Out-of-pocket and Catastrophic Health Expenditures Puzzle… · · 2014-07-04Out-of-pocket and Catastrophic Health Expenditures Puzzle: ... Out-of-pocket and Catastrophic Health

Out-of-pocket and Catastrophic Health Expenditures Puzzle:

The Costa Rican experienceMaría Paola Zúñiga-Brenes*, Juan Rafael Vargas**

Alberto Vindas***

* Professor in Economics, University of Costa Rica. Researcher Associate at Development Observatory and Central American Population Center.** Professor of Economics, University of Costa Rica and Research Associate at Central American Population Center*** Research assistant at Central American Population Center, University of Costa Rica

* Professor in Economics, University of Costa Rica. Researcher Associate at Development Observatory and Central American Population Center .

** Professor of Economics, University of Costa Rica and Research Associate at Central American Population Center

*** Research assistant at Central American Population Center, University of Costa Rica

Out-of-pocket and Catastrophic Health Expenditures Puzzle: The Costa Rican experience

María Paola Zúñiga-Brenes* Juan Rafael Vargas**

Alberto Vindas***

Abstract

This chapter addresses the paradox of why the out-of-pocket spending in Costa Rica is about 20% while catastrophic expenditures are small. Out-of-pocket expenses, the financial burden of out-pocket, catastrophic health expenditures and impoverishing expenditures are computed from the National Income and Expenditure Survey 2004. This opens the search for which are the factors that affect these variables. The results show that spending on medical consultation (where dental expenses are high) and expenditures on drugs are the main components of out- of-pocket spending. Catastrophic health expenditures may occur when families incur the expenses mentioned above and not in hospitalization outlays as it is the case at other LDC experiences. Although these expenses are very small, the results are very sensitive to the indicator used, and the definition of capacity to pay. The results on risk factors analysis show the conditions under which the health care services network is accessed and the role waiting lists have on out-of-pocket spending but not on the financial burden of out-of-pocket expenditures for households with catastrophic expenditures. Costa Rica could be viewed as a feasible target health system arrangement for countries of similar size and relative resource lacking, so it could lead to financial protection environments of the nature suggested by World Health Organization.

1. Introduction

The World Health Organization (WHO) recommends reducing the dependence of out-of-pocket

expenditures to fund the people’s own health care and that of their families. Because a funding

mechanism for the health systems has the purpose of raising funds to financially protect the

individual when he is facing a health event (risk pooling)) .

This chapter discusses the financial burden of out-of-pocket spending (HFB), out-of-pocket

expenses (OOPS), the catastrophic expenditure in health (CHE) and impoverishment expenditure

2

(IHE). Catastrophic health expenditures are defined as the proportion of out-of-pocket payments

that exceeds a certain percentage (threshold) of capacity to pay (CP). The impoverishing health

expenditures (IHE) are those that make the person extremely poor. To measure health

expenditures the National Income and Expenditure Survey (ENIG) 2004 is used. The capacity to

pay is defined as total expenditure less subsistence expenditure (SE), and this rises three criteria

depending on how SE is defined: a) SE is the household food expenditure (WHO, 2000), b) SE as

the poverty line, and c) SE as endogenous poverty line(Xu et al. 2003)1. Based on these

definitions six indicators are constructed: a) (WHO, 2000), b) Xu et al (2003), c) which is called

as Wagstaff et al, because of national poverty line (NPL) (WvD1, d) Wagstaff et al, WVD-2

because international poverty line (IPL) e) an hybrid with NPL (hybrid1)2, f) an hybrid with IPL

(hybrid2).

The contribution of this chapter is empirical as well as methodological one. From the first area of

knowledge, it combines composition analysis with progressiveness approach to help

understanding what lies behind HFB, OOPS and CHE in Costa Rica. It also explores the health

system characteristics as risk factors that may explain HFB and OOPS behaviour. From the

methodological point of view, the main contribution is discussing the robustness of results to

different definitions of affordability indicators, and its thresholds. Additionally, it has the novel

1 (Xu et al, 2003) use the average food expenditure of families, for those whose participation in food expenditure are in the percentile 45-55.I If food expenditure is less than this household food expenditure is used instead. 2 The hybrid is similar to WvD, indicator but it is replaced by total spending on food for the family when the CP <0, and SE is NPL.

3

feature of combining survey data with geographical data about access on health care services

networks and its associated waiting times for regression analysis.

The second section presents the data supporting this research. The next section briefly illustrates

why it is important to measure equity in health financing, CHE and IHE. It reports on some early

studies for Costa Rica. The fourth section sets out the composition of health spending. The fifth

part presents the results in terms of progressiveness of OOPS, HFB, CHE, IHE, along with the

robustness of results to the indicators, their definitions of affordability and the stated thresholds.

The sixth section discusses the possible risk factors that affect these variables (OOPS, HFB,

CHE), when the main hypothesis is OOPS are high, but not catastrophic due to the nature of the

national health system and the features of the waiting lists syndrome. ENIG data is

complemented with geographic information on the network access to health care services and its

associated waiting lists.

2. Health and expenditure data

The 2004 National Income and Expenditure Survey (ENIG), prepared by the National Statistics

and Censuses Institute (INEC), provides the quantitative basis for these study. Its main task is to

develop the basic consumer basket and its weights to construct the Consumer Price Index and the

basic needs basket to measure poverty. The survey is representative of the whole country and the

sample size is 4132 families. ENIG provides with no information on health status or on the

characteristics of the health system and that means a limitation for the analysis. Its design is

4

probabilistic for geographic areas, stratified (17 layers), and it follows a two-stage procedure. The

family is the unit of study for this survey and its capacity to pay is defined as total spending3

minus the subsistence expenditure, estimated on any of the above mentioned four poverty lines.

The expenditure ability includes all monetary expenditures. Furthermore, it excludes expenditure

in kind (either donated by another household or by an institution), but it includes imputed rental

value of owner-occupied housing, which it is often the most important asset for poor families and

for senior citizens households. The (Informe de la Persona Mayor 2008) shows families with

homeownership are less vulnerable to poverty. The imputed rental value of owner-occupied

housing, is also taken into account for the GDP definition on National Accounts. (Barrios 2005)

reinforces that finding.

Out-of-pocket spending on health, as mentioned above is the sum of payments for medical

consultation, medicines, hospitalization (at public or private institutions), laboratory tests,

therapeutic equipment such as prosthetics, eyeglasses, etc. It does not include private health

insurance. The summary statistics of the variables used for analysis are included in Table A1.1.

The analysis of determinants of HFB has three further limitations: i) it provides no information

on the health status variable, thus it possibly results in an omitted variable problem. ii) the

3 The World Bank recommends (Deaton, 1997) using expenditure rather than income because i) income is often under-reported, especially when a significant proportion of the workforce is self-employed and / or employers in the economy, ii) income has a seasonal component, and income approaches better the notion of permanent income, and iii) finally the use of cost is based on the micro approach of the monetary gain metric

5

insurance status could be considered endogenous, as the family can be assured to cope with high

health costs, iii) there is a very small percentage of families with CHE (at the 20% threshold).

The 42 alternative ways by which the families surveyed by the ENIG access the hospital network

were explicitly defined. Given its geographical location4 it yields dummy variables for the

regressions later on. Furthermore, data on the number of people on waiting lists was gathered.

This kind of information is available at the hospital and clinic basis for specialist medical

consultation, for some special procedures5 and for surgery as well. For these cases, it was not

possible to establish where the health expenditure was incurred, and the assumption adopted is

the person faces the public system specific institution where it is geographically ascribed;

therefore it has to deal with the relevant waiting list. Knowledge of this situation might lead the

person to incur into the out-of-pocket expense rather than waiting for the relevant place on the

queue. As, it will be discuss this seems to occur only for waiting list in medical procedures.

Matching survey with geographic data has the disadvantage that information is partial, as it only

includes information from the districts selected in the sample. There are four specific difficulties:

i) there are hospitals that have waiting lists in certain specialties or procedures, but those districts

were not selected in the ENIG sample, ii) no information for waiting list in medicines and

therapeutic devices is available, iii) despite institutional efforts made it has not been possible to

4 Only for some cases could not be established if the family belongs to a network or to other according to the district of residence. In the case of having to allocate waiting lists it was assigned the average of both networks 5 The list of procedures in 2004 included general ultrasound, mammography, gastroscopy, gynecologic ultrasound, electrocardiogram, etc.

6

detect the waiting list in all cases, especially those when the doctor tells the patient not to return

within a period of time, so there a place on the queue becomes available, iv) it is not feasible to

combine the length of the waiting list with the number of people listed. The behaviour of waiting

lists fluctuates. In 2004, the total waiting list was 103.604 while it exceeded 300,000 in 2000. By

2008 the total amounted to 295.621 cases.

3. Why is it important to measure OOPS along with CHE, IHE in Costa Rica?

The provision and financing of health care are key goals for health systems worldwide. Equity

financing is understood to mean each agent contributes according to its ability to pay while equal

access requires the individual receiving care according to its need. (Wagstaff & Van Doorslaer,

1993). This is so stated in order to separate the use of health care service from the capacity to

pay. A system that depends on out-of-pocket expenditures may be inequitable, because it may

exclude individuals from access to health services, therefore causing them a financial disaster or

to making them impoverished.

150 million persons worldwide have catastrophic health expenses and 100 million are

impoverished by having to face out-of-pocket health spending (Xu et al, 2005). WHO also notes

that expenditure is catastrophic when the prepaid mechanism is insufficient or the capacity to pay

is reduced. They also recommend to increase coverage of pre-payment mechanism, to design a

benefits package to protect the poor, and to determine an appropriate level of co-payments.

These measures appear to be logical to reduce the dependence of out-of-pocket spending, but in a

7

country like Costa Rica, where CHE is small, the questions then are: what does high OOPS

spending matters, which are the relevant mitigation measures, and what is its relationship with

catastrophic spending?.

Equity financing can be assessed through the progressiveness feature of the OOPS or by

measuring CHE and GE. (Zuniga-Brenes, 2008)6 measured progressiveness, using the (Davidson

and Duclos, 1997) approach. Other studies agree that CHE is relatively small in Costa Rica.

Zuniga (2002, 2006) estimated a version in financial health contribution index, and CHE using

data from the 1988 ENIG and from the 1992 Social Investment Survey (ENISO). (Xu et al.,

2003) presented results on CHE for Costa Rica in a multi-country and (Briceno et al., 2006)

calculated it using the ENIG 2004. Table 1 summarizes the results of CHE of the studies

mentioned above. They differ because: i) the polls are different (in 1988 and 2004 the survey

used is the ENIG, while in 1992 is a more aggregated type survey quality of life), ii) they use

different thresholds (30%, between 30% and 50%, 40% and over 50%, etc.), iii) uses different

measures of subsistence income, iv) different adjustments are made for consistency with national

health accounts, v) they apply different scales for family size. Finally, past performance, as well

as the (Zuniga-Brenes 2008) study shines the importance of taking into account how they are

made, measurements and definitions of capacity for purposes of comparative studies.

6 (Zuniga-Brenes, 2008) examines whether health funding is equitable from the point of view of progressivity, using the stochastic dominance approach posed by (Davidson and Duclos, 1997) to analyze progressivity in taxes.

8

Table 1

Summary of catastrophic health spending studies in Costa Rica

Study Indicator

CHE Año

Threshold Subsistence income or exp

Xu et al. (2003) 0,12 1992 40 Endogenous Zúñiga (2006) 0,12 1992 30 Household 0,13 1992 50

0,52 1988 50

1,72 1988 50

Food/exp (WHO)

0,79 2004 30 Briseño, Elizondo et al. (2006) 0,42 2004

40 Houshold Food /income

1,6 2004 30 NPL 0,73 2004 40

Source: authors’elaboration

4. Why is it important to study the Costa Rican health system and health expenditures?

Costa Rica is a low middle income country, but it is a high human development country; it

ranked number 54 in 2009 Index. United Nations (UN, 2009) ranks it as the second highest life

expectancy country in the Western Hemisphere, after Canada. Probably these achievements were

result of no army since 1948 and a 20% of gross domestic product (GDP) being devoted to social

programs, with public expenditure on health amounting around 6% of GDP. In Costa Rica the

prepaid mechanism coverage is high, there is a protection system for the poor (non-contributory)

and there is no co-payments in the public insurance system, with all prescriptions granted with no

patient’s payment. In a parallel chapter the organization of the health system is described.

Figure 1 shows the composition of OOPS, which integrates private expenditure on medicines,

doctor visits, hospitalization, surgery and laboratory tests, etc.

9

Figure 1

Total health expenditure and out-of-pocket expenditure breakdown in Costa Rica

Source: Author’s calculation from ENIG 2004

In 2004 about 84% of the private expenditure is made up of private medical consultation and

medicines, but the private hospital expenditure amounts to less than 0.37%. The right hand side

graph shows the composition of total health expenditures, around 70% is financed through social

health insurance.

5. Out-of-pocket, progressivity of out of pocket, catastrophic expenditures, and

impoverishment

This section is divided into three parts: the first presents an analysis of progressivity on

OOPS by type, the second the results of CHE and the third IHE. The results of CHE are

presented for the six indicators mentioned and the unit of analysis is the family.

10

5.1. Progressivity of out-of-pocket expenditures

One way to measure equity is through the progressivity curve based on (Davidson and Duclos,

1997), which shows the difference between the Lorenz curve for capacity to pay and the

concentration curve for health payments. When the difference between them is positive, it means

that the lower p percent (p%) of the population have a greater cumulative share of capacity to

pay than cumulative share of health expenditures in which they incur. For example if the p% of

population at the bottom part of the distribution have 10% of its whole distribution of capacity to

pay (cumulative share), and 8% of its whole distribution of health expenditures, progressivity is

of 2%, because people are contributing 2% less relatively to its cumulative share of capacity to

pay. The results for two measures of capacity to pay are presented in Figure 2.

Figure 2

Progressivity of out-of-pocket expenditures

Source: Authors’ calculation from ENIG 2004

11

Results highlight the progressive pattern of OOPS especially for medium capacity to pay deciles.

Families between the 40% and 80%, are contributing with health expenditures less than

proportional while for top deciles (up to 85%), households are contributing more than

proportionately to health expenditures7. For the bottom deciles is slightly proportional, with some

negative differences. This measure is different from the average of the ratio of health payments

relative to capacity to pay because it takes into account what happens in the entire cumulative

distribution, not just at the specific decile rank. Progressivity for OOPS in drugs, therapeutic

devices, medical, dental and laboratory tests is shown in Figure 3.

Figure 3

Progressiveness curve of out-of-pocket spending by type

Source: Authors´ calculation from ENIG 2004

7 These results are consistent for several measures of ability to pay. The only difference is when the total spending of out-of-pocket spending is progressive for about 90% of distribution

12

5.2. Catastrophic health expenditures

Catastrophic spending is defined as the consumer's out-of-pocket expense when it exceeds a

given threshold, namely k% of his CP. The percentage of households reporting catastrophic

expenditure is sensitive to: i) the definition of capacity to pay; ii) how expenditure is measured,

and iii) the indicator of catastrophic expenditures used. Figure 4 presents the proportion of

households with CHE for the three thresholds and for each of the six indicators mentioned in

section 1. The CHE is relatively small, varying between 1.56% and 0.31% for a threshold

k=30%. This is an outstanding feature of Costa Rica, as it was first shown by (Xu et al., 1993).

Figure 4 Costa Rica: Catastrophic Health Expenditures

Source: Authors’ calculation from ENIG 2004.

At the 30% threshold, the percentage of households with catastrophic expenditure is higher for

WvD1 poverty line indicator (1.56), and it is lower when the (Xu et al, 1993) (0.31) is used. The

13

later is slightly lower than the WHO (0.43) procedure calculation. The results are shown in Table

A1.3. Those results are due to the way WvD1, and WvD2 treats all poor people with catastrophic

expenditures, not allowing (as the other approaches do) for the family to reduce their subsistence

income to cope with health costs, namely, to reduce the expense of subsistence food expenditure.

These results are consistent for a 20% threshold. For 40% threshold there are just too few

observations to yield a reliable measure. The national poverty line is almost double the

international line. Therefore, WvD2 which use IPL tends to be closer to the rest of the

calculations that includes food adjustment, although it does not consider the possible reduction

in the subsistence expenditures to smooth spending on health. A person may have CHE not

because his OOPS are high in absolute terms, but because his CP is very small. Therefore,

people with low CP may have higher incidence of CHE, however as it will be discuss results are

sensitive to how each indicator treats people with very low CP.

Figure 5 shows the percentage of households with CHE to be lower in the third and fourth

quintile, and to be greater for the first quintile for WvD1, WvD2, Hybrid1, Hybrid2. That does

not happen for the WHO or by Xu indicators, where CHE is highest for the fifth quintile. The

difference in these results is due to the use of an absolute or a relative poverty line. However, for

the lowest quintile, the proportion of households with catastrophic expenditure is important for

all indicators.

14

Figure 5 Costa Rica: Percentage of households with CHE by expenditure quintile

Source: Authors’ calculation from ENIG

By using the (Xu et al., 2003) indicator, the percentage of households with catastrophic

expenditure (at the 30% threshold) is 0.31%. However, for the last quintile of income of

households, it is higher (0.5%) than in the first quintile (0.2%). The frequency of households with

catastrophic expenditure is very low, so the robustness of the results is further reduced as the

analysis is broken down by quintiles or by family composition. Clearly there is a compromise

between specificity of the results or their relevance and the statistical soundness.

Costa Rica has a long standing public policy of increasing the coverage of the health system and

it has enacted measures to protect the poor (people in the first quintile). While it had the non

intended consequence of reducing the importance of OOPS and the possibility of experiencing

15

CHE, it has notably moved ahead in both directions. Figure 6 shows CHE by insurance

condition.

Figure 6

Costa Rica: % of CHE at least one member with and without health insurance


.

Nonetheless, for the 30% threshold, the percentage of households with catastrophic spending is

zero for some indicators when no member is insured, but it is positive for families with at least

one insured member. This may suggest there is no CHE for the uninsured because they do not

use the system. That features a conceptual problem with the catastrophic health expenditure

analysis. However, this seems not to be the case of Costa Rica. At least 90% of the households

have a family member insured, and 96% of them reported using health services. Yet, despite of

that, 35% of households reported no health expenditure budgets. According to the 2006 ENSA

16

only 3% of the population report to need a medical consultation in EBAIS but did not receive

care. From those 3%, around 70% reported that the reason was that he couldn’t get the

appointment. In addition, CHE could be low because National Health System has no co-

payments and health care is a universal right in Costa Rica, so no one is denied first time or

emergency care, even if they are not insured. That is so, even though there will be a due

procedure to try to collect, but clearly the payment is contingent.

Other household feature that makes them experience catastrophic expenditures is having

dependent members within. The reason is those families require special health care and they are

more likely to face higher costs. Figure 7 shows the results.

Figure 7

Costa Rica: Percentage of households with CHE by family composition


17

Figure 7 shows families with no older adults or children have less CHE and households with

older adults have a relatively larger proportion of catastrophic expenditures, except by WvD1,

WvD2 (K30, k40) Those are also followed by households with children in the household.

However, only two extreme values reach 5% of the relevant income and one of them is under

WvD1.

Figure 8 shows the average composition of out-of-pocket spending for household with and

without CHE.

Figure 8 Composition of pocket spending and spending without catastrophic

Source: author’s calculations out of 2004 ENIG

18

Spending on hospitalization care is likely to be a major source of catastrophic expenditures. It is

uncommon, it is unexpected and it usually can not be postponed. However, it does not seem to

be the case of the Costa Rican households. For families with CHE figure 8 shows: i) medical

visits have a greater role in the OOPS, for the (six) indicators, ii) expenditure on hospitalization

(be that public or private and even if it involves surgery) representing 0% of out-of-pocket

expenditures of households with CHE at the 30% threshold. Interestingly, for families without

CHE: i) drugs are a relatively very important component of out-of-pocket spending, ii)

expenditure on therapeutic devices and laboratory tests is a sizeable part of out-of-pocket

spending, and the first item expense is quite regressive. The latter result is also found in the

estimation approach. However, it is important to keep in mind the small sample size of CHE.

So far the discussion has dealt with robustness of the results, with the indicator used, and with the

threshold as well. However, the results can differ according to the capacity to pay definition.

The above results compared the 6 capacity to pay measures: all include various spending levels

minus subsistence income. Table A1.3 presents a summary of those results at the 30% threshold

The results may differ if income instead of expenditures is used or how ability to pay is defined.

Table A1.3 shows the use income instead of expenditure rising the percentage of households with

catastrophic expenditure; it is slightly higher for all cases. For instance, the WvD1 indicator with

LNP (WvD1) for the 30% and 20% thresholds amount to 3.25% and 1.83% of the households,

respectively. Zuniga-Brenes (2006) also found that for households in the lower income decile

19

reported expenditures higher than their income, in contrast to what the higher income deciles

families did, for that reason it is possible the CHE are less when using expenditures as CP.

The inclusion or exclusion imputed rental value of owner-occupied housing has an important

effect on the measurement of CHE greater than that coming from the alternative use of income or

expense. Table 3 shows differences in the maximum range of variation are greater the higher the

threshold. For instance, at the 20% threshold, the inclusion of imputed income arising from

housing usage shifts the number of households from 3.25% to 6.07%.

Table 3 Extreme values of the indicator of catastrophic expenditures

CP Expenditure-SE Income- SE Expenditure-SE Income- SE

with no imputed rental value

with no imputed rental value

with imputed rental value

with imputed rental value

Threshold 20 30 40 20 30 40 20 30 40 20 30 40 Max value 5,78 3,58 2,76 6,07 3,66 2,82 2,99 1,56 1,18 3,25 1,84 1,3

Min value 2,76 1,14 0,56 3,48 1,97 1,04 1,28 0,31 0,05 1,88 0,65 0,33

Source: authors’´ calculation from ENIG 2004

These results show the relationship between CHE and some other variables. However, it is

necessary taking into account the correlation that may exist among all the variables, such is the

case of the proportion of families with at least one insured member may increased along the

spending quintile distribution, or with the size of the family. That is why trying to explain what

factors that influence CHE calls for multiple regression analysis.

20

5.3. Impoverishing health expenditures (IHE)

Total expenditures including imputed rental value of owner-occupied housing, imputed rental

value of owner-occupied housing, for the ENIG 2004 yields 3.6% of families in extreme poverty.

CHE results show that only 0.14% of families became impoverished because of the OOPS when

the national poverty line is used. That figure is 0.10% for the international poverty line. IHE

appears only in the first quintile with reading of 0,007 (LNP) and 0,005 (LPI). Yet not much

could be said because there are very few observations (6 families).8 The exclusion of imputed

income from home ownership increases the percentage of households who reported IHE to 0.40%

Figure 9

Percentage of households with Impoverishment by denominator

Source: Authors´ calculation from 2004 ENIG

8 This is not a robustness feature of the Survey, it is a result of a quasi-universal national health system

21

6. Analysis of determinants of out-of-pocket expenditures, financial burden of

private health and catastrophic expenditures

6.1. Network access to health care services and waiting lists

The Costa Rican Health System is organized into three disjoint services networks lead by one of

the main hospitals. Each network organizes services into three levels of complexity with the

Integrated Basic Health Services Delivery Centers (EBAIS) acting as the primary, neighborhood

site. There are almost 1000 centers each one of them meant to take care of 1000 families

throughout the land and they are designed to be the gates to the system. (Muiser and Vargas,

2010) provides more details on the working of the networks.

Waiting lists happen because at a specific health service there are more patients than service

delivery instances. They arise out of the interaction of supply factors (the provision of resources

and the efficiency of their use) and demand factors (arising from a complex conjunction of

preferences and protocols of patients and physicians). They are endemic of public health services

but they could occur in private or non health care services, too. The problem has been managed

with different levels of priority and it is usually a program within the office of the Vice President

for Medical Affairs at CCSS.

22

In 2004, 47% of the waiting lists were credited to specialized medical consultation, while 43%

were therapeutic procedures and 10% were surgeries. 28% of the waiting list goes to

ophthalmology care while orthopedic amounts for 14%. Orthopedic surgery had the largest queu

(20%) among elective surgeries. For medical procedures the highest numbers were ultrasound

procedures (31%) and X-ray (20%). A single national hospital (H. Mexico) accounted for 20%

of the waiting lists.

The length of waiting lists data could not be cross tabulated with the number of visits on the lists.

Yet, partial evidence showed ophthalmology services at Mexico Hospital had a waiting time of

more than 360 days, and the same was true for general surgery, urology, and peripheral vascular

care at San Juan de Dios Hospital. For the regional hospitals 5 out of 7 had waiting lists longer

than 1 year, one of the hospitals had them in 6 specialties. For health network ophthalmology has

a waiting lists longer than 360 days, and Children's Hospital has waiting lists of over a year in

dentistry, reconstructive surgery, orthopedics and urology. All in all, the waiting lists are a major

problem and they could induce out-of-pocket expenditure if households could afford it.

6.2. Health financial burden, out of pocket and catastrophic and catastrophic

expenditures

This section presents the results for risk factors analysis related to i) OOPS, ii) (Health Financial

Burden) HFB for those households with and without CHE. CHE and HFB are seen from a

23

quintile regression perspective, which analyzes the financial health burden of the 98th percentile

and 99th percentile instead of using probit or logit estimation. That is due to very few families

experiencing CHE. Around the 98th percentile of the HFB, Costa Rican families experience a

OOPs over 20% of the CP (considering a 20% catastrophic threshold). Results are also presented

for the 60, 80, 90 percentiles for analyzing the effects of the variables for the different segments

of the HFB distribution. In addition HFB and OOPs are estimated by a tobit procedure (with

standard errors are corrected for clusters). All estimates are conducted in two stages: first of all,

the probability that at least one family member is insured is obtained and then, the second stage

proceeds with the estimates described above.

(Hart, 2006) recommends to include a Heckman adjustment to correct for selection bias when

the probability of having CHE is estimated, because conditioned on health status people first

choose whether to use and pay of health services if they need it, in order to account for those who

need and did not use the health care services. However, in the Costa Rican ENIG household

survey has no information about health status (or need), but 96% uses health care services,

although 35% did not spend in OOPS something that makes that procedure less relevant.

To explain risk factors that may contribute to HFB for families with and without CHE, and

OOPS, specified socioeconomic characteristics were analyzed. It is expected when households

face an adverse health event, those located at the lower quintiles tend to spend less on private

health care, because its capacity to pay is lower. Families without insurance coverage are likely

24

to spend more out of their pockets. Families with older adults are bound to spend more on health.

Other variables that are incorporated in the estimation analysis are: place of residence and family

size (scale effect), years of schooling for the head and the living conditions of the family, having

access to basic conditions such as electricity, drinking water, sanitation and similar. The share of

health spending that was not financed using cash is included in the equations to somehow capture

the room the family has to soften the health loss shock.

Several alternative estimations were performed for the HFB. They included i)only

socioeconomic household characteristics based on survey information and ii) household

information matched with geographic records. In the latter case, the explanatory variables

included either: a) several dichotomous variables specifying how families access the health

services network were added, or b) waiting list information: three dummies are used to identify

the waiting lists in procedures, surgeries and specialist medical consultation. Waiting lists are

also related to geographic information given the referral system.

Table 4 presents the quintile regression for HFB including dichotomous variables for health

services network given different percentiles using (Xu et al, 2003) definition of CP. In the

appendix results are presented for others measures of CP. Adjusted Pseudo R2 shows a better

adjustment when health services network is included instead of waiting list information, or when

no information about health care appears.

25

For people with catastrophic health expenditures, those at percentile 99%. Being in second to

fifth quintiles decrease its health financial risk. Therefore, people in first quintile tends to have

higher HFB. This result seems contradictory with the per quintile analysis from the previous

section. However correlation between different variables is brought into the equation and, as it

will be discuss below, results are only different when household food is included as subsistence

income. Having elderly people at home increases HFB as it is expected, because older people

use more often health services and need treatments that are relatively more expensive. Living in

urban areas, having children, having a house with electric energy, having access to improved

sanitary condition show increased HFB, but the coefficients are not significant. Having access

to improved water or to waste disposal affects HFB negatively, but they are not significantly.

Years of schooling is not significant either, and the share of health expenditures that is not

financed with cash has negative impact, but it is not significant.

Some health access networks decrease the HFB. Such cases are: i) Calderón Guardia-Clorito

Picado, ii) Calderón Guardia/México- Clorito Picado-Jiménez Núñez, iii) México-San Carlos-

Los Chiles, and iv) México/San Juan de Dios-San Rafael Alajuela. The rest of dichotomous

geographical variables are not significant and they alternate considerably of algebraic sign.

There are only two health networks that show households at 98% percentile increasing HFB and

they are: San Juan de Dios-Tomas de las Casas and San Juan de Dios-Escalante-Pradilla.

26

Table 4:

Quintile regression for HFB with health network coverage

With Endogenous poverty line 99 98 90 80 60

Urban area 0.182 -0.018 0.098 0.094 0,128* Quintile 2 -0,591** -0,441*** -0,296** -0.149 0,492*** Quintile 3 -0,86*** -0,564*** -0,272* -0.128 0,607*** Quintile 4 -0,755*** -0,586*** -0,345** -0.046 0,89*** Quintile 5 -0,598** -0,461** -0,293* 0.049 1,073*** hhd with children 0.044 -0.013 0.013 0,132** 0,137*** hhd with elderly 0,589** 0,627*** 0,72*** 0,701*** 0,431*** hhd with elderly and children 0.206 -0.072 0.15 0.096 0.05 hhd size -0.162 -0.137 -0.057 -0.044 0,151*** Prob of having insurance -1.091 -1.883 -2,239** -1.056 -0,974* hhd with electrical energy 0.664 0.605 0.358 -0.031 0.097 hhd with access to improved sanitary conditions 0.204 0,564* 0,388* 0.293 0,298** hhd with access to improved water -0.113 -0.118 -0.156 -0.127 -0.014 hhd with access to waste disposal -0.298 -0.219 -0.066 -0.121 -0.1 years of schooling (head) -0.004 0.001 0.007 0.011 0.018 share of household expenses in health not in cash -0.185 0.014 -0.101 -0,264* -0,151* Calderón Guardia-Catedral Noreste 0.131 0.536 0,459** 0.198 0.224 Calderón Guardia-Coronado 0.361 0.688 0.269 0.11 0.169 Calderón Guardia-Clorito Picado -1,097** -0.721 -0.327 -0.534 -0.119 Calderón Guardia-Jiménez Núñez 0.157 0.693 0.301 0.124 -0.1 Calderón Guardia-Max Peralta 0.033 0.315 0.205 0.049 0.17 Calderón Guardia-Max Peralta-Catedral Noroeste -0.196 0.115 0.36 0.16 -0.042 Calderón Guardia-Max Peralta-Marcial Fallas 0.159 0.609 0,571* 0,543** 0.455 Calderón Guardia-William Allen 0.155 0.351 0.169 0.016 0.019 Calderón Guardia-Tony Facio -0.115 0.217 0.005 -0.288 -0.123 Calderón Guardia-Tony Facio-Guápiles 0.12 0.462 0.376 0.209 -0.022 Calderón Guardia/México-Clorito Picado-JN -1,452*** -1,023*** -0,628*** -0.458 -0.438 Calderón Guardia/México-Jiménez Núñez 0.159 0.717 0.446 0.308 0.322 México -0.295 0.276 0.41 0.248 0.03 México-Clorito Picado 0.704 -0.13 0.238 0.041 -0.13 México-San Vicente 0.46 0.357 0.165 0.033 0.111 México-Baltodano -0.06 0.278 0.146 -0.086 0.002 México-Baltodano-Anexión 0.234 0.418 -0.161 -0.389 -0.139 México-Baltodano-Upala 0.097 0.679 -0.138 -0.351 -0.113 México-Mons. Sanabria 0.097 0.407 0.154 -0.164 0.004 México-Mons. Sanabria-Max Terán -0.304 -0.092 0.435 0.479 0,671* México-San Carlos 0.049 0.271 0.058 -0.072 -0.036 México-San Carlos-Los Chiles -1,887*** -1,523*** -0,878* -1,3*** -0.215 México-San Carlos-San Vicente 0.192 0.533 0.736 -0.02 0.219 México-San Carlos/San Rafael Alajuela-LCV -0.441 0.099 0.7 0.412 -0.28 México-San Rafael Alajuela-Marcial Rodríguez 0.324 0.181 0.249 0.199 0.287 México-San Rafael Alajuela-Luis C. Valverde 0.799 0.745 0,624*** 0,48** 0,747*** México-San Rafael Alajuela-Guápiles/Los Chiles -0.896 -0.215 0.386 0.589 1,176** México/San Juan de Dios-SRafael Alajuela-SFA -0.017 0.27 0,602** 0,453** 0,487* México/San Juan de Dios-Mons. Sanabria -1,334*** -0,801* -0.232 0.307 -0.015

27

Continuation

San Juan de Dios 0.724 0.421 0.296 -0.003 0.061 San Juan de Dios-Carlos Durán -0.355 0.212 0.677 0.112 -0.158 San Juan de Dios-Clorito Picado -0.505 -0.058 0.357 -0.82 -0.234 San Juan de Dios-Marcial Fallas 0.338 0.576 0.168 0.039 -0.036 San Juan de Dios-Moreno Cañas 0.212 0.493 0.154 0.049 0.124 San Juan de Dios-Solón Núñez 0.024 0.356 0.171 0.055 -0.087 San Juan de Dios-Escalante Pradilla 0.277 0,649* 0.193 0.104 0.021 San Juan de Dios-Escalante Pradilla-Ciudad Neily 0.71 0.221 0.141 -0.017 0.04 San Juan de Dios-Escalante Pradilla-Golfito -0.257 0.127 0.184 0.008 -0.1 San Juan de Dios-Tomás Casas 0.768 1,432* 0.006 -0.426 0.058 Constant 4,252*** 4,053*** 3,706*** 2,565*** 0.262

Pseudo R2 0.1418 0.1090 0.060 0.0434 0.0683 Note: * significant at 10%, ** at 5% y *** at 1%. standard errors in parenthesis.

Source: authors´ elaboration using ENIG 2004 and geographic information for health access

For people with no CHE, those at percentile 60 of HFB, being in quintile two to five increase

of HFB, contrary to results for 98 and 99 percentile. These results might suggest that out of

pocket is progressive for people with no CHE. Household size, household with access to

improved sanitary conditions and with children affects positively the HFB (and it becomes

significant). The share of health expenditures not financed with cash become negative, as it is to

be expected because having access to credit will have a negative impact on HFB. The probability

of having insurance decreases HFB, as expected too. The health networks mentioned above are

no longer significant. Instead other health network seem to have a positive impact on HFB such

as i) México-Monseñor Sanabria, ii) México-San Rafael de Alajuela-Luis Carlos Valverde iii)

México-San Rafael de Alajuela-Guápiles/ Los Chiles iv) México/San Juan de Dios-San Rafael de

Alajuela-San Francisco. Therefore, those health networks have an impact on HFB but none of

them on HFB for people with CHE which is a remarkable result.

28

Determinants of HFB affect differently HFB conditional on its distribution, showing some degree

of heterogeneity, as it is presented in Table A2.3. They are consistent for 99, 98 percentile and in

some cases still for the 90 percentile. However, as mentioned before differences are evident for

percentile 60 and 80, compared to 99 and 98 percentiles. Those results shows that determinants

of HFB are different for those with out of pocket expenditures from those with catastrophic

expenditures (percentile 98/99).

Table A2.1 and A2.3 also shows the same estimations for different definitions of SE, household

food, and international poverty line. Household characteristics and health coverage are quite

consistent with results mentioned above, for national, international, and endogenous poverty line

definition of SE.

For the case of household food, results are slightly different: It is important to note that for those

with CHE, quintiles do not have a significant impact on HFB. In addition some health networks

will increase the HFB such as: San Juan de Dios alone, San Juan de Dios with: Carlos Durán,

Marcial Fallas, Solon Nunez and Escalante Pradilla.

Table 5 shows quintile regression with waiting lists instead of health coverage network. Results

are consistent with those found above. However, for household with CHE at percentile 99%,

29

being in fifth quintile do not have a significant impact on HFB, but being in quintile two to four

reduce HFB, except when household food is used as subsistence income. With endogenous

poverty line household size reduce HFB, while having children increases HFB. However, those

results are not robust for others measures (Table A2.2 and Table A2.4). Waiting lists do not have

an impact on HFB (for people at percentile 99%). For percentile 98% results are not robust,

when considering household food and international as subsistence income. Waiting lists in

medical procedures increase HFB.

For those with no CHE (percentile 60%), results are consistent with those found for health

services network before. Additional variables, such us having kids, the probability of being

insured, access to improved sanitary conditions and share of the HE financed with no cash seem

to determine HFB. In addition, being in higher quintiles increases HFB.

The most interesting results is that waiting lists in medical procedures seem to increase HFB for

percentile 60% but not for 99%, and waiting list in medical consultations seems to reduce HFB

also for 60% percentile, but not for 99%. Waiting lists in surgeries are not significant. This

could be explained, as mentioned above, by the fact waiting lists in medical consultations are

ophthalmology and orthopedics; therefore people are still willing to stay on line with no

additional outlay. Surgeries represented only 10% of the waiting lists. However, people may not

willing to wait for diagnose once doctors have requested special exams.

30

Table 5:

Quintile regression for HFB with health network coverage

Quintile / Endogenous poverty Line 99 98 90 80 60 Urban area 0.019 0.032 0.069 0.094 0.055 Quintile 2 -0,56** -0,445** -0,329*** -0.122 0,55*** Quintile 3 -0,888*** -0,682*** -0,349** -0.179 0,731*** Quintile 4 -0,704*** -0,597*** -0,433*** -0.112 0,976*** Quintile 5 -0.482 -0,487** -0,337** 0.008 1,194*** hhd with children 0,299* -0.134 0.039 0.095 0,14*** hhd with elderly 0,54*** 0,706*** 0,626*** 0,648*** 0,497*** hhd with elderly and children 0.504 -0.224 0.067 0.109 -0.054 hhd size -0,209** -0,139* -0,15** -0.094 0,164*** Prob of having insurance -0.735 -2.036 -0.543 -0.354 -1,162** hhd with electrical energy 0.578 0.514 0.053 -0.103 0.167 hhd with access to improved sanitary conditions -0.078 0.55 0.175 0.327 0,349*** hhd with access to improved water -0.181 -0.121 -0.174 -0.129 -0.048 hhd with access to waste disposal -0.211 -0.264 -0.136 -0,176** -0.065 years of schooling (head) -0.009 0.003 0.003 0.011 0.019 share of household expenses in health not in cash -0.12 0.106 -0.14 -0,282** -0,194*** No of cases in waiting list for surgery 0.021 0.008 0.003 0.016 0.008 No of cases in waiting list for specialized medical consultation -0.019 -0.022 -0,044** -0,034** -0,025** No of cases in waiting list for procedures 0.023 0.018 0,039*** 0,029** 0,023*** Constant 4,47*** 4,789*** 3,361*** 2,189*** 0.32 Pseudo R2 0.107 0.0844 0.04878 0.0302 0.0551 Quintile / National poverty line 99 98 90 80 60 Urban area 0.029 0.09 0.03 0,164** 0.082 Quintile 2 -0,83*** -0,63*** -0.074 -0.083 0,502*** Quintile 3 -1,137*** -0,715*** -0.074 -0.122 0,685*** Quintile 4 -0,977*** -0,678*** -0.193 -0.023 0,921*** Quintile 5 -0,851*** -0,555** -0.056 0.109 1,164*** hhd with children -0.012 -0.13 -0.098 0.085 0,153*** hhd with elderly 0,496** 0,691*** 0,597*** 0,569*** 0,5*** hhd with elderly and children 0.364 -0.088 0.013 0.075 -0.061 hhd size -0.154 -0.018 -0.05 0.015 0,189*** Prob of having insurance -0.705 -2.469 -0.803 -0.19 -1,246** hhd with electrical energy -0.329 -0.289 0.143 0.089 0.171 hhd with access to improved sanitary conditions 0.295 0,712* 0.261 0.286 0,399*** hhd with access to improved water -0.302 -0,542* -0.074 -0.167 -0.06 hhd with access to waste disposal -0.041 -0.077 -0.05 -0.114 -0.073 years of schooling (head) -0.005 0.001 0.005 0.009 0.019 share of household expenses in health not in cash -0.029 -0.069 -0.085 -0.212 -0,219*** No of cases in waiting list for surgery 0.003 -0.009 0.009 0.013 0.009 No of cases in waiting list for especialized medical consultation -0.022 -0,042* -0,036** -0.023 -0,025**

31

Cont/ No of cases in waiting list for procedures 0.023 0.03 0.022 0,023* 0,024*** Constant 5,249*** 6,058*** 2,835*** 1,492*** 0.317 Pseudo R2 0.0943 0.065 0.0364 0.0239 0.0567 Note: * significant at 10%, ** at 5% y *** at 1%. standard errors in parenthesis.

Source: authors´ elaboration using ENIG 2004 and geographic information for health access

Table A2.5 and A2.6 presents a tobit estimations for OOPs and for HFB. In this case they do not

take into account the heterogeneity of the HFB distribution, but results are consistent with those

found for percentile 60%. An interesting difference is waiting list in surgeries increases HFB, a

variable that was not significant before.

7. Discussion of results

Costa Rica is a country shows relatively high OOPS along with small CHE. The two main

components of out-of-pocket spending are doctor visits and drugs, over 80% and lab exams

represent about 7% of total health expenditures. Health expenditures are progressive with minor

exceptions in the bottom of the distribution (depending on the definition of ability to pay).

Medicines and therapeutic devices are regressive for the whole distribution, and lab exams are

progressive. Out-of-pocket hospital staying expenses are minimal and this is a very important

finding. Interestingly, the families that tested positive for catastrophic expenditure (at the 30%

threshold) had no hospital stays, and doctor visits represent their larger spending item. Even

though catastrophic health spending is shown to be small, the results are very sensitive to the

chosen indicator, and the definition of capacity to pay. That may say more about the nature of

32

the expenditure definitions than of the Costa Rican specific results. Whether or not the imputed

income of home ownership is used, it significantly affects the results, much more than the use of

income rather than expense does. The WvD1 (with NPL) indicator brings up the highest

proportion of households with catastrophic expenses. This is the result of viewing all poor

people as having catastrophic expenditures, and of the national poverty line being much higher

than the international or endogenous ones, and therefore the result is magnified.

The analysis by socioeconomic is less robust, when it is done in one dimension (i.e CHE are

analyzed by insurance condition, household composition, quintile, etc). Households without any

member insured do not have catastrophic expenditures for some indicators, and for the others the

share of households with CHE is less for households without any member insured. This may

suggest that families do not have CHE, because they do not use health services, but as it was

mentioned above 96% reported to use them. The same occurs when CHE is analyzed by

quintile. This analysis gives already important results as households with elderly members have

higher CHE, however it is important to take into account the correlation between variables. For

example, when other control variables are included in the regression analysis being in the fifth

quintile increases HFB for those with no CHE.

In addition, the estimation procedure shows that medical procedures ( which include ultrasound

and other exams) increase HFC but it does not do it for households with CHE. This may be due

to not being so important in the OOPS share (7%), and because they are very progressive. It is

33

also important to study health networks that are affecting in a significant way HFC, for example

it would important to study if population that receive care in San Juan de Dios-Tomas de las

Casas, and San Juan de Dios-Escalante Pradilla are relatively poor than those in other geographic

areas or whether they have larger health expenses due to health system characteristics.

Geographic variables are capturing the combined effect of having less capacity to pay or higher

OOPS, but it would be very interesting to disentangle those effects.

Finally is important to understand that having low CHE is the result of the Costa Rican health

system design (Muiser and Vargas, 2010), very high insurance coverage, no co-payment in health

care services, and a complementary private health care system. Challenges such as the population

ageing and the epidemiological transition are present and will play a larger role, as it does in

OECD countries, and that may increase health care costs questioning the financial sustainability

of the Costa Rican health care system.

8. Concluding remarks

The purpose of this chapter is to explain the paradox of high out-of-pocket health expenditures

while catastrophic health expenditures are very small. Enough evidence was provided on the

composition of out-of-pocket expenditures and their progressivity, the financial burden of-out-

pocket expenditures, the catastrophic costs and the impoverishing costs for different measures of

ability to pay, rates and thresholds. Besides, the risk factors that may explain OOPS and HFB,

including socioeconomic variables were successfully tested. It was so, also for the characteristics

34

of the health system, which involved mixing further information from the survey with

geographical access to the network of services and the waiting times. There was a novel feature

of the study as that has never been done before for Costa Rican National Health Systems cases.

Furthermore, in order to explain the risk factors that account for the financial burden of pocket

spending, but not necessarily those of the high catastrophic expenditures, a quintile regression to

determine the effect of variables in different parts of the distribution was estimated. That was

again a novel feature in Costa Rican health studies.

From the standpoint of public policies, the whereabouts and interactions of these networks of

services that were significant in explaining the catastrophic expenditure should be further

developed for future research. It could be done more by case studies than from a general

approach.

Finally, the quality of databases is a pending issue. The household expenditure survey was

exploited to their highest yield. A welcome alternative will be quality-of-life type surveys, which

combines information on health expenditures, health status, health system utilization, and health

system characteristics, such as access, quality, waiting lists, etc. If those surveys becomes

available, it will take the risk factors analysis that explain out-of-pocket costs and catastrophic

expenses to a higher plateau.

35

9. References

Briceño, R., M. Elizondo, et al. (2006). Catástrofe económica generada por mala salud: Una

aplicación para Costa Rica 2005. En Simposio: Costa Rica a la Luz de la Encuesta de

Ingresos y Gastos. I.N.E.C. San José.

Davidson R. y D. J.Y. (1997). "Statistical inference for the measurement of the incidence of taxes

and transfers." Econometrica 65(6): 1453-1465.

Deaton, A., Ed. (1997). The Analysis of Households Surveys: A Microeconomic Approach to

Development Policy, The World Bank and the Jonh Hopkings University Press.

Gakidou, E., C. J. L. Murray, et al. (2000) "A Framework for Measuring Health Inequality."

Global Programme on Evidence for Health Policy Discussion Paper Series.

Hatt, L. (2007). Measuring risk factors for catastrophic health expenditures in Peru and their

effects over family in time, John Hopkins University. Doctor in Philosophy.

Kawkani, N. C. (1976). "Measurement of tax progressivity: An international Comparison."

Economic Journal 87: 71-80.

Kawkani, N. C. (1977). "Application of Lorenz Curves in Economic Analysis." Econometrica

45(3): 719-727.

Kawkani, N. C. y N. Podder (1976). "Efficient Estimation of Lorenz Curve and associated

inequality measures from grouped observations." Econometrica 44(1): 87-103.

36

Kawkani, N. C., A. Wagstaff, et al. (1997). "Socio-economic inequalities in health:

Measurement, computation and statistical inference." Journal of Econometrics

77(1): 87-103.

Knaul, F. M., H. Arreola-Ornelas, et al. (2005). "Protección Financiera en Salud: México 1992 a

2004." Salud Pública de México 47( 6): 430-439.

Muiser, J and JR Vargas (2010). The design of the Costa Rican health financing system in view

of financial risk protection. Forthcoming In Health Care Financing and Social Protection

in LA. Vol.1

Murray, C. J. L., Xu, K., et al. Assessing the distribution of household financial contributions to

the health system: concepts and empirical application. In Health Systems Performance

Assessment: Debates, Methods and Empiricism. Eds. C. J. L. Murray and D. B. Evans.

Murray, C. J. L., F. M. Knaul, et al. (2000) "Defining and Measuring Fairness in Financing

Contribution to health system." GPD Discussion Paper Series No.24.WHO

O'Donell, O., E. van Doorslaer, et al. (2008). Analyzing Health Equity using Household Survey

Data A Guide Techniques and their implementation. World Bank. Washington.

Poullier, J. P. y P. Hernández (2000) "Estimates of National Health Accounts (NHA) for 1997."

WHO/EIP Discussion Paper. No 27.

Van Doorslaer, E. (1999). "The redistributive effect of health care finance in twelve countries."

Journal of Health Economics 18: 291-313.

37

Van Doorslaer, E. y A. Wagstaff (1997). "Income related inequalities in health: some

international comparisons." Journal of Health Economics 16: 93-112.

Wagstaff A, et al (2003). Catastrophe and impoverish in paying for health care with applications

to Vietnam 1993-1998. Health Economics 12(11): 921-933

Wagstaff, A. (2001). "Measuring equity in health care financing: Reflections on and alternatives

to the World Health Organization's Fairness and Financing Index."

Wagstaff, A. y P. Paci (1991). "On the measurement of inequalities in health." Social Science

Medicine 33(5): 263-290.

Wagstaff, A. y E. van Doorslaer (1997). "Progressivity, horizontal equity and reranking in health

care finance: a decomposition analysis for Netherlands." Journal of Health Economics

16(5): 499-516.

Wagstaff, A. y E. van Doorslaer (1999). "Redistributive effect, progressivity and differential tax

treatment: Personal Income taxes in twelve OECD countries." Journal of Public

Economics 72(1): 73-98.

Wagstaff, A. y E. van Doorslaer, (1999). "Equity in finance of health care: some further

comparisons." Journal of Health Economics 18: 263-290.

WHO (2000). The World Health Report 2000. Switzerland, Geneva.

Xu, K. (2005) "Distribution of Health Payments and Catastrophic Expenditures." WHO Health

Financing Policy Discussion Paper.

38

Xu, K., D. Evans, et al. (2003). "Household Catastrophic Health Expenditures: A multicountry

Analysis." The Lancet 362: 111-117.

Xu, K., D. B. Evans, et al. (2005) "Designing Health Financing Systems to reduce catastrophic

expenditures." Technical brief for policy-makers.

Xu, K., J. Klavus, et al. Summary measures of the distribution of household financial

contributions to health in Health Systems Performance Assessment: Debates, Methods

and Empiricism. Eds. C. J. L. Murray and D. B. Evans, World Health Organization.

Xu, K., J. Klavus, et al. Household health system contributions and capacity to pay: definitional,

empirical and technical challenges

Health Systems Performance Assessment: Debates, Methods and Empiricism. C. J. L. Murray

and D. Evans, World Health Organization.

Zuniga-Brenes, M. P. (2006). Equidad y justicia en el financiamiento del sistema de salud: Una

aplicación al sistema de salud costarricense. San José, Universidad de Costa Rica. Mimeo

Zúñiga-Brenes, M. P. (2006). Progresividad en el financiamiento al gasto en salud, en gasto

privado en educación y en algunos impuestos: Una aplicación de dominancia. En

Simposio: Costa Rica a la Luz de Ingresos y Gastos 2004. I.N.E.C. San José.

Zúñiga-Brenes, M. P. (2008). Three essays in Health Economics. Economics. Lausanne. Thesis

Ph.D in Economics: 183pp.

39

Zúñiga-Brenes, M. P., M. Saborío, et al. (2006). Una mirada al bienestar de las familias

costarricenses, a través del consumo. En Simposio: Costa Rica a la Luz de la Encuesta de

Ingresos y Gastos. I.N.E.C. San José.

1

Appendix A1

Table A1.1

Summary Statistics of main variables

Variable Mean Std. Dev. Variable Mean

Std. Dev.

Households in rural areas 37.91% 0.03 National poverty line at hhd level 57,857.17 552.39

Composition of households by age International pov line 32,065.36 285.56

With children 24.95% 0.01 Endogenous pov line 58,307.73 306.98

With elderly 16.41% 0.01 Different indicators of Capacity to pay

(CP) With children and elderly 1.74% 0.00 Food expenditures 397,430.70 19222.66 With children and elderly 56.90% 0.01 National pov line 391,019.40 19854.49

Household size International pov line 416,811.20 19894.99

2 members or less 25.91% Endogenous pov line 390,568.80 19887.36

3-4 members 44.59% Health financial burden of Oops

5 members or more 29.50%

0.01

Food expenditures 2.56% 0 % hhd with at least one person insured 90.20% 0.01 National pov line 3.34% 0.01 Total expenditures 448,876.60 19846.04 Internacional pov line 2.63% 0 Total Oops expenditures 11,430.48 705.57 Endogenous pov line 3.60% 0.03

Food expenditures 51,445.85 1069.23 No of households 4231

Table A1.2

Kawkani indices for OOPS

Capacity to pay with ES Food Nat Pov Line Int Pov Line Endogenous Pov

Line Subsistence

Expenditure=0

Drugs -0.047 -0.072 -0.042 -0.077 -0.008 Medical Consultations 0.045 0.013 0.044 0.008 0.080 Odontological Consultations 0.084 0.055 0.086 0.050 0.120 Laboratory exams 0.083 0.055 0.086 0.050 0.122

Priv. Hospitalization 0.381 0.345 0.377 0.341 0.416 Pub Hospitalization 0.183 0.192 0.225 0.188 0.261 Surgeries 0.383 0.355 0.387 0.351 0.426 Therapeutically suppl -0.033 -0.068 -0.037 -0.074 -0.003 OOPs 0.032 0.002 0.033 -0.003 0.068

2

Table A1.3

Catastrophic Expenditures, Thresholds and Capacity to Pay (mean values in percentages)

Exp- ES with no rental value

Income –ES with no rental value

Expend-ES with rental value

Inc-ES with rental value

Mean Mean Mean Mean In % In % In % In %

Threshold

k=20% 2,91 5,55 1,41 3,25 [2.37 - 3.44] [4.80 – 6.30] [1.04 - 1.79] [2.67 - 3.84]

k=30% 1,32 3,89 0,43 1,54 [0.95 - 1.68] [3.24 – 4.54] [0.24 - 0.62] [1.13 - 1.94]

WHO

k=40% 0,62 2,84 0,07 1,17 WvD1 [0.39 - 0.85] [2.29 – 3.38] [-0.01 - 0.15] [0.81 - 1.52]

k=20% 5,78 6,07 2,99 3,25 [5.04 - 6.51] [5.30 – 6.84] [2.45 - 3.54] [2.66 - 3.84]

k=30% 3,58 3,66 1,56 1,84 [2.99 - 4.17] [3.04 – 4.28] [1.18 - 1.93] [1.40 - 2.29]

k=40% 2,76 2,82 1,18 1,30 [2.26 - 3.26] [2.28 – 3.37] [0.85 - 1.52] [0.93 - 1.68]

k=20% 4,51 5,49 2,21 2,79 [3.85 - 5.16] [4.74 – 6.24] [1.74 - 2.68] [2.24 - 3.33]

k=30% 2,13 3,00 0,74 1,31 [1.66 - 2.59] [2.42 – 3.59] [0.49 - 1.00] [0.94 - 1.69]

Hybrid1

k=40% 1,31 2,11 0,37 0,75 [0.96 - 1.65] [1.61 – 2.61] [0.19 - 0.55] [0.46 - 1.04]

k=20% 3,27 3,41 1,66 1,93 [2.71 - 3.83] [2.81 – 4.01] [1.26 - 2.07] [1.46 - 2.41]

k=30% 1,87 1,90 0,60 0,71 [1.44 - 2.31] [1.43 – 2.37] [0.36 - 0.84] [0.41 - 1.00]

WvD2

k=40% 1,20 0,99 0,39 0,39 [0.88 - 1.53] [0.66 – 1.32] [0.20 - 0.59] [0.20 - 0.59]

k=20% 2,79 3,48 1,51 1,88 [2.27 - 3.31] [2.86 – 4.10] [1.12 - 1.89] [1.41 - 2.34]

k=30% 1,29 1,97 0,44 0,65 [0.92 - 1.65] [1.47 – 2.47] [0.24 - 0.64] [0.37 - 0.93]

Hybrid2

k=40% 0,61 1,04 0,24 0,33 [0.39 - 0.84] [0.68 – 1.41] [0.08 - 0.39] [0.17 - 0.50]

k=20% 2,76 4,54 1,28 2,46 [2.24 - 3.29] [3.86 – 5.23] [0.92 - 1.63] [1.93 - 2.98]

k=30% 1,14 2,89 0,31 0,90 [0.79 - 1.48] [2.31 – 3.47] [0.15 - 0.47] [0.58 - 1.22]

Xu

k=40% 0,56 1,81 0,05 0,55 [0.34 - 0.78] [1.36 – 2.26] [-0.02 - 0.12] [0.29 - 0.80]

NA 0,40 0,33 0,14 0,03 Impoverishment NPL [0.21 - 0.60] [0.15 – 0.50] [0.03 - 0.25] [-0.03 - 0.09]

NA 0,13 0,09 0,10 0,07 Impoverishment IPL [0.01 - 0.24] [0.00 – 0.18] [-0.01 - 0.22] [-0.01 - 0.15]

3

Table A2.3 Catastrophic Health Expenditures according different variables (mean value: not percentages)

Variable Area of residence Quintile of CP

Thresholds

Urban Rural

Q I (+ poor)

Q II

Q III

Q IV

Q V (- poor)

k=20% 0.016 0.012 0.009 0.014 0.005 0.019 0.024 k=30% 0.004 0.004 0.002 0.005 0.002 0.005 0.007 WHO

k=40% 0.001 0.000 0.001 0.000 0.000 0.000 0.002 k=20% 0.025 0.038 0.073 0.024 0.014 0.016 0.022 k=30% 0.009 0.026 0.062 0.008 0.001 0.001 0.006 WvD1 k=40% 0.006 0.021 0.055 0.003 0.000 0.000 0.001 k=20% 0.021 0.024 0.034 0.024 0.014 0.016 0.022 k=30% 0.005 0.011 0.021 0.008 0.001 0.001 0.006 Hybrid1 k=40% 0.002 0.006 0.014 0.003 0.000 0.000 0.001 k=20% 0.015 0.019 0.031 0.015 0.004 0.014 0.019 k=30% 0.005 0.008 0.021 0.004 0.001 0.000 0.004 WvD2 k=40% 0.003 0.006 0.017 0.002 0.000 0.000 0.001 k=20% 0.015 0.016 0.023 0.015 0.004 0.014 0.019 k=30% 0.004 0.005 0.013 0.004 0.001 0.000 0.004 Hybrid2 k=40% 0.002 0.003 0.009 0.002 0.000 0.000 0.001 k=20% 0.014 0.011 0.009 0.013 0.004 0.016 0.021 k=30% 0.003 0.003 0.002 0.004 0.002 0.002 0.005 Xu k=40% 0.001 0.000 0.001 0.000 0.000 0.000 0.001

NA 0.001 0.002 0.007 0.000 0.000 0.000 0.000

Impoverishment NPL [000 - 000] [000 - 000] [000 - 000] [000 - 000] [000 - 000] [000 - 000] [000 - 000]

NA 0.001 0.002 0.005 0.000 0.000 0.000 0.000 Impoverishment IPL [000 - 000] [000 - 000] [000 - 000] [000 - 000] [000 - 000] [000 - 000] [000 - 000]

4

Cont/

With children With

elderly

With children

and elderly

No children no elderly

2 members

or less 3-4

members 5 or more members No HI With HI

k=20% 0.007 0.036 0.018 0.011 0.031 0.008 0.009 0.008 0.015 k=30% 0.002 0.015 0.000 0.002 0.009 0.002 0.003 0.000 0.005 WHO k=40% 0.000 0.003 0.000 0.000 0.003 0.000 0.000 0.000 0.001 k=20% 0.040 0.057 0.042 0.018 0.039 0.020 0.037 0.015 0.032 k=30% 0.029 0.021 0.042 0.007 0.012 0.011 0.026 0.011 0.016 WvD1 k=40% 0.026 0.011 0.042 0.005 0.007 0.007 0.023 0.009 0.012 k=20% 0.017 0.054 0.018 0.015 0.038 0.016 0.018 0.009 0.024 k=30% 0.005 0.018 0.018 0.005 0.011 0.006 0.006 0.004 0.008 Hybrid 1 k=40% 0.002 0.008 0.018 0.003 0.006 0.003 0.003 0.003 0.004 k=20% 0.020 0.033 0.018 0.010 0.027 0.010 0.018 0.007 0.018 k=30% 0.014 0.008 0.018 0.002 0.006 0.003 0.011 0.000 0.007 WvD2 k=40% 0.010 0.004 0.000 0.001 0.002 0.002 0.008 0.000 0.004 k=20% 0.015 0.033 0.018 0.010 0.027 0.010 0.012 0.007 0.016 k=30% 0.008 0.008 0.018 0.001 0.006 0.003 0.006 0.000 0.005 Hybrid2 k=40% 0.005 0.004 0.000 0.001 0.002 0.002 0.003 0.000 0.003 k=20% 0.007 0.034 0.018 0.009 0.028 0.007 0.009 0.006 0.013 k=30% 0.002 0.010 0.000 0.002 0.008 0.002 0.001 0.000 0.003 Xu

k=40% 0.000 0.002 0.000 0.000 0.002 0.000 0.000 0.000 0.001 Impoverishment NPL

0.001 0.006 0.000 0.000 0.003 0.002 0.000 0.003 0.001

Impoverishment IPL

0.004 0.001 0.000 0.000 0.001 0.001 0.002 0.000 0.001

5

Appendix A2: Estimations Table A2.1

Quintile regressions for Health Financial Burden of CP with health access variables: Endogenous and National poverty line

Endogenous poverty line National poverty line Quintile 60 80 90 98 99 60 80 90 98 99

0,128 0,094 0,098 -0,018 0,182 0,144 0,162 0,108 0,141 -0,027 Urban [0.068]* [0.087] [0.088] [0.157] [0.210] [0.069]** [0.086]* [0.072] [0.144] [0.159]

0,492 -0,149 -0,296 -0,441 -0,591 0,479 -0,126 -0,081 -0,46 -0,379 Quintil 2 [0.099]*** [0.140] [0.131]** [0.161]*** [0.236]** [0.109]*** [0.118] [0.122] [0.194]** [0.191]**

0,607 -0,128 -0,272 -0,564 -0,86 0,609 -0,095 -0,052 -0,553 -0,462 Quintil 3 [0.120]*** [0.129] [0.141]* [0.166]*** [0.246]*** [0.118]*** [0.129] [0.121] [0.214]*** [0.203]**

0,89 -0,046 -0,345 -0,586 -0,755 0,881 -0,002 -0,115 -0,654 -0,42 Quintil 4 [0.108]*** [0.135] [0.143]** [0.180]*** [0.253]*** [0.111]*** [0.129] [0.105] [0.222]*** [0.198]**

1,073 0,049 -0,293 -0,461 -0,598 1,08 0,104 -0,027 -0,501 -0,267 Quintil 5 [0.111]*** [0.149] [0.161]* [0.210]** [0.282]** [0.115]*** [0.120] [0.138] [0.236]** [0.253]

0,137 0,132 0,013 -0,013 0,044 0,144 0,119 0,012 -0,12 0,007 Hhd with children [0.050]*** [0.066]** [0.076] [0.139] [0.162] [0.059]** [0.062]* [0.073] [0.111] [0.135]

0,431 0,701 0,72 0,627 0,589 0,453 0,63 0,693 0,506 0,694 Hhd with elderly [0.111]*** [0.130]*** [0.117]*** [0.182]*** [0.244]** [0.107]*** [0.121]*** [0.128]*** [0.201]** [0.230]***

0,05 0,096 0,15 -0,072 0,206 0,032 0,091 0,195 -0,052 0,464 Hhd with children and elderly [0.181] [0.214] [0.219] [0.568] [0.479] [0.176] [0.249] [0.206] [0.638] [0.518] 0,151 -0,044 -0,057 -0,137 -0,162 0,188 0,041 -0,028 -0,002 -0,051 Household size

[0.047]*** [0.064] [0.072] [0.092] [0.113] [0.051]*** [0.061] [0.072] [0.105] [0.124] -0,974 -1,056 -2,239 -1,883 -1,091 -1,173 -0,896 -1,692 -1,382 -2,306 Probability of having insurance

[0.540]* [1.106] [1.031]** [1.173] [1.468] [0.482]** [1.056] [1.029] [1.626] [1.799] 0,224 0,198 0,459 0,536 0,131 0,21 0,176 0,529 0,66 0,295 Calderón Guardia-Catedral

Noreste [0.181] [0.166] [0.203]** [0.371] [0.396] [0.216] [0.175] [0.184]*** [0.410] [0.404] 0,169 0,11 0,269 0,688 0,361 0,159 0,099 0,341 0,793 0,452 Calderón Guardia-Coronado

[0.223] [0.203] [0.251] [0.439] [0.449] [0.234] [0.184] [0.219] [0.405]* [0.425]

6

Continuation Endogenous poverty line National poverty line Quantile 60 80 90 98 99 60 80 90 98 99

-0,119 -0,534 -0,327 -0,721 -1,097 -0,139 -0,5 -0,288 -0,774 -1,18 Calderón Guardia-Clorito Picado [0.355] [0.378] [0.393] [0.480] [0.460]** [0.347] [0.359] [0.357] [0.404]* [0.436]***

-0,1 0,124 0,301 0,693 0,157 -0,099 0,095 0,356 0,704 0,202 Calderón Guardia-Jiménez Núñez [0.261] [0.369] [0.314] [0.476] [0.433] [0.271] [0.314] [0.379] [0.463] [0.423]

0,17 0,049 0,205 0,315 0,033 0,185 0,072 0,303 0,431 0.00 Calderón Guardia-Max Peralta [0.189] [0.137] [0.242] [0.377] [0.346] [0.216] [0.119] [0.177]* [0.404] [0.386] -0,042 0,16 0,36 0,115 -0,196 -0,071 0,14 0,354 0,224 -0,187 Calderón Guardia-Max

Peralta-Catedral Noreste [0.347] [0.247] [0.261] [0.355] [0.378] [0.325] [0.253] [0.208]* [0.384] [0.385] 0,455 0,543 0,571 0,609 0,159 0,443 0,447 0,667 0,382 0,046 Calderón Guardia-Max

Peralta-Marcial Fallas [0.331] [0.246]** [0.301]* [0.454] [0.470] [0.333] [0.241]* [0.226]*** [0.490] [0.519] 0,019 0,016 0,169 0,351 0,155 0,01 -0,053 0,155 0,304 -0,164 Calderón Guardia-William

Allen [0.199] [0.304] [0.303] [0.421] [0.468] [0.210] [0.255] [0.255] [0.436] [0.471] -0,123 -0,288 0,005 0,217 -0,115 -0,122 -0,298 0,154 0,213 -0,176 Calderón Guardia-Tony Facio [0.179] [0.176] [0.266] [0.369] [0.494] [0.200] [0.150]** [0.202] [0.375] [0.411] -0,022 0,209 0,376 0,462 0,12 -0,032 0,06 0,564 0,609 0,23 Calderón Guardia-Tony Facio-

Guápiles [0.206] [0.330] [0.270] [0.369] [0.383] [0.215] [0.232] [0.243]** [0.440] [0.391] -0,438 -0,458 -0,628 -1,023 -1,452 -0,415 -0,532 -0,605 -0,954 -1,421 Calderón Guardia/México-

Clorito Picado-Jiménez Núñez [0.330] [0.325] [0.187]*** [0.360]*** [0.400]*** [0.325] [0.213]** [0.173]*** [0.380]** [0.425]*** 0,322 0,308 0,446 0,717 0,159 0,324 0,342 0,547 0,571 0,142 Calderón Guardia/México-

Jiménez Núñez [0.309] [0.348] [0.344] [0.483] [0.390] [0.332] [0.339] [0.343] [0.432] [0.456] 0,03 0,248 0,41 0,276 -0,295 0,028 0,351 0,442 0,369 -0,189 México

[0.587] [0.427] [0.399] [0.423] [0.429] [0.558] [0.409] [0.349] [0.489] [0.399] -0,13 0,041 0,238 -0,13 0,704 -0,142 0,101 0,283 0,007 0,724 México-Clorito Picado

[0.235] [0.337] [0.291] [0.661] [0.672] [0.271] [0.312] [0.240] [0.651] [0.626] 0,111 0,033 0,165 0,357 0,46 0,107 0,061 0,179 0,456 0,36 México-San Vicente

[0.171] [0.135] [0.205] [0.365] [0.402] [0.197] [0.109] [0.148] [0.472] [0.436] 0,002 -0,086 0,146 0,278 -0,06 -0,008 -0,072 0,227 0,471 0,151 México-Baltodano

[0.188] [0.172] [0.230] [0.395] [0.417] [0.197] [0.152] [0.175] [0.422] [0.483] -0,139 -0,389 -0,161 0,418 0,234 -0,148 -0,313 -0,259 0,488 0,163 México-Baltodano-Anexión [0.180] [0.264] [0.422] [0.505] [0.466] [0.215] [0.262] [0.358] [0.491] [0.528] -0,113 -0,351 -0,138 0,679 0,097 -0,127 -0,323 -0,128 -0,388 -1,2 México-Baltodano-Upala [0.221] [0.408] [0.454] [0.622] [0.590] [0.220] [0.356] [0.292] [0.401] [0.520]**

7

Continuation Endogenous poverty line National poverty line Quintile 60 80 90 98 99 60 80 90 98 99

0,004 -0,164 0,154 0,407 0,097 -0,001 -0,08 0,313 0,757 0,207 México-Mons. Sanabria [0.172] [0.193] [0.230] [0.402] [0.403] [0.198] [0.162] [0.179]* [0.420]* [0.410] 0,671 0,479 0,435 -0,092 -0,304 0,648 0,327 0,31 -0,037 -0,567 México-Mons. Sanabria-Max

Terán [0.380]* [0.383] [0.287] [0.414] [0.456] [0.323]** [0.332] [0.244] [0.448] [0.471] -0,036 -0,072 0,058 0,271 0,049 -0,038 -0,061 0,107 0,479 0,036 México-San Carlos [0.193] [0.168] [0.231] [0.395] [0.416] [0.186] [0.146] [0.151] [0.435] [0.427] -0,215 -1,3 -0,878 -1,523 -1,887 -0,233 -1,14 -0,793 -1,022 -2,052 México-San Carlos-Los Chiles [0.227] [0.427]*** [0.454]* [0.532]*** [0.544]*** [0.248] [0.385]*** [0.443]* [0.485]** [0.546]*** 0,219 -0,02 0,736 0,533 0,192 0,23 0,077 0,746 0,896 0,004 México-San Carlos-San

Vicente [0.580] [0.869] [0.909] [0.772] [0.827] [0.684] [0.847] [0.826] [0.788] [0.669] -0,28 0,412 0,7 0,099 -0,441 -0,295 0,385 0,751 0,346 -0,676 México-San Carlos/San Rafael

Alajuela-Luis C. Valverde [0.626] [0.604] [0.514] [0.536] [0.583] [0.710] [0.628] [0.374]** [0.600] [0.531] 0,287 0,199 0,249 0,181 0,324 0,271 0,203 0,281 0,192 -0,016 México-San Rafael Alajuela-

Marcial Rodríguez [0.189] [0.136] [0.185] [0.366] [0.474] [0.201] [0.130] [0.147]* [0.369] [0.410] 0,747 0,48 0,624 0,745 0,799 0,778 0,501 0,599 0,694 -0,143 Meixco-San Rafael Alajuela-

Luis C. Valverde [0.237]*** [0.188]** [0.204]*** [0.572] [0.635] [0.273]*** [0.182]*** [0.148]*** [0.552] [0.697] 1,176 0,589 0,386 -0,215 -0,896 1,18 0,824 0,766 1,306 0,446 Meixco-San Rafael Alajuela-

Guápiles/Los Chiles [0.542]** [0.374] [0.322] [0.466] [0.575] [0.477]** [0.551] [0.741] [0.817] [0.774] 0,487 0,453 0,602 0,27 -0,017 0,498 0,436 0,549 0,378 -0,248 México/San Juan de Dios-San

Rafael Alajuela-San Francisco Asís [0.266]* [0.195]** [0.243]** [0.418] [0.452] [0.256]* [0.185]** [0.178]*** [0.378] [0.371]

-0,015 0,307 -0,232 -0,801 -1,334 -0,033 0,416 0,006 -0,836 -1,656 México/San Juan de Dios-Mons. Sanabria [0.630] [0.603] [0.368] [0.461]* [0.517]*** [0.711] [0.471] [0.293] [0.430]* [0.476]***

0,061 -0,003 0,296 0,421 0,724 0,075 -0,024 0,249 0,446 0,832 San Juan de Dios [0.187] [0.194] [0.241] [0.498] [0.522] [0.222] [0.142] [0.193] [0.518] [0.604] -0,158 0,112 0,677 0,212 -0,355 -0,162 0,091 0,64 0,056 -0,346 San Juan de Dios-Carlos Durán [0.403] [0.522] [0.629] [0.700] [0.724] [0.451] [0.570] [0.616] [0.650] [0.657] -0,234 -0,82 0,357 -0,058 -0,505 -0,261 -0,806 0,385 0,081 -0,501 San Juan de Dios-Clorito

Picado [0.357] [0.756] [0.673] [0.595] [0.583] [0.495] [0.804] [0.717] [0.650] [0.593] -0,036 0,039 0,168 0,576 0,338 -0,015 0,045 0,17 0,633 0,354 San Juan de Dios-Marcial

Fallas [0.184] [0.158] [0.200] [0.374] [0.399] [0.207] [0.131] [0.143] [0.377]* [0.372]

8

Continuation

Endogenous poverty line National poverty Quintile 60 80 90 98 99 60 80 90 98 99

0,124 0,049 0,154 0,493 0,212 0,13 0,03 0,21 0,478 0,221 San Juan de Dios-Moreno Cañas [0.197] [0.160] [0.208] [0.421] [0.358] [0.226] [0.128] [0.164] [0.399] [0.387]

-0,087 0,055 0,171 0,356 0,024 -0,074 0,113 0,179 0,305 -0,099 San Juan de Dios-Solón Núñez [0.261] [0.224] [0.294] [0.390] [0.385] [0.290] [0.242] [0.243] [0.406] [0.402] 0,021 0,104 0,193 0,649 0,277 0,006 0,1 0,234 0,64 0,53 San Juan de Dios-Escalante

Pradilla [0.283] [0.184] [0.321] [0.380]* [0.444] [0.218] [0.138] [0.226] [0.447] [0.461] 0,04 -0,017 0,141 0,221 0,71 0,024 0,007 0,376 0,807 0,771 San Juan de Dios-Escalante

Pradilla-Ciudad Neily [0.216] [0.278] [0.229] [0.558] [0.671] [0.216] [0.243] [0.240] [0.584] [0.589] -0,1 0,008 0,184 0,127 -0,257 -0,043 -0,064 0,418 0,394 -0,449 San Juan de Dios-Escalante

Pradilla-Golfito [0.364] [0.484] [0.437] [0.422] [0.450] [0.369] [0.471] [0.492] [0.451] [0.499] 0,058 -0,426 0,006 1,432 0,768 0,054 -0,434 0,549 1,294 0,591 San Juan de Dios-Tomás Casas

[0.307] [0.658] [0.909] [0.829]* [1.018] [0.313] [0.632] [1.000] [0.769]* [0.852] 0,097 -0,031 0,358 0,605 0,664 0,11 0,052 0,152 -0,402 0,493

House with access to electrical energy

[0.170] [0.420] [0.412] [0.445] [0.428] [0.161] [0.341] [0.348] [0.641] [0.809] 0,298 0,293 0,388 0,564 0,204 0,33 0,329 0,363 0,491 0,318 House with access to improved

sanitary conditions [0.116]** [0.250] [0.222]* [0.322]* [0.388] [0.094]*** [0.234] [0.167]** [0.323] [0.409] -0,014 -0,127 -0,156 -0,118 -0,113 0,001 -0,09 -0,153 -0,431 -0,437 House with access to improved

water [0.112] [0.194] [0.150] [0.230] [0.290] [0.132] [0.163] [0.157] [0.292] [0.358] -0,1 -0,121 -0,066 -0,219 -0,298 -0,108 -0,105 0,004 -0,044 -0,221 House with access to waste

disposal [0.077] [0.103] [0.124] [0.168] [0.201] [0.070] [0.093] [0.105] [0.197] [0.193] 0,018 0,011 0,007 0,001 -0,004 0,019 0,011 0,007 -0,002 -0,005 Years of schooling of head

[0.026] [0.025] [0.019] [0.023] [0.026] [0.028] [0.026] [0.018] [0.027] [0.020] -0,151 -0,264 -0,101 0,014 -0,185 -0,163 -0,238 -0,189 -0,048 -0,201 Share of health expenditures

financed with no cash [0.081]* [0.151]* [0.173] [0.214] [0.226] [0.086]* [0.124]* [0.174] [0.212] [0.272] 0,262 2,565 3,706 4,053 4,252 0,303 1,97 3,009 4,431 5,293 Constant

[0.327] [0.704]*** [0.608]*** [0.715]*** [0.821]*** [0.308] [0.637]*** [0.604]*** [1.041]*** [1.060]*** Pseudo R2 0.0683 0.0434 0.0600 0.1090 0.1418 0.0702 0.0378 0.0512 0.0977 0.1345 Note: * significant at 10%, ** at 5% y *** at 1%. standard errors in parenthesis.

Source: authors´ elaboration using ENIG 2004 and geographic records

9

Table A2.2 Quintile regressions for Health Financial Burden of CP with waiting list: Endogenous and National poverty line

Endogenous poverty line National poverty line Quantile 60 80 90 98 99 60 80 90 98 99

0,055 0,094 0,069 0,032 0,019 0,082 0,164 0,03 0,09 0,029 Urban [0.060] [0.068] [0.099] [0.122] [0.150] [0.066] [0.066]** [0.087] [0.124] [0.151]

0,55 -0,122 -0,329 -0,445 -0,56 0,502 -0,083 -0,074 -0,63 -0,83 Quintile 2 [0.113]*** [0.125] [0.114]*** [0.202]** [0.243]** [0.103]*** [0.108] [0.100] [0.238]*** [0.209]***

0,731 -0,179 -0,349 -0,682 -0,888 0,685 -0,122 -0,074 -0,715 -1,137 Quintile 3 [0.111]*** [0.139] [0.146]** [0.167]*** [0.255]*** [0.110]*** [0.105] [0.120] [0.241]*** [0.251]***

0,976 -0,112 -0,433 -0,597 -0,704 0,921 -0,023 -0,193 -0,678 -0,977 Quintile 4 [0.105]*** [0.144] [0.135]*** [0.168]*** [0.245]*** [0.106]*** [0.115] [0.126] [0.258]*** [0.254]***

1,194 0,008 -0,337 -0,487 -0,482 1,164 0,109 -0,056 -0,555 -0,851 Quintile 5 [0.116]*** [0.142] [0.142]** [0.205]** [0.313] [0.115]*** [0.110] [0.131] [0.246]** [0.281]***

0,14 0,095 0,039 -0,134 0,299 0,153 0,085 -0,098 -0,13 -0,012 Hhd with children [0.048]*** [0.058] [0.075] [0.136] [0.180]* [0.048]*** [0.060] [0.074] [0.131] [0.204]

0,497 0,648 0,626 0,706 0,54 0,5 0,569 0,597 0,691 0,496 Hhd with elderly [0.110]*** [0.120]*** [0.120]*** [0.154]*** [0.202]*** [0.086]*** [0.121]*** [0.117]*** [0.169]*** [0.216]**

-0,054 0,109 0,067 -0,224 0,504 -0,061 0,075 0,013 -0,088 0,364 Hhd with elderly and children [0.191] [0.215] [0.201] [0.580] [0.474] [0.204] [0.262] [0.199] [0.623] [0.518]

0,164 -0,094 -0,15 -0,139 -0,209 0,189 0,015 -0,05 -0,018 -0,154 Household size [0.047]*** [0.060] [0.069]** [0.082]* [0.089]** [0.045]*** [0.060] [0.080] [0.092] [0.120]

-1,162 -0,354 -0,543 -2,036 -0,735 -1,246 -0,19 -0,803 -2,469 -0,705 Prob of having insurance [0.532]** [0.980] [0.910] [1.250] [1.428] [0.495]** [0.834] [1.134] [1.501] [1.587]

0,167 -0,103 0,053 0,514 0,578 0,171 0,089 0,143 -0,289 -0,329 House with access to electrical energy [0.183] [0.542] [0.361] [0.439] [0.521] [0.198] [0.401] [0.315] [0.675] [0.763]

0,349 0,327 0,175 0,55 -0,078 0,399 0,286 0,261 0,712 0,295 House with access to improved sanitary cond [0.119]*** [0.304] [0.219] [0.369] [0.395] [0.093]*** [0.281] [0.183] [0.414]* [0.474]

-0,048 -0,129 -0,174 -0,121 -0,181 -0,06 -0,167 -0,074 -0,542 -0,302 House with access to improved water [0.141] [0.173] [0.145] [0.197] [0.214] [0.108] [0.138] [0.137] [0.295]* [0.276]

-0,065 -0,176 -0,136 -0,264 -0,211 -0,073 -0,114 -0,05 -0,077 -0,041 House with access to waste disposal [0.086] [0.089]** [0.123] [0.167] [0.180] [0.077] [0.087] [0.109] [0.160] [0.191]

0,019 0,011 0,003 0,003 -0,009 0,019 0,009 0,005 0,001 -0,005 Years of schooling head [0.027] [0.026] [0.024] [0.036] [0.061] [0.032] [0.025] [0.027] [0.025] [0.047]

10

Continuation Endogenous poverty line National poverty line Quintile 60 80 90 98 99 60 80 90 98 99

Share of HE that is not -0.194 -0.282 -0.14 0.106 -0.12 -0.219 -0.212 -0.085 -0.069 -0.029 financed with cash [0.073]*** [0.143]** [0.128] [0.294] [0.286] [0.079]*** [0.160] [0.156] [0.263] [0.259]

0,008 0,016 0,003 0,008 0,021 0,009 0,013 0,009 -0,009 0,003 No of cases in waiting list in surgery [0.008] [0.011] [0.013] [0.015] [0.017] [0.006] [0.010] [0.008] [0.017] [0.019]

-0,025 -0,034 -0,044 -0,022 -0,019 -0,025 -0,023 -0,036 -0,042 -0,022 No of cases in waiting list in medical consultations [0.010]** [0.017]** [0.018]** [0.027] [0.037] [0.011]** [0.018] [0.018]** [0.025]* [0.031]

0,023 0,029 0,039 0,018 0,023 0,024 0,023 0,022 0,03 0,023 No of cases in waiting list in procedures [0.007]*** [0.012]** [0.014]*** [0.022] [0.033] [0.008]*** [0.013]* [0.014] [0.021] [0.030]

0,32 2,189 3,361 4,789 4,47 0,317 1,492 2,835 6,058 5,249 Constant [0.282] [0.691]*** [0.516]*** [0.649]*** [0.869]*** [0.260] [0.544]*** [0.678]*** [1.029]*** [0.880]***

PseudoR2 0.0551 0.0302 0.04878 0.0844 0.1070 0.0567 0.0239 0.0364 0.065 0.0943 Nota: * significant at 10%, ** 5% y *** 1%. standard errors in parenthesis.

Source: authors´ calculation with ENIG 2004 and waiting list

11

Table A2.3 Quintile regressions for Health Financial Burden of CP with health access coverage: Food and International poverty line

International poverty line Food Quintile 60 80 90 98 99 60 80 90 98 99

0,118 0,106 0,116 -0,004 0,153 0,104 0,105 0,118 -0,004 -0,117 Urban [0.051]** [0.086] [0.091] [0.159] [0.117] [0.072] [0.075] [0.080] [0.112] [0.150]

0,421 0,111 0,056 -0,294 -0,308 0,475 0,356 0,313 0,151 0,136 Quintile 2 [0.085]*** [0.119] [0.128] [0.194] [0.201] [0.085]*** [0.116]*** [0.103]*** [0.173] [0.174]

0,608 0,208 0,2 -0,3 -0,294 0,694 0,46 0,48 0,224 0,012 Quintile 3 [0.104]*** [0.121]* [0.141] [0.198] [0.189] [0.087]*** [0.130]*** [0.128]*** [0.181] [0.142]

0,891 0,343 0,211 -0,254 -0,14 0,959 0,632 0,486 0,155 0,145 Quintile 4 [0.093]*** [0.122]*** [0.137] [0.202] [0.192] [0.076]*** [0.139]*** [0.127]*** [0.202] [0.191]

1,113 0,488 0,32 -0,116 0,063 1,177 0,724 0,571 0,367 0,371 Quintile 5 [0.096]*** [0.125]*** [0.141]** [0.232] [0.223] [0.087]*** [0.143]*** [0.133]*** [0.226] [0.210]*

0,111 0,091 0,06 0,129 0,182 0,085 0,067 -0,012 -0,061 0,022 Hhd with children [0.054]** [0.055]* [0.074] [0.126] [0.146] [0.052] [0.053] [0.078] [0.128] [0.144]

0,44 0,643 0,689 0,445 0,684 0,441 0,673 0,635 0,647 0,588 Hhd with elderly [0.099]*** [0.102]*** [0.129]*** [0.168]*** [0.214]*** [0.092]*** [0.116]*** [0.116]*** [0.192]*** [0.199]***

0,027 0,083 0,199 -0,256 0,082 0,043 0,096 0,125 0,103 0,578 Hhd with children and elderly [0.140] [0.243] [0.189] [0.510] [0.417] [0.156] [0.238] [0.187] [0.585] [0.480] 0,172 0,027 -0,054 -0,139 -0,036 0,173 0,035 -0,059 -0,071 -0,146 Household size

[0.046]*** [0.062] [0.073] [0.095] [0.111] [0.044]*** [0.063] [0.074] [0.117] [0.109] -1,112 -0,974 -1,37 -0,854 -2,306 -1,155 -1,343 -0,986 -2,085 -0,356 Probability of having insurance

[0.406]*** [0.991] [1.095] [1.327] [1.771] [0.399]*** [1.045] [1.028] [1.574] [1.238] 0,238 0,177 0,579 0,521 0,304 0,231 0,214 0,54 0,539 0,337 Calderón Guardia-Catedral

Noreste [0.184] [0.177] [0.150]*** [0.384] [0.369] [0.177] [0.169] [0.193]*** [0.283]* [0.317] 0,175 0,11 0,375 0,757 0,615 0,131 0,098 0,259 0,795 0,5 Calderón Guardia-Coronado

[0.189] [0.196] [0.224]* [0.462] [0.432] [0.199] [0.194] [0.222] [0.337]** [0.347] -0,114 -0,491 -0,252 -0,71 -1,139 -0,037 -0,406 -0,257 -0,519 -0,829 Calderón Guardia-Clorito Picado [0.317] [0.341] [0.381] [0.480] [0.429]*** [0.272] [0.373] [0.383] [0.318] [0.366]** -0,08 0,095 0,392 0,633 0,166 -0,058 0,199 0,249 0,893 0,573 Calderón Guardia-Jiménez Núñez

[0.253] [0.291] [0.335] [0.477] [0.412] [0.268] [0.314] [0.340] [0.425]** [0.404] 0,183 0,077 0,368 0,137 -0,058 0,208 0,179 0,411 0,465 0,4 Calderón Guardia-Max Peralta

[0.183] [0.145] [0.169]** [0.407] [0.408] [0.152] [0.144] [0.207]** [0.272]* [0.292]

12

Continuation International poverty line Food Quintile 60 80 90 98 99 60 80 90 98 99

-0,047 0,131 0,402 0,095 -0,005 -0,047 0,179 0,416 0,293 0,121 Calderón Guardia-Max Peralta-Catedral Noreste [0.334] [0.226] [0.191]** [0.409] [0.360] [0.295] [0.240] [0.220]* [0.280] [0.291]

0,45 0,462 0,667 0,283 -0,244 0,443 0,586 0,55 0,218 0,091 Calderón Guardia-Max Peralta-Marcial Fallas [0.306] [0.248]* [0.227]*** [0.483] [0.471] [0.250]* [0.290]** [0.259]** [0.334] [0.373]

0,019 -0,091 0,174 0,253 0,03 0,03 -0,104 0,081 0,241 0,171 Calderón Guardia-William Allen [0.188] [0.259] [0.213] [0.505] [0.461] [0.175] [0.217] [0.299] [0.349] [0.377] -0,08 -0,265 0,082 0,013 -0,335 -0,056 -0,189 0,002 0,183 0,015 Calderón Guardia-Tony Facio

[0.175] [0.161]* [0.197] [0.431] [0.416] [0.161] [0.155] [0.237] [0.282] [0.349] -0,004 -0,018 0,573 0,67 0,287 0,02 0,132 0,563 0,468 0,359 Calderón Guardia-Tony Facio-

Guápiles [0.208] [0.285] [0.216]*** [0.482] [0.433] [0.173] [0.242] [0.227]** [0.266]* [0.304] -0,352 -0,485 -0,519 -1,076 -1,514 -0,183 -0,522 -0,565 -1,029 -1,206 Calderón Guardia/México-

Clorito Picado-Jiménez Núñez [0.280] [0.274]* [0.172]*** [0.428]** [0.428]*** [0.277] [0.233]** [0.207]*** [0.278]*** [0.322]*** 0,348 0,331 0,616 0,668 0,066 0,31 0,347 0,584 0,614 0,315 Calderón Guardia/México-

Jiménez Núñez [0.336] [0.359] [0.349]* [0.472] [0.422] [0.283] [0.311] [0.340]* [0.384] [0.435] 0,062 0,314 0,465 0,322 -0,108 0 0,357 0,423 0,092 -0,293 México

[0.602] [0.405] [0.321] [0.438] [0.467] [0.600] [0.431] [0.224]* [0.266] [0.299] -0,122 0,073 0,304 -0,14 0,933 -0,102 0,214 0,29 0,34 0,895 México-Clorito Picado [0.203] [0.321] [0.232] [0.682] [0.784] [0.184] [0.366] [0.203] [0.627] [0.570] 0,097 0,049 0,255 0,351 0,419 0,115 0,082 0,27 0,399 0,293 México-San Vicente

[0.161] [0.128] [0.119]** [0.448] [0.450] [0.157] [0.109] [0.174] [0.276] [0.360] 0,002 -0,067 0,261 0,446 0,087 0,039 -0,026 0,202 0,321 0,224 México-Baltodano

[0.186] [0.178] [0.178] [0.423] [0.451] [0.172] [0.168] [0.209] [0.273] [0.275] -0,094 -0,312 -0,207 0,446 0,33 -0,066 -0,241 -0,014 0,391 0,611 México-Baltodano-Anexión [0.190] [0.275] [0.293] [0.508] [0.455] [0.168] [0.264] [0.310] [0.386] [0.514] -0,086 -0,34 -0,084 0,915 0,457 -0,072 -0,346 -0,116 -0,495 -0,783 México-Baltodano-Upala [0.196] [0.350] [0.423] [0.873] [0.916] [0.191] [0.314] [0.326] [0.307] [0.412]* 0,022 -0,094 0,391 0,393 0,309 0,036 -0,049 0,211 0,622 0,364 México-Mons. Sanabria

[0.177] [0.198] [0.168]** [0.423] [0.454] [0.177] [0.164] [0.213] [0.286]** [0.345] 0,649 0,505 0,576 -0,118 -0,323 0,59 0,1 0,384 -0,138 -0,389 México-Mons. Sanabria-Max

Terán [0.352]* [0.392] [0.342]* [0.444] [0.414] [0.313]* [0.362] [0.332] [0.341] [0.435] -0,005 -0,121 0,084 0,247 -0,083 0,012 -0,077 0,155 0,162 0,072 México-San Carlos [0.185] [0.167] [0.149] [0.418] [0.373] [0.156] [0.185] [0.214] [0.279] [0.372]

13

International poverty line Food Quintile 60 80 90 98 99 60 80 90 98 99

-0,181 -0,931 -0,772 -1,555 -2,265 -0,092 -0,608 -0,848 -1,316 -1,73 México-San Carlos-Los Chiles [0.188] [0.389]** [0.399]* [0.517]*** [0.525]*** [0.195] [0.275]** [0.324]*** [0.348]*** [0.386]*** 0,147 0,237 0,831 0,598 0,178 0,155 0,4 0,649 0,622 0,14 México-San Carlos-San Vicente

[0.670] [0.797] [0.775] [0.720] [0.832] [0.615] [0.831] [0.785] [0.657] [0.697] -0,224 0,421 0,729 -0,251 -0,816 -0,21 0,542 0,949 0,532 -0,118 México-San Carlos/San Rafael

Alajuela-Luis C. Valverde [0.671] [0.684] [0.404]* [0.505] [0.579] [0.694] [0.647] [0.600] [0.559] [0.453] 0,261 0,202 0,337 0,206 0,177 0,284 0,241 0,352 0,489 0,335 México-San Rafael Alajuela-

Marcial Rodríguez [0.177] [0.139] [0.134]** [0.406] [0.373] [0.163]* [0.143]* [0.185]* [0.279]* [0.301] 0,745 0,482 0,496 0,278 0,214 0,657 0,466 0,494 0,296 0,15 Meixco-San Rafael Alajuela-Luis

C. Valverde [0.251]*** [0.168]*** [0.159]*** [0.540] [0.742] [0.184]*** [0.166]*** [0.194]** [0.377] [0.576] 1,051 0,914 0,777 -0,47 -1,04 1,212 0,7 0,582 0,189 -0,488 Meixco-San Rafael Alajuela-

Guápiles/Los Chiles [0.470]** [0.435]** [0.365]** [0.465] [0.488]** [0.438]*** [0.347]** [0.387] [0.406] [0.447] 0,486 0,352 0,569 0,16 -0,029 0,497 0,423 0,401 0,208 -0,029 México/San Juan de Dios-San

Rafael Alajuela-San Francisco Asís [0.229]** [0.198]* [0.159]*** [0.406] [0.377] [0.192]*** [0.169]** [0.204]** [0.269] [0.351]

-0,005 0,557 0,158 -1,153 -1,535 0,021 0,404 0,212 -0,591 -0,872 México/San Juan de Dios-Mons. Sanabria [0.517] [0.630] [0.431] [0.493]** [0.525]*** [0.580] [0.478] [0.374] [0.422] [0.431]**

0,083 -0,006 0,297 0,362 0,773 0,07 0,04 0,256 0,373 0,955 San Juan de Dios [0.188] [0.185] [0.216] [0.582] [0.552] [0.178] [0.163] [0.231] [0.400] [0.527]* -0,136 0,096 0,712 0,146 -0,409 -0,146 0,033 0,574 0,108 -0,226 San Juan de Dios-Carlos Durán [0.398] [0.554] [0.606] [0.572] [0.756] [0.364] [0.539] [0.604] [0.588] [0.629] -0,201 -0,779 0,427 -0,058 -0,513 -0,236 -0,689 0,34 -0,022 -0,307 San Juan de Dios-Clorito Picado [0.438] [0.759] [0.709] [0.575] [0.688] [0.439] [0.787] [0.708] [0.501] [0.649] -0,006 0,065 0,235 0,562 0,171 0,012 0,107 0,212 0,6 0,631 San Juan de Dios-Marcial Fallas [0.191] [0.148] [0.130]* [0.397] [0.381] [0.173] [0.126] [0.177] [0.282]** [0.303]** 0,155 0,048 0,289 0,415 0,281 0,166 0,097 0,198 0,504 0,372 San Juan de Dios-Moreno Cañas

[0.189] [0.148] [0.150]* [0.437] [0.376] [0.199] [0.140] [0.191] [0.287]* [0.261] -0,071 0,092 0,21 0,378 -0,066 -0,074 0,045 0,104 0,551 0,645 San Juan de Dios-Solón Núñez [0.228] [0.264] [0.199] [0.415] [0.367] [0.259] [0.196] [0.294] [0.340] [0.340]*

0,02 0,183 0,22 0,793 0,99 0,049 0,3 0,224 0,625 0,644 San Juan de Dios-Escalante Pradilla [0.232] [0.165] [0.283] [0.646] [0.509]* [0.236] [0.171]* [0.251] [0.328]* [0.321]**

0,042 -0,01 0,31 0,142 0,289 0,052 0,216 0,259 0,114 0,262 San Juan de Dios-Escalante Pradilla-Ciudad Neily [0.225] [0.258] [0.198] [0.495] [0.489] [0.220] [0.176] [0.221] [0.338] [0.353]

-0,036 -0,056 0,003 -0,005 -0,507 -0,016 0,475 0,388 0,167 -0,276 San Juan de Dios-Escalante Pradilla-Golfito [0.252] [0.361] [0.443] [0.423] [0.447] [0.280] [0.510] [0.376] [0.339] [0.360]

14

Continuation International poverty line Food Quintile 60 80 90 98 99 60 80 90 98 99

0,071 -0,391 0,379 0,75 0,323 0,07 -0,329 0,387 0,495 0,373 San Juan de Dios-Tomás Casas [0.232] [0.547] [0.747] [1.009] [1.107] [0.242] [0.442] [0.644] [0.646] [0.647] 0,119 0,075 0,397 1,18 0,184 0,133 0,281 0,324 0,589 0,278 Hhd with electrical energy

[0.145] [0.389] [0.374] [0.715]* [0.997] [0.132] [0.311] [0.363] [0.583] [0.554] 0,287 0,428 0,231 0,032 0,119 0,287 0,472 0,25 0,625 0,259 House with access to improved

sanitation [0.084]*** [0.215]** [0.222] [0.410] [0.503] [0.091]*** [0.210]** [0.199] [0.306]** [0.339] 0,002 -0,106 -0,18 0,086 0,032 -0,006 -0,081 -0,156 -0,369 -0,043 House with access to improved

water [0.082] [0.148] [0.194] [0.243] [0.269] [0.101] [0.157] [0.136] [0.292] [0.277] -0,096 -0,077 -0,135 -0,339 -0,571 -0,093 -0,128 -0,165 -0,072 -0,208 House with access to waste

disposal [0.057]* [0.103] [0.133] [0.195]* [0.183]*** [0.053]* [0.110] [0.093]* [0.166] [0.160] 0,018 0,012 0,004 -0,004 -0,003 0,016 0,012 0,004 -0,003 -0,004 Years of schooling of head

[0.022] [0.027] [0.028] [0.022] [0.024] [0.024] [0.024] [0.011] [0.019] [0.031] -0,145 -0,27 -0,185 -0,063 -0,443 -0,116 -0,263 -0,267 -0,056 -0,117 Share of HE financed with no

cash [0.066]** [0.154]* [0.156] [0.189] [0.200]** [0.070] [0.113]** [0.152]* [0.208] [0.222] 0,264 1,582 2,394 2,638 5,18 0,265 1,446 1,976 3,375 2,843 Constant

[0.270] [0.554]*** [0.630]*** [0.960]*** [1.200]*** [0.272] [0.561]** [0.624]*** [0.880]*** [0.763]*** Pseudo R2 0.0845 0.0488 0.0542 0.0788 0.1104 0.0928 0.0645 0.01714 0.0906 0.1058 Note: * significant at 10%, ** at 5% and at 1%.*** Standard errors in parenthesis.

Source: authors´ calculation with ENIG 2004 and geographic data

15

Table A2.4 Quintile regressions for Health Financial Burden of CP with waiting list: Food and International poverty line

International poverty line Food poverty line Quintile 60 80 90 98 99 60 80 90 98 99

0,068 0,098 0,047 0,183 0,074 0,057 0,104 0,076 0,092 0,101 Urban area [0.063] [0.068] [0.103] [0.129] [0.153] [0.057] [0.069] [0.087] [0.105] [0.133] 0,459 0,138 0,063 -0,401 -0,414 0,503 0,323 0,324 0,207 -0,014 Quintile 2

[0.087]*** [0.117] [0.118] [0.168]** [0.197]** [0.096]*** [0.112]*** [0.113]*** [0.154] [0.172] 0,69 0,166 0,135 -0,366 -0,658 0,795 0,447 0,477 0,133 -0,068 Quintile 3

[0.101]*** [0.130] [0.131] [0.173]** [0.205]*** [0.102]*** [0.122]*** [0.112]*** [0.160] [0.159] 0,948 0,286 0,119 -0,287 -0,445 1,036 0,564 0,421 0,289 0,138 Quintile 4

[0.089]*** [0.136]** [0.131] [0.188] [0.209]** [0.088]*** [0.133]*** [0.112]*** [0.164]* [0.191] 1,203 0,45 0,263 -0,099 -0,229 1,269 0,686 0,553 0,399 0,178 Quintile 5

[0.093]*** [0.132]*** [0.131]** [0.193] [0.227] [0.088]*** [0.137]*** [0.110]*** [0.174]** [0.240] 0,119 0,054 -0,009 0,013 0,251 0,102 0,033 -0,062 -0,073 -0,027 Hhd with children

[0.052]** [0.056] [0.090] [0.162] [0.174] [0.045]** [0.045] [0.078] [0.113] [0.162] 0,475 0,592 0,647 0,607 0,413 0,469 0,674 0,621 0,44 0,429 Hhd with elderly

[0.091]*** [0.126]*** [0.117]*** [0.163]*** [0.193]** [0.098]*** [0.108]*** [0.092]*** [0.169]*** [0.180]** -0,061 0,048 0,047 -0,187 0,424 -0,087 0,068 0,008 -0,223 0,657 Hhd with children and

elderly [0.192] [0.226] [0.189] [0.661] [0.505] [0.171] [0.185] [0.201] [0.562] [0.608] 0,182 -0,007 -0,081 -0,042 -0,174 0,188 0,009 -0,057 -0,078 -0,138 Household size

[0.041]*** [0.066] [0.065] [0.091] [0.112] [0.037]*** [0.060] [0.062] [0.084] [0.117] -1,245 -0,318 -0,946 -1,93 -0,039 -1,276 -0,972 -0,64 -1,152 -0,351 Probability of having health

insurance [0.408]*** [1.093] [1.025] [1.371] [1.780] [0.402]*** [0.865] [0.790] [1.258] [1.254] 0,188 0,239 0,254 1,066 -0,071 0,195 0,224 0,267 0,621 -0,063 House with electrical energy

[0.128] [0.387] [0.409] [0.816] [0.940] [0.108]* [0.331] [0.284] [0.607] [0.521] 0,345 0,291 0,215 -0,111 -0,317 0,335 0,463 0,3 0,555 0,277 House with access to

improved sanitary conditions [0.082]*** [0.258] [0.221] [0.593] [0.583] [0.079]*** [0.190]** [0.151]** [0.282]** [0.358] -0,05 -0,177 -0,169 -0,32 -0,271 -0,063 -0,072 -0,221 -0,393 -0,478 House with access to

improved water [0.106] [0.181] [0.174] [0.280] [0.315] [0.111] [0.156] [0.136] [0.266] [0.242]** -0,077 -0,124 -0,104 -0,262 -0,173 -0,07 -0,23 -0,183 -0,148 -0,107 House with access to waste

disposal [0.067] [0.079] [0.103] [0.169] [0.204] [0.062] [0.080]*** [0.091]** [0.136] [0.168] 0,018 0,01 0,004 -0,005 -0,013 0,016 0,011 0,002 -0,004 -0,01 Years of schooling of head

[0.029] [0.025] [0.028] [0.029] [0.043] [0.030] [0.029] [0.020] [0.018] [0.029]

16

Continuation

International poverty line Food poverty line Quintile 60 80 90 98 99 60 80 90 98 99

-0,182 -0,252 -0,142 -0,117 -0,194 -0,164 -0,223 -0,203 -0,03 -0,074 Share of health expenditures financed with no cash [0.069]*** [0.147]* [0.153] [0.189] [0.246] [0.059]*** [0.124]* [0.154] [0.222] [0.270]

0,008 0,014 0,001 -0,008 0,009 0,008 0,012 0,013 0,011 0,004 No of cases in waiting list for surgery [0.006] [0.009] [0.012] [0.017] [0.020] [0.006] [0.010] [0.011] [0.014] [0.017]

-0,024 -0,026 -0,047 -0,045 -0,033 -0,019 -0,033 -0,049 -0,056 -0,036 No of cases in waiting list for medical consultations [0.011]** [0.016]* [0.017]*** [0.024]* [0.023] [0.010]* [0.015]** [0.016]*** [0.020]*** [0.023]

0,021 0,024 0,035 0,033 0,04 0,019 0,03 0,035 0,057 0,037 No of cases in waiting list for procedures [0.008]*** [0.011]** [0.014]** [0.020]* [0.022]* [0.006]*** [0.011]*** [0.014]** [0.017]*** [0.017]**

0,339 1,203 2,78 4,485 4,429 0,322 1,458 2,171 3,053 3,817 Constant [0.242] [0.592]** [0.599]*** [0.911]*** [0.974]*** [0.222] [0.479]*** [0.458]*** [0.708]*** [0.817]***

Pseudo R2 0.0710 0.0354 0.0393 0.0488 0.0689 0.0795 0.0514 0.0586 0.0652 0.0688 Note: * significant at 10%, ** at 5% y *** at 1%. Standard errors in parenthesis.

Source: own authors´ calculation with ENIG 2004 and geographic information

17

Table A2.5 Tobit for Oops and HFB to different measures of CP: using waiting lists

Health Financial Burden Oops Food National International Endogenous

0,243 0,066 0,089 0,069 0,068 Urban [0.199] [0.052] [0.053]* [0.051] [0.055] 2,924 0,479 0,32 0,373 0,306 Quintile 2 [0.365]*** [0.085]*** [0.087]*** [0.082]*** [0.092]*** 4,402 0,702 0,482 0,559 0,47 Quintile 3 [0.431]*** [0.095]*** [0.099]*** [0.097]*** [0.108]*** 6,339 0,976 0,751 0,838 0,737 Quintile 4 [0.468]*** [0.108]*** [0.113]*** [0.106]*** [0.122]*** 9,199 1,324 1,108 1,2 1,093 Quintile 5 [0.450]*** [0.110]*** [0.115]*** [0.109]*** [0.126]*** 1,162 0,188 0,214 0,208 0,24 Hhd with children [0.221]*** [0.042]*** [0.051]*** [0.042]*** [0.052]*** 0,949 0,419 0,414 0,39 0,447 Hhd with elderly [0.390]** [0.093]*** [0.107]*** [0.096]*** [0.107]*** -0,712 -0,077 -0,088 -0,091 -0,074 Hhd with children and elderly [0.921] [0.188] [0.208] [0.188] [0.206] 1,4 0,147 0,165 0,144 0,128 Household size [0.206]*** [0.050]*** [0.053]*** [0.049]*** [0.059]** -1,641 -0,604 -0,639 -0,46 -0,719 Probability of having insurance [3.039] [0.799] [0.873] [0.832] [0.962] 0,711 0,237 0,153 0,202 0,229 House with electrical energy [1.123] [0.250] [0.297] [0.271] [0.310] 1,837 0,417 0,425 0,339 0,397 House with access to improved

sanitary conditions [0.683]*** [0.166]** [0.189]** [0.183]* [0.202]** 0,007 -0,087 -0,109 -0,074 -0,078 House with access to improved

water [0.500] [0.108] [0.125] [0.109] [0.128] -0,286 -0,11 -0,096 -0,105 -0,124 House with access to waste disposal [0.303] [0.067] [0.076] [0.071] [0.079] 0,053 0,009 0,011 0,01 0,012 Years of schooling of head [0.048] [0.011] [0.011] [0.011] [0.012] -0,998 -0,227 -0,212 -0,23 -0,237 Share of HE that is financed with no

case [0.346]*** [0.066]*** [0.079]*** [0.068]*** [0.076]*** 0,069 0,014 0,016 0,013 0,018 Number of cases in waiting list in

surgery [0.024]*** [0.005]*** [0.006]*** [0.006]** [0.006]*** -0,097 -0,026 -0,025 -0,025 -0,027 Number of cases in waiting list in

medical consultations [0.038]** [0.009]*** [0.009]*** [0.009]*** [0.009]*** 0,131 0,032 0,031 0,031 0,034 Number of cases in waiting list in

procedures [0.034]*** [0.007]*** [0.008]*** [0.007]*** [0.008]*** -5,109 -0,664 -0,433 -0,603 -0,341 Constant [2.000]** [0.466] [0.530] [0.467] [0.572]

Observations 4231 4231 4231 4231 4231 Pseudo R-2 0,05 0,05 0,03 0,04 0,03 Log of pseudo-likelihood -9822,95 -5531,87 -5763,02 -5559,88 -5843,47 Nota: * significant at 10%, ** 5% and *** 1%. Standard errors in parenthesis

18

Table A2.6

Tobit for Oops and HFB to different measures of CP: geographic variables

Poverty Line OOPS Food National International Endogenous

0,356 0,102 0,126 0,101 0,113 Urban area [0.247] [0.066] [0.070]* [0.066] [0.075] 2,908 0,475 0,316 0,369 0,3 Quintile 2 [0.355]*** [0.082]*** [0.083]*** [0.078]*** [0.089]*** 4,326 0,686 0,465 0,543 0,45 Quintile 3 [0.430]*** [0.094]*** [0.098]*** [0.095]*** [0.109]*** 6,297 0,968 0,743 0,829 0,726 Quintile 4 [0.466]*** [0.106]*** [0.111]*** [0.103]*** [0.121]*** 9,108 1,311 1,094 1,185 1,077 Quintile 5 [0.452]*** [0.109]*** [0.114]*** [0.107]*** [0.126]*** 1,205 0,2 0,226 0,22 0,253 Hhd with children [0.228]*** [0.044]*** [0.052]*** [0.044]*** [0.054]*** 0,94 0,424 0,421 0,397 0,452 Hhd with elderly [0.396]** [0.095]*** [0.108]*** [0.098]*** [0.109]*** -0,686 -0,057 -0,062 -0,067 -0,048 Hhd with children and elderly [0.925] [0.191] [0.210] [0.190] [0.208] 1,388 0,145 0,164 0,143 0,127 Household size [0.202]*** [0.049]*** [0.052]*** [0.049]*** [0.058]** -1,869 -0,701 -0,761 -0,571 -0,835 Prob of having health insurance [3.085] [0.816] [0.881] [0.850] [0.975] 0,376 0,164 0,05 0,128 0,138 Calderón Guardia-Catedral Noreste [1.130] [0.251] [0.296] [0.274] [0.309] 1,879 0,431 0,441 0,354 0,409 Calderón Guardia-Coronado [0.697]*** [0.168]** [0.193]** [0.185]* [0.205]** 0,175 -0,044 -0,059 -0,024 -0,027 Calderón Guardia-Clorito Picado [0.501] [0.106] [0.123] [0.107] [0.126] -0,39 -0,122 -0,116 -0,121 -0,148 Calderón Guardia-Jiménez Núñez [0.277] [0.059]** [0.067]* [0.062]* [0.072]** 0,049 0,009 0,011 0,01 0,011 Calderón Guardia-Max Peralta [0.046] [0.010] [0.011] [0.011] [0.012] -1,065 -0,236 -0,225 -0,241 -0,249 Calderón Guardia-Max Peralta-

Catedral Noreste [0.352]*** [0.068]*** [0.080]*** [0.070]*** [0.077]*** -0,121 0,054 0,055 0,063 0,044 Calderón Guardia-Max Peralta-

Marcial Fallas [0.039]*** [0.009]*** [0.009]*** [0.009]*** [0.010]*** -0,115 0,03 0,026 0,044 0,02 Calderón Guardia-William Allen [0.054]** [0.011]*** [0.012]** [0.011]*** [0.012]* 0,098 -0,141 -0,179 -0,178 -0,172 Calderón Guardia-Tony Facio [0.122] [0.023]*** [0.024]*** [0.023]*** [0.025]*** 0,19 0,124 0,083 0,09 0,085 Calderón Guardia-Tony Facio-

Guápiles [0.135] [0.032]*** [0.034]** [0.032]*** [0.035]** 0,384 0,062 0,061 0,063 0,063 Calderón Guardia/México-Clorito

Picado-Jiménez Núñez [0.018]*** [0.003]*** [0.004]*** [0.003]*** [0.004]*** 1,083 0,371 0,392 0,366 0,429 Calderón Guardia/México-Jiménez

Núñez [0.286]*** [0.066]*** [0.075]*** [0.070]*** [0.075]***

19

Continuation

Poverty Line Oops Food National International Endogenous

-0,562 -0,076 -0,087 -0,085 -0,054 México [0.221]** [0.057] [0.058] [0.056] [0.060] -1,375 -0,253 -0,261 -0,254 -0,284 México-Clorito Picado [0.195]*** [0.046]*** [0.049]*** [0.047]*** [0.050]*** -0,538 -0,003 -0,027 0,019 -0,024 México-San Vicente [0.229]** [0.054] [0.057] [0.055] [0.059] -1,187 -0,46 -0,493 -0,461 -0,5 México-Baltodano [0.068]*** [0.015]*** [0.017]*** [0.015]*** [0.017]*** 1,241 0,302 0,309 0,325 0,319 México-Baltodano-Anexión [0.089]*** [0.019]*** [0.020]*** [0.020]*** [0.021]*** -0,537 0,057 0,057 0,068 0,056 México-Baltodano-Upala [0.056]*** [0.013]*** [0.014]*** [0.013]*** [0.016]*** -0,492 -0,081 -0,077 -0,087 -0,115 México-Mons. Sanabria [0.069]*** [0.014]*** [0.015]*** [0.013]*** [0.016]*** -0,106 0,022 0,021 0,032 0,024 México-Mons. Sanabria-Max Terán [0.071] [0.017] [0.018] [0.017]* [0.018] -0,787 -0,067 -0,096 -0,064 -0,123 México-San Carlos [0.128]*** [0.031]** [0.033]*** [0.033]** [0.035]*** -1,556 -0,314 -0,38 -0,356 -0,391 México-San Carlos-Los Chiles [0.216]*** [0.052]*** [0.056]*** [0.055]*** [0.060]*** -2,285 -0,441 -0,549 -0,387 -0,491 México-San Carlos-San Vicente [0.316]*** [0.070]*** [0.079]*** [0.072]*** [0.083]*** -0,601 -0,082 -0,086 -0,065 -0,11 México-San Carlos/San Rafael

Alajuela-Luis C. Valverde [0.135]*** [0.033]** [0.034]** [0.033]** [0.036]*** -0,814 -0,153 -0,156 -0,143 -0,162 México-San Rafael Alajuela-Marcial

Rodríguez [0.161]*** [0.038]*** [0.040]*** [0.038]*** [0.041]*** -5,877 -1,078 -1,237 -1,146 -1,316 México-San Rafael Alajuela-Luis C.

Valverde [0.292]*** [0.072]*** [0.077]*** [0.077]*** [0.081]*** -0,479 -0,038 -0,049 -0,034 -0,055 México-San Rafael Alajuela-

Guápiles/Los Chiles [0.251]* [0.067] [0.068] [0.067] [0.071] 0,162 0,178 0,114 0,1 0,09 México/San Juan de Dios-San

Rafael Alajuela-San Francisco Asís [0.381] [0.087]** [0.096] [0.090] [0.100] 0,281 0,153 0,128 0,137 0,132 México/San Juan de Dios-Mons.

Sanabria [0.129]** [0.030]*** [0.032]*** [0.030]*** [0.033]*** 0,591 0,303 0,342 0,316 0,342 San Juan de Dios [0.175]*** [0.040]*** [0.044]*** [0.042]*** [0.045]*** 5,024 1,12 1,365 1,107 1,14 San Juan de Dios-Carlos Durán [0.312]*** [0.072]*** [0.079]*** [0.074]*** [0.083]*** 0,913 0,312 0,354 0,317 0,397 San Juan de Dios-Clorito Picado [0.172]*** [0.042]*** [0.044]*** [0.041]*** [0.045]*** 0,198 0,087 0,092 0,11 0,124 San Juan de Dios-Marcial Fallas [0.362] [0.083] [0.091] [0.085] [0.093] 0,206 0,001 -0,001 0,02 -0,004 San Juan de Dios-Moreno Cañas [0.051]*** [0.011] [0.011] [0.011]* [0.012] -2,611 -0,423 -0,43 -0,388 -0,424 San Juan de Dios-Solón Núñez [0.082]*** [0.022]*** [0.025]*** [0.022]*** [0.025]***

20

Continuation

Poverty Line Oops Food National International Endogenous

0,909 0,01 0,016 0,028 0,012 San Juan de Dios-Escalante Pradilla [0.075]*** [0.017] [0.018] [0.017]* [0.019] -1,067 -0,127 -0,136 -0,124 -0,146 San Juan de Dios-Escalante Pradilla-

Ciudad Neily [0.070]*** [0.014]*** [0.015]*** [0.014]*** [0.015]*** 0,221 0,078 0,083 0,09 0,079 San Juan de Dios-Escalante Pradilla-

Golfito [0.067]*** [0.016]*** [0.017]*** [0.016]*** [0.018]*** -1,543 -0,215 -0,207 -0,186 -0,203 San Juan de Dios-Tomás Casas [0.078]*** [0.016]*** [0.018]*** [0.016]*** [0.018]*** 0,006 0,1 0,059 0,104 0,063 House with electrical energy [0.232] [0.051]** [0.056] [0.052]** [0.058] -0,245 -0,027 0,049 -0,018 0,004 House with access to improved

sanitation [0.141]* [0.036] [0.038] [0.035] [0.039] -0,258 0 -0,1 -0,065 -0,089 House with access to improved

water [0.391] [0.088] [0.101] [0.090] [0.106] 0,638 0,094 0,194 0,094 0,176 House with access to waste disposal [0.368]* [0.090] [0.096]** [0.090] [0.101]* -1,267 -0,184 -0,192 -0,174 -0,202 Years of schooling of head [0.063]*** [0.014]*** [0.016]*** [0.015]*** [0.017]*** 1,753 0,382 0,507 0,527 0,567 Share of health expenditures that is

financed with no cash [0.072]*** [0.077]*** [0.073]*** [0.082]*** -3,776 -0,449 -0,151 -0,391 -0,048 constant [2.114]* [0.495] [0.558] [0.505] [0.607]

N 4231 4231 4231 4231 4231 Pseudo R-2 0,06 0,05 0,04 0,04 0,03 Log of pseudolikelihood -9793,15 -5501,26 -5729,9 -5528,93 -5812,23 Note: * significant at 10%, ** 5% and *** 1%. Standard errors in parenthesis.

Source: authors´calculation with ENIG 2004 and geographic information

Out-of-pocket and Catastrophic Health Expenditures Puzzle… · · 2014-07-04Out-of-pocket and Catastrophic Health Expenditures Puzzle: ... Out-of-pocket and Catastrophic Health

Documents