenditure,by quintile
Quintiles Poorest (%)
2(%)
3(%)
4(%)
Richest (%)
Albania General
expenditure 8788919292Informal
expenditure 86554Transportation expenditure 6
7
4
3
2
Montenegro General
expenditure 10099999791Informal
expenditure a.a.n.a.Transportation expenditure 0
1
1
3
9
Serbia
General
expenditure 5869717477Informal
expenditure 11113Transportation expenditure 28
22
14
13
13
Kosovo General
expenditure 8180818082Informal
expenditure 22121Transportation expenditure
17
15
17
17
15
Notes :‘General expenditure’includes formal ur fees,medicine and a.¼not applicable.
352HEALTH POLICY AND PLANNING
by guest on January 18, 2017
pop设计While statistics for a particular country may reveal a fairly low average share of catastrophic expenditure,the distribu-tion of tho expenditures can be quite uneven across the population.Indeed,we find that mean out-of-pocket health expenditure (as a percentage of total houhold expenditure)substantially exceeds the median,producing large coefficients of variation.In Montenegro and in Bosnia and Herzegovina,in particular,while the catastrophic impact of health expend-itures is low,this effect is rather unevenly distributed (e Table 5).
Table 3Health expenditure as a percentage of total gross expenditure (among tho who ek care),by quintile
Quintile Poorest
山西自驾游2
3
4
Richest
Total
Albania
General official expenditure (%)766545Informal expenditure (%)111000Transport expenditure (%)100000Total health expenditure (%)877546Health expenditure (monthly,pc)449.68665.99737.28748.23939.80709.58Total gross exp.(monthly,pc)
冠族4708.047182.299354.4012171.2720008.0610755.93Total net expenditure (excluding health)(pc)4258.37
6516.30
8617.12
11423.04
19068.27
10046.36王者五排名字
Bosnia and Herzegovina General official expenditure (%)a.a.a.Informal expenditure (%)a.a.a.Transport expenditure (%)n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
Total health expenditure (%) 2.3 1.6 1.6 1.5 1.2 1.7Health expenditure (monthly,pc) 4.16 3.95 5.07 6.497.71 5.1992Total gross exp.(monthly,pc)
157.99231.65301.82398.29643.05315.9Total net expenditure (excluding health)(pc)153.83
227.71
296.75
胡影秋391.80
635.35
310.7
Montenegro
General official expenditure (%)0.80.8 1.2 1.2
1.1
1.0Informal expenditure (%)n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
Transport expenditure (%)0.00.00.00.00.00.0Total health expenditure (%)0.80.8 1.2 1.2 1.1 1.1Health expenditure (monthly,pc)0.74 1.08 2.16 3.73 4.72 2.81Total gross exp.(monthly,pc)
84.81131.33174.34229.35398.28225.69Total net expenditure (excluding health)(pc)84.07
130.24
172.17
225.62
393.56
222.87
Serbia
General official expenditure (%) 3.8 3.9 4.3 2.8 3.3 3.6Informal expenditure (%)0.030.010.020.020.070.03Transport expenditure (%)0.620.570.360.280.180.41Total health expenditure (%) 4.4 4.4 4.6 3.1 3.6 4.1Health expenditure (monthly,pc)216.99350.19483.55372.16703.26417.33Total gross exp.(monthly,pc)
3912.356134.718190.0510508.4817548.369022.11Total net expenditure (excluding health)(pc)3695.35
5784.52
7706.50
10136.33
16845.10
8604.78
Kosovo
General official expenditure (%)1186537Informal expenditure (%)000000Transport expenditure (%)211101Total health expenditure (%)1397648Health expenditure (monthly,pc)12.1410.1410.710.0911.2110.88Total gross exp.(monthly,pc)
63.4792.59120.42157.77272.66141.71Total net expenditure (excluding health)(pc)
讨论组51.34
82.46
109.71
147.69
261.45
130.83
Notes :pc ¼per capita;n.a.¼not applicable.
CATASTROPHIC AND IMPOVERISHING EFFECTS IN THE BALKANS
353
营养结构by guest on January 18, 2017
