Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine The Author

r 2010; all rights reserved. Advance Access publication 25 October 2010

Health Policy and Planning 2011;26:349356 doi:10.1093/heapol/czq070

Catastrophic and impoverishing effects of health expenditure: new evidence from the Western Balkans
Caryn Bredenkamp,1* Mariapia Mendola2 and Michele Gragnolati3
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, 1818 H Street NW, Washington DC 20433, USA. E-mail: cbredenkamp@worldbank.org

Accepted

5 August 2010 This paper investigates the effect of health-related expenditure on household welfare in Albania, Bosnia and Herzegovina, Montenegro, Serbia and Kosovo, all of which have undertaken major health sector reform. Two methodologies are used: (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 Measurement Surveys, 200005. While our analyses 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 householdsincreasing the incidence of poverty and pushing poor households into deeper povertyin each country. Both the catastrophic and the impoverishing effects of health expenditures are particularly severe 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.
Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

Keywords

Equity, financial protection, health expenditure, catastrophic, poverty

KEY MESSAGES
 Health expenditure contributes substantially to household impoverishment in Albania, Bosnia and Herzegovina, Montenegro, Serbia and Kosovo, increasing the incidence of poverty and pushing poor households into deeper poverty. Both the catastrophic and the impoverishing effects of health expenditures are particularly severe in Albania and Kosovo. Expenditures on transportation and on informal payments account for a large share of total health expenditures, especially in Albania. As these 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.

 

349

350

HEALTH POLICY AND PLANNING

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 seek to ensure that, in seeking care, households are protected from incurring health care expenditure that is so high that it adversely affects household 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 case in most developing countries (Roberts et al. 2004), the effect of health expenditure on household economic status can be severe, particularly among the poor. There is a growing interest in the effects of health expenditure on household 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 Zambia (Ekman 2007) and Uganda (Xu et al. 2006). A recent article, using survey data from 89 countries, found that 3% of households in low-income countries, 1.8% of households in middle-income countries and 0.6% of households in high-income countries incur catastrophic health expenditures (Xu et al. 2007).1 It is difficult to compare the findings of these studies because of, inter alia, variation in the comprehensiveness 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 household 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 section, we present and discuss our findings on inequalities in health expenditure and the effects of health care expenditure on household welfare; and in the final section, conclusions are drawn.

Description of country health systems

The countries under analysis have undergone significant transitions in the past decade or two, which have been complicated by a series of dramatic regional conflicts. After an initial phase focused 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 sector, a number of major health reforms have been undertaken in recent years aimed at improving access to comprehensive quality services, protecting vulnerable groups from the impoverishing effects of ill health and ensuring the systems fiscal sustainability (see Bredenkamp and Gragnolati 2008 for 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 because 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 financed 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. Albanias health system, by contrast, is based 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-based norms. Health insurance was only introduced in 1995 and 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 increase the efficiency of the hospital network. These 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 comprehensive, often including a range of non-essential services and even non-medical benefits (such as maternity leave and funeral expenses). 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 sector 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 expenditure on household wellbeing and poverty in these 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 2008 show that, in all five countries, household 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 database, http://www.who.int/nha/en/, accessed May 2010). The burden is

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

CATASTROPHIC AND IMPOVERISHING EFFECTS IN THE BALKANS

351

similar in Kosovo at 50% (World Bank 2008). Figures are 42%, 22% and 22% in Bosnia and Herzegovina, Montenegro and Serbia, respectively (WHO NHA database, http://www.who.int/ nha/en/, accessed May 2010). While these percentages constitute a much smaller share of total health expenditure than in many countries in Africa and Asia, the figures are far larger than in most of the 25 countries of the European Union.

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 expenditure includes official treatment fees, expenditure on medicines and laboratory expenses. 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 surrounding 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 disclose 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 seasonally. 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 14 months was used. Figures for Bosnia and Herzegovina were adjusted to reflect a 4-week period. Changes in total per capita expenditure are used to capture the extent of impoverishment. To obtain this measure, households were ranked by real total expenditure (including food, non-food, utilities and education expenditure, as well as the use value of durable goods owned by the household), adjusted for household size. Health expenditure is not included in the construction of quintile measures because health expenditure is considered to be non-discretionary. The concepts poor and non-poor refer to households below and above the respective National Poverty Lines close to the time of the surveys, calculated in local currency units (LCU) by the World Bank Poverty Assessment team. The national poverty lines used are 5145.33 new Lek per capita per month in Albania, 2223.146 KM per year in Bosnia and Herzegovina, 90.34 Euro per capita per month in Montenegro, 4111.31 dinars per capita per month in Serbia, and 106.689 DM per capita per month in Kosovo.

Methods
Methodological approach
In order to assess the effect of out-of-pocket health expenditure on household welfare, two methodologies are used: (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. In short, the analysis of catastrophic expenditure involves measuring the extent to which health costs incurred exceed different fractions of pre-payment household income. The second approach looks at the effect of health care expenditure on the incidence and depth of poverty, i.e. whether, and by how much, out-of-pocket payments push households below the poverty line. These two approaches capture two different aspects of financial protection. The first places the emphasis on the extent to which households are able to insure themselves against income loss due to health expenditures. Households that are relatively well-off may incur catastrophic health expenditures, but not become impoverished as a result of these expenditures. By contrast, the second methodology focuses purely on impoverishment. Households 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. These methodologies and their underlying assumptions and limitations are described in detail by Wagstaff (2008) and ODonnell et al. (2008).

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

Data
Data are drawn from recent household 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 observations on all variables, are 15 434 individuals in Albania, 2325 in Bosnia and Herzegovina, 8205 in Montenegro, 7871 in Serbia and 16 013 in Kosovo (Table 1). Throughout the analysis, sample weights are used to produce population estimates at the country level. Data

Table 1 Description of the data Country Albania Bosnia and Herzegovina Montenegro Serbia Kosovo Data source Living Standards Measurement Survey (LSMS) Living in Bosnia and Herzegovina Survey Institute for Strategic Studies and Prognoses (ISSP) Household Survey Living Standards Measurement Survey (LSMS) Living Standards Measurement Survey (LSMS) Year 2005 2004 2004 2003 2000 Sample size 15 434 2325 8205 7871 16 013

352

HEALTH POLICY AND PLANNING

Findings
Inequalities in health expenditure
The composition of health expenditure, in some countries, varies by socio-economic quintile. Table 2 shows that most of the health expenditure incurred by those who seek care can be categorized as treatment fees, expenditure on medicines and laboratory expenses, i.e. what we will refer to as general expenditure. Transportation costs and informal payments also represent 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 see that in Albania (which has the highest mean share of informal payments among the five countries), households 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 considerable share of total expenditure. In Table 3, we present health expenditure as a percentage of total gross expenditure, by quintile. On average, households in the bottom fifth of the distribution spend less in absolute terms, but more in percentage terms, on total health care (including transportation costs and informal payments) than households in the richest quintiles. In Albania the poorest spend about half the amount that the richest spend on health care, but this expenditure represents 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 expense represents 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 household expenditure devoted to health care is more similar across quintiles.

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

The incidence and intensity of catastrophic health care expenditure

Table 4 presents the incidence (headcount) and the intensity (mean gap) of catastrophic out-of-pocket payments. The headcount is the percentage of individuals whose health care costs, expressed as a proportion of income, exceeds a given discretionary 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.3 The sensitivity of the analyses 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% raises the proportion of the population with catastrophic payments to almost 21%, while the mean gap rises 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 these 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 observed there. Both the incidence and intensity is higher at lower thresholds and, in all cases, as thresholds increase, the MPG increases. Most of the increase in the MPG is due to a modest decline in the mean gap relative to the headcount as the threshold is raised. The interpretation is that the catastrophic effect of health costs manifests itself more as an increase in poverty incidence than a deepening of poverty among those who are already poor.

Table 2 Composition of out-of-pocket health expenditure, by quintile Quintiles Poorest (%) Albania General expenditure Informal expenditure Transportation expenditure Montenegro General expenditure Informal expenditure Transportation expenditure Serbia General expenditure Informal expenditure Transportation expenditure Kosovo General expenditure Informal expenditure Transportation expenditure 81 2 17 80 2 15 81 1 17 80 2 17 82 1 15 58 1 28 69 1 22 71 1 14 74 1 13 77 3 13 100 n.a. 0 99 n.a. 1 99 n.a. 1 97 n.a. 3 91 n.a. 9 87 8 6 88 6 7 91 5 4 92 5 3 92 4 2 2 (%) 3 (%) 4 (%) Richest (%)

Notes: General expenditure includes formal user fees, medicine and laboratory expenses. n.a. not applicable.

CATASTROPHIC AND IMPOVERISHING EFFECTS IN THE BALKANS

Table 3 Health expenditure as a percentage of total gross expenditure (among those who seek care), by quintile Quintile Poorest Albania General official expenditure (%) Informal expenditure (%) Transport expenditure (%) Total health expenditure (%) Health expenditure (monthly, pc) Total gross exp. (monthly, pc) Total net expenditure (excluding health) (pc) Bosnia and Herzegovina General official expenditure (%) Informal expenditure (%) Transport expenditure (%) Total health expenditure (%) Health expenditure (monthly, pc) Total gross exp. (monthly, pc) Total net expenditure (excluding health) (pc) Montenegro General official expenditure (%) Informal expenditure (%) Transport expenditure (%) Total health expenditure (%) Health expenditure (monthly, pc) Total gross exp. (monthly, pc) Total net expenditure (excluding health) (pc) Serbia General official expenditure (%) Informal expenditure (%) Transport expenditure (%) Total health expenditure (%) Health expenditure (monthly, pc) Total gross exp. (monthly, pc) Total net expenditure (excluding health) (pc) Kosovo General official expenditure (%) Informal expenditure (%) Transport expenditure (%) Total health expenditure (%) Health expenditure (monthly, pc) Total gross exp. (monthly, pc) Total net expenditure (excluding health) (pc)
Notes: pc per capita; n.a. not applicable.

353

2 6 1 0 7 665.99 7182.29 6516.30

3 6 1 0 7 737.28 9354.40 8617.12

4 5 0 0 5 748.23 12 171.27 11 423.04

Richest 4 0 0 4 939.80 20 008.06 19 068.27

Total 5 0 0 6 709.58 10 755.93 10 046.36

7 1 1 8 449.68 4708.04 4258.37

n.a. n.a. n.a. 2.3 4.16 157.99 153.83

n.a. n.a. n.a. 1.6 3.95 231.65 227.71

n.a. n.a. n.a. 1.6 5.07 301.82 296.75

n.a. n.a. n.a. 1.5 6.49 398.29 391.80

n.a. n.a. n.a. 1.2 7.71 643.05 635.35

n.a. n.a. n.a.

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

1.7 5.1992 315.9 310.7

0.8 n.a. 0.0 0.8 0.74 84.81 84.07

0.8 n.a. 0.0 0.8 1.08 131.33 130.24

1.2 n.a. 0.0 1.2 2.16 174.34 172.17 n.a.

1.2 n.a. 0.0 1.2 3.73 229.35 225.62

1.1 n.a. 0.0 1.1 4.72 398.28 393.56

1.0

0.0 1.1 2.81 225.69 222.87

3.8 0.03 0.62 4.4 216.99 3912.35 3695.35

3.9 0.01 0.57 4.4 350.19 6134.71 5784.52

4.3 0.02 0.36 4.6 483.55 8190.05 7706.50

2.8 0.02 0.28 3.1 372.16 10 508.48 10 136.33

3.3 0.07 0.18 3.6 703.26 17 548.36 16 845.10

3.6 0.03 0.41 4.1 417.33 9022.11 8604.78

11 0 2 13 12.14 63.47 51.34

8 0 1 9 10.14 92.59 82.46

6 0 1 7 10.7 120.42 109.71

5 0 1 6 10.09 157.77 147.69

3 0 0 4 11.21 272.66 261.45

7 0 1 8 10.88 141.71 130.83

While statistics for a particular country may reveal a fairly low average share of catastrophic expenditure, the distribution of those expenditures can be quite uneven across the population. Indeed, we find that mean out-of-pocket health expenditure (as a percentage of total household 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 expenditures is low, this effect is rather unevenly distributed (see Table 5).

354

HEALTH POLICY AND PLANNING

Table 6 Change in the incidence and depth of poverty as a result of out-of-pocket health expenditure
Bosnia and Albania Herzegovina Montenegro Serbia Kosovo Poverty headcount 5% 36.55 3.58 9.79 10% 20.79 2.19 10.53 15% 12.58 1.36 10.81 25% 5.12 0.52 10.16 1 Pre-payment headcount 2 Post-payment headcount 3 Percentage point change 4 Percentage change 13.40% 17.75% 16.20% 19.48% 2.80 20.90% 1.73 9.75% 7.20% 7.60% 0.40 5.59% 9.37% 40.86% 10.61% 47.12% 1.24 6.26

Table 4 Catastrophic impact of health expenditure, at various threshold levels

Catastrophic expenditure measures Health expenditure as a share of non-health expenditure per capita Albania Headcount (%) Mean gap (%) Mean positive gap (%) Threshold level z

Bosnia and Herzegovina

Headcount (%) Mean gap (%) Mean positive gap (%)

7.83 0.47 6.00

3.10 0.21 6.77

1.29 0.12 9.30

0.35 0.04 11.43

13.23% 15.32%

Poverty gaps Montenegro Headcount (%) Mean gap (%) Mean positive gap (%) 5.84 0.23 3.94 1.14 0.12 10.53 0.70 0.07 10.00 0.15 0.04 26.67 5 Pre-payment poverty gap 6 Post-payment poverty gap 7 Percentage point change 8 Percentage change 138.33 185.14 46.81 34% 83.16 92.03 8.87 11% 1.33 1.36 0.03 1% 76.75 91.85 15.10 20% 12.40 15.82 3.42 28%

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

Serbia

Headcount (%) Mean gap (%) Mean positive gap (%)

23.83 2.28 9.58

12.22 1.44 11.76

7.64 0.97 12.67

3.52 0.46 13.12

Kosovo

Headcount (%) Mean gap (%) Mean positive gap (%)

44.73 4.59 10.26

26.32 2.87 10.90

15.35 1.86 12.08

6.73 0.83 12.29

Table 5 Health expenditure, as share of total expenditure

Mean (%) Albania Bosnia and Herzegovina Kosovo Montenegro Serbia 6 2 8 1 4 Median (%) 3 0 4 0 1 Coeff. of variation 1.44 2.16 1.33 2.84 1.96

increases by 20%. In other countries, the impact of health expenditure on the poverty headcount is also substantial: health payments increase the incidence of poverty by 15% in Kosovo, 13% in Serbia, 10% in Bosnia and Herzegovina and 6% in Montenegro. Transition into poverty is not the only concern. The extent to which households that were already poor (in terms of their pre-payment income) were further impoverished by health care payments is also important. A comparison of the pre- and post-payment poverty gap shows that health expenditure increases the poverty gap by 28% in Kosovo, 20% in Serbia, 11% in Bosnia and Herzegovina and 1% in Montenegro.

Note: The coefficient of variation is equal to the standard deviation divided by the mean.

Discussion
Dramatic effects of health expenditure on poverty are observed in Albania where health expenditure is found to increase the poverty headcount by 21% and the poverty gap by 34%. Thirty-seven per cent of the population incurred health expenditures that exceeded 5% of non-health expenditure, and as many as 5% incurred expenditures in excess of 25% of non-health expenditure. Informal health expenditure is an important driver, accounting for between 4% (in the top quintile) and 8% (in the bottom quintile) of total health expenditure. The burden of health expenditures on the poor is particularly heavy: those in the bottom quintile spend 8% of their income on health compared to 4% in the top quintile. In trying to explain these results, it is notable that, after Kosovo, Albania is the country in the region in which health insurancecommonly assumed to provide financial protection from direct health expendituresis the least developed. Health insurance was only introduced in 1995 and, at the time of data collection, only 28% of the bottom quintile and 47% of the top quintile had a health insurance card (authors own

Effect of out-of-pocket expenditure on poverty measures

A limitation of the catastrophic expenditure approach is that it does not provide an indication of the extent to which the catastrophic payment causes hardship. For relatively well-off households, spending 25% of pre-payment income on health may not bring them anywhere near the poverty line, while for other households spending only a very small percentage of income on health may be impoverishing. The second approach looks explicitly at the effect of health care payments on the incidence and depth of poverty. It compares the headcount and poverty gap measures before and after expenditure on health care is taken into consideration. Table 6 shows that out-of-pocket health expenditure results in an increase in the percentage of poor Albanian households from 13% to 16%. In other words, the poverty headcount

CATASTROPHIC AND IMPOVERISHING EFFECTS IN THE BALKANS

355

analysis), due in part to the high unemployment rate (57%, by the expanded definition) and widespread informal employment (Arandarenko and Vukojevic 2008). It is in Kosovo that we find the largest economic disparities in health expenditures. While in absolute terms the bottom quintile and the top quintile spend the same on health care, expressed as a percentage of total expenditure the poor spend as much as 13% of their income on health whereas the rich spend around 4% of their income. On average, health expenditure results in a 15% increase in the poverty headcount and a 28% increase in the poverty gap. Seven per cent of the population incurred health expenditures in excess of 25% of non-health expenditure, while almost half of the population (45%) spent an amount equivalent to more than 5% of their non-health expenditure on health care. Part of the explanation may be that in Kosovo, there is currently no opportunity to pool risks through health insurance. It is also very expensive for households to overcome the physical barriers to accessing care: around 62% of the sample lives in rural areas and households in Kosovo spend a much larger share of their income on health-related transportation expenditure than elsewhere in the sub-region. However, since data are from 2000, we caution against too much reliance on these figures for current policy-making. Poverty measures in Serbia worsen substantially when health expenditure is taken into consideration, despite compulsory health insurance. Health expenditure raises the poverty headcount by 13% and the poverty gap by 20%. Twenty-four per cent of the population incurs health expenditure that exceeds 5% of non-health expenditure and 4% incur expenditures in excess of 25% of non-health expenditure. The poorest quintile spends more than three times what the wealthiest quintile spends on health care, but wealth inequalities mean that the poor and the rich devote approximately the same share of total expenditure to health care. Transportation expenditure is responsible for a very large share of total health expenditure (13% in the richest quintile and as much as 28% in the poorest quintile). Moreover, since the prospect of high transportation expenditure may discourage households, and especially poor households, from seeking care, it is likely that these figures are an underestimation of total health care expenditure and, thus, the magnitude of poverty. In Bosnia and Herzegovina, health expenditure increases the poverty headcount by 10% and the poverty gap by 11%. However, these figures are likely to understate the effect of health expenditure on poverty. First, this survey does not capture transportation expenditure. Yet, because 70% of the population lives in urban centres, it is unlikely that this expenditure will be as large as that incurred in Albania, Kosovo and Serbia. More significantly, the survey does not capture informal payments, which around a quarter of residents have been reported to pay (Lewis 2006). Households in Montenegro appear to be among the most protected. Health expenditure raises the poverty headcount by 6% and the poverty gap by 1%. In 2004, 6% of households incurred health expenditure in excess of 5% of non-health expenditure. For both poor and rich households, health expenditure constitutes a similar percentage of total expenditure. The fact that around 95% of the population had health insurance

at the time of data collection is likely to be a contributing factor. As in Bosnia and Herzegovina, one limitation of the Montenegrin estimates is that they survey did not capture informal payments, so these figures are likely to underestimate the true impact of health-related expenditure on poverty. More importantly, expenditure on inpatient services is not collected by the household survey. Yet, it is difficult to assess to what extent this latter omission results in an overestimation of financial protection because the cost of inpatient care is included in the fairly comprehensive benefit package to which almost all of the population has access. The interpretation of these findings needs to be tempered by the limitations of the methodologies employed. First, since health expenditure can only be incurred if sick individuals actually seek care, and those towards the lower end of the income distribution tend to face greater physical and financial obstacles to seeking care, we expect that, in general, the estimates generated by this analysis will underestimate of the true effect of health expenditure on poverty, creating the impression of a greater degree of financial protection than the system actually provides. Second, we are only measuring financial protection in the current period. Out-of-pocket payments in the current period may be financed by sources other than current income, such as dis-saving, borrowing and depletion of assets, which allows households to smooth non-health consumption in the period in which ill health occurs. While this coping mechanism may protect households from impoverishment in the short run, and result in estimates that (correctly) suggest adequate financial protection in the current period, eventually these expenditures will have to be financed. The replenishment of assets and the repayment of loans may impose substantial financial hardship in subsequent periods. Third, these results should not be interpreted as a ranking of the degree of financial protection offered by the health systems of the different countries in this analysis. This is because, across countries, there are differences in the types of health care expenditure on which information was collected, and it is likely that those countries for which the range of expenditure variables is more comprehensive will be observed as providing less financial protection. The results should be interpreted as country-specific analyses.

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

Conclusion
In this paper we used data from household surveys to examine the variation in out-of-pocket expenditures on health and their relationship to financial protection and impoverishment in Albania, Bosnia and Herzegovina, Montenegro, Serbia and Kosovo. All are in the process of fairly deep health system reform and the effect of health expenditures on poverty is likely to be of concern to policy makers. While our analyses are not without their limitations, there is no doubt that health expenditure contributes substantially to the impoverishment of households, increasing the incidence of poverty and pushing poor households into deeper poverty. Transportation expenditure accounts for a large share of total health expenditures, and contributes to impoverishment, especially in Albania and Serbia where transportation expenditure

356

HEALTH POLICY AND PLANNING

constitutes a greater share of total health expenditure among the poor than among the rich. Informal payments are significant in all countries, and are particularly high in Albania. As countries in the sub-region continue the process of health system reform, they may want to consider how to protect vulnerable groups from the impoverishing effects of health care expenditure. One step could be to revisit the user fee structureboth its design and implementationto consider different exemption criteria, the progressivity of co-payment schedules and the interaction between formal and informal payments. Countries could also more closely examine the constraints on the expansion of health insurance to uncovered groups, especially agricultural workers and the informally employed. This is especially critical in Albania, Serbia and Bosnia and Herzegovina where informal employment rates are 75%, 43% and 42%, respectively (Arandarenko and Vukojevic 2008), and because agriculture is a major employer in the subregion. Also, since transportation costs have been identified as a major component of health expenditure, policy makers could explore how to subsidize transport for the rural poor or, where feasible, try to achieve a more equitable geographic distribution of health care facilities. Finally, since this period remains one of rapid change in the region (including health sector reform, political change and economic growth), we would recommend repeated analysis of the impoverishing effects of health expenditure in the sub-region as new data become available.

Enhancing Efficiency and Equity: Challenges and Reform Opportunities Facing Health and Pension Systems in the Western Balkans. HNP Discussion Paper series. Washington, DC: World Bank. Baeza C, Packard T. 2005. Beyond Survival: Protecting Households from the Impoverishing Effects of Health Shocks in Latin America. Washington, DC: World Bank. Berman P, Ahuja R, Bhandari L. 2010. The impoverishing effect of health care payments in India: new methodology and findings. Economic and Political Weekly 45: 6571. Bredenkamp C, Gragnolati M. 2008. Sustainability of healthcare financing in the Western Balkans: an overview of progress and challenges. South East Europe Review 11: 15184. Ekman B. 2007. Catastrophic health payments and health insurance: some counterintuitive evidence from one low-income country. Health Policy 83: 30413. Lewis M. 2006. Governance and corruption in public health care systems. CGD Working Paper 78. Washington, DC: Center for Global Development. Limwattananon S, Tangcharoensathien V, Prakongsai P. 2007. Catastrophic and poverty impacts of health payments: results from national household surveys in Thailand. Bulletin of the World Health Organization 85: 6006. Lindelow M, Wagstaff A. 2005. Health shocks in China: are the poor and uninsured less protected? Policy Research Working Paper series #3740. Washington, DC: The World Bank. ODonnell O, Van Doorslaer E, Wagstaff A, Lindelow M. 2008. Analyzing Health Equity using Household Survey Data: A Guide to Techniques and their Implementation. Washington, DC: World Bank. Roberts M, Hsiao W, Berman P, Reich M. 2004. Getting Health Reform Right: A Guide to Improving Performance and Equity. Oxford: Oxford University Press. Van Doorslaer E, ODonnell O, Rannan-Eliya Catastrophic expenditures on health care Economics 16: 115984. R et al. 2007. in Asia. Health

Downloaded from http://heapol.oxfordjournals.org/ by guest on March 16, 2012

Disclaimer
The findings, interpretations and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank, its affiliated organizations or members of its Board of Executive Directors or the countries they represent.

Wagstaff A. 2005. The economic consequences of health shocks. World Bank Policy Research Working Paper WPS3644. Washington, DC: World Bank. Wagstaff A. 2008. Measuring financial protection in health. World Bank Policy Research Working Paper WPS4554. Washington DC: World Bank. Wagstaff A, Van Doorslaer E. 2003. Catastrophe and impoverishment in paying for health care: with applications to Vietnam 199398. Health Economics 12: 92134. World Bank. 2008. Kosovo Healthcare Financing Reform Study. Report No. 43183-XK. Human Development Sector Unit, Europe and Central Asia. Washington, DC: World Bank. World Bank. 2009. LSMS data table. Online at: http://go.worldbank.org/ PDHZFQZ6L0, accessed November 2009. Xu K, Evans D, Kadama P et al. 2006. Understanding the impact of eliminating user fees: utilization and catastrophic health expenditures in Uganda. Social Science and Medicine 62: 86676.

Endnotes
1

They consider catastrophic expenditure as having occurred when a household spends 40% of its capacity to pay (defined as total expenditure minus estimated food needs) on out-of-pocket health payments. The Former Yugoslav Republic of Macedonia is the only country of the former Yugoslavia that is excluded from our analysis. This is because its last LSMS-type household survey was conducted in 1996. The mean positive gap is defined as MPG G/H. Because this implies G H MPG, it means that the overall mean catastrophic gap equals the fraction of catastrophic payments times the mean positive gap.

References
Arandarenko M, Vukojevic V. 2008. Labor costs and labor taxes in the Western Balkans. In: Bredenkamp C, Gragnolati M, Ramljak V (eds).

Xu K, Evans D, Carrin G et al. 2007. Protecting households from catastrophic health spending. Health Affairs 26: 97283.