(LUẬN văn THẠC sĩ) inequity in household health care finance in vietnam

TRAN NGOC THANH INEQUITY IN HOUSEHOLD HEALTH CARE FINANCE IN VIETNAM MASTER OF ART IN DEVELOPMENT ECONOMICS SPECIALIZATION IN HEALTH ECONOMICS AND MANAGEMENT Ho Chi Minh City – 2015.

TRAN NGOC THANH

INEQUITY IN HOUSEHOLD HEALTH CARE FINANCE

IN VIETNAM

MASTER OF ART IN DEVELOPMENT ECONOMICS

(SPECIALIZATION IN HEALTH ECONOMICS AND MANAGEMENT)

Ho Chi Minh City – 2015

TRAN NGOC THANH

INEQUITY IN HOUSEHOLD HEALTH CARE FINANCE

IN VIETNAM

Major : Economics Code : 60310105

MASTER OF ART IN DEVELOPMENT ECONOMICS

(SPECIALIZATION IN HEALTH ECONOMICS AND MANAGEMENT)

Advisor: Dr Pham Khanh Nam

Dr ArdeshirSepehri

Ho Chi Minh City – 2015

I swear this is my research The data and conclusions of Research presented in this

thesis are honest and have not been published in other research

I am responsible for my research

Student

Tran Ngoc Thanh

1.1 Background 1

1.2 Research Objectives 3

1.3 Data source 3

1.4 Study Design 3

CHAPTER 2: LITERATURE REVIEW 4

2.1 Definition 4

2.1.1 Social equity 4

2.1.2 Equity in health care 5

2.1.3 Inequality and Inequity 6

2.1.4 Vertical equity and Horizontal equity 9

2.1.5 Ability to pay – ATP 9

2.2 Concentration index and Concentration curve 9

2.3 1 Concentration index 9

2.3 2 Concentration curve 10

2.3 Katwani indices and Concentration curves 10

2.4 Inequity or Progressivity of health care finance 12

2.5 Decomposition 13

2.6 Review emperical studies about health equity finance 14

CHAPTER 3: METHODOLOGY 19

3.1 Analytical framework 19

3.2 Model 20

3.3 Data 24

3.4 Variables 24

CHAPTER 4: RESULTS 27

4.1 Vietnam Health Care System 27

4.2 Delivery of Health care 28

4.3 Financing of Health Care 29

4.4 Results 32

4.4.1 OLS and Quantile Regression of Household Total expenditure 32

4.4.2 Average Per household Health Finance, Shares of Total Financing 35

4.4.3 Distributional Incidence of Sources of Household Health Finance 39

4.4.4 Decomposition inequality of Household Total expenditure 43

4.4.5 Decomposition inequality of Health Care 43

4.4.6 Concentration Curves 47

4.4.7 Distribution of Health Payments 48

4.5 Compare with international studies 49

4.6 Discusion 50

CHAPTER 5: CONCLUSION AND POLICY IMPLICATION 51

5.1 Conclusion 51

5.2 Policy implication 52

5.3 Limitation 53 REFERENCE

LIST OF ABBREVIATIONS

CONTENT OF TABLE

Table 1: Some brief definitions 5

Table 2: The magnitude of inequality based on the value of CIs 10

Table 3: Summary formulae analyzing inequity 20

Table 4: Variables of socioeconomic factors and expenditures 25

Table 5: Health Expenditure in Vietnam 31

Table 6: OLS and Quantile Regression of Household Total expenditure 33

Table 7: Average Per household Health Finance (‘000 VND) and Shares of Total Financing (%) 37

Table 8: Distributional Incidence of Sources of Household Health Finance in Vietnam, 2012 and 2010 41

Table 9: Decomposition inequality of Household Total expenditure 46

Table 10: Decomposition inequality of Health Care 45

Table 11: Compared results with international studies 49

CONTENT OF FIGURE

Figure 1: Social determinants of health and health equity 6

Figure2: Health inequality vs health inequity 7

Figure3: Process to analyze inequity 8

Figure4: Three dimensions of health coverage 8

Figure5: Lorenz curve for prepayment income and concentration curve for health care payment 11

Figure6: Framework of analysing inequity 19

Figure7: The structure of health care system in Vietnam 28

Figure8: Channels of financing sources for Viet Nam health care system 30

Figure9: Social Insurance Contribution, Inpatient and Outpatient payments, Out-of-pocket for health care 47

Figure10: Health Payment Shares by Quintiles 49

CHAPTER 1: INTRODUCTION

Equity is one of the most important problems on the world, especially in health care finance Many countries are working to establish a health financing system that allows them promote, prevent, curate and rehabilitate health interventions for all at

an affordable cost – thereby achieving equity in access and financial riskprotection

as well as in health financing (WHO, 2005) Moreover, this is particularly challenging for low- and middle-income countries in light of their heavy reliance on out-of-pocket (OOP) payments for health care (WHO, 2010) Viet Nam is a developing country, withoutthe exception The challenge is to improve the health financing system in order to achieve universal coverage asan overall policy goal Equitable financing is a key objective of health care systems Its importance is evidenced in policy documents, policy statements, the work of health economists and policy analysts The financing of health care is a subject of major concern throughout the world The conventional categorisations of finance source for health care are taxation, social health insurance, and out-of-pocket payments An understanding of the equity implications would help policy makers in achieving equitable financing

The main purpose of this research was to comprehensively assess the equity of health care financing in Vietnam, which represents a new country context for the quantitative techniques used In this research, author uses the concentration index to assess inequality and Katwani index to assess the inequity of health care finance The study evaluated each of the four financing sources (outpatient and inpatient expenditures, health insurance,out-of-pocket payments) independently, and subsequently by combined the financing sources to evaluate the whole financing system The author also assesses inequality of expenditure only in health care and total expenditure including food and non-food expenditures of households Moreover, the author also uses additional methodology to assess that which sources

mostly affect inequity of health care finance by applying the method decomposition

of expenditure Definition of Equity involves a value judgment of fairness on the variations from the equality in the population Equity in health care financing is assessed by the degree of inequality in paying for health care between households of unequal Ability To Pay (ATP) (Doorslaer, Wagstaff, 1993), ATP is the factor used to evaluate inequity of health care finance system– ATP can be measured by the total expenditure of household, including food, non-food payments and healthcare expenses

To strengthen the important of health care finance related to ATP, many studies have used ATP to evaluate the inequity in health care field such as ATP in Denmark and the UK; Ireland, Portugal and Spain; Italy and the Netherlands; and tax financing in Switzerland Furthermore, the accordance of health payments to ATP is regarded as an important objective in the finance of health care in Belgium, France, Germany, the Netherlands Policy makers in various countries are seen to commit towards financing health care according to ATP

Kakwani (1997), Doorslaer (1997,2000),Doorslaer andMasseria(2004), Wagstaffand Doorslaer (1993, 1997),Wagstaff(2002)have studied income- relatedinequalityinhealthcareutilization, equity in health care delivery, equity in health care finance, and inequalities in health by using ATP

The Ministry of Health (MOH) in Vietnam also agree to use the new national health financing scheme be related to ATP (PAHE, 2011)

With all reasons above, the author also uses ATP tomeasureandexplaininequality and inequity in health care finance in Vietnam

In summary, this study usetheconcentrationindex and Kakwani indexforthe measurement of ATPinequality and inequity in health care finance proposedbyWagstaffandDoorslaer (2000) to assess whether there are inequity and inequality in health care system andwhichfactors affect mostly to the inequity in health care finance system in Vietnam

1.2 Research Objectives

This study presents an inequity assessment of the health financing system, and draws together all finance sources in Vietnam to evaluate the whole financing system The general objective is to analyzethe inequity of health care finance with quintile of ability-to-pay of Vietnam households Specific objectives are:

1 To calculate the inequality indices (CIs) and the inequity indices (Katwani indices)of healthcare finance variables of households such astotal expenditure, health payments, out-of-pocket for health, food or non-food payments

2 To decompose the inequality of households’ totalexpenditure and total health expenditure

3 To calculate the factors affect to total expenditure or ATP through both OLS and Quantile regression models

Finally, Chapter 5 briefly discusses the conclusions, policy implications and

limitation of this study

CHAPTER 2: LITERATURE REVIEW

2.1.1 Social equity

Today, there are many definitions about equity of different schools, here are some perspectives:

 Libertarians emphasize a respect for natural rights, focusing in particular on two

of the rights: rights to life and to possessions

 Utilitarians aim at maximizing the sum of individual utilities or welfare, though some utilitarian writers have incorporated a concern for individual autonomy into this maximand

 Rawlsians (1971) proposes two principles of social justice, namely that individuals should havethe maximal liberty compatible with the same degree of liberty for everyone and that deliberate inequalities are unjust unless they work

to the advantage of the least well off

 Marxists emphasize “needs”, principle of “distribution according to need” And this principle is can be interpreted as “from each according to his ability to pay” Health equity also has many perspectives of many different reseachers and institutions on the world, specific described at Table 1:

Table 1: Some brief definitions No

1 Mooney 1983 (and others

Horizontal equity requires equal treatment for equal need

Vertical equity: different treatment for different need

2 Aday 1984 Health care is equitable when resource allocation

and access are determined by health needs

3 Whitehead 1990, 1992 Health inequities are differences in health that are

avoidable, unjust and unfair

4 Culyer &Wagstaff

1993

Equity in health care means equal utilization, distribution according to need, equal access and equal health outcomes

5

International Society for Equity in Health (ISEqH), 2005

Health equity is the absence of systematic and potentially remediable differences in one or more aspects of health across populations or population subgroups defined socially, economically,

demographically or geographically

“Health Equity is the absence of potentially avoidable differences in health (or health risks that policy can influence) between groups of people who aremore and less advantaged socially”

Source: Braveman, 2006, forthcoming PAHE 2013

2.1.2 Equity in health care

Hurst (1985) studies of inequity in health care finance have tended to take as their starting point the premise that health care ought to be financed according to ability to pay

The egalitarians who are concerned to ensure that health care is financed according to ability to pay and that the delivery of health care is organized in such a way that everyone enjoys the same access to care and that the care is allocated on the basis of need with a view to promoting equality of health

The general picture of health care finance which was affected by many determinants such as individual lifestype factors, social and community networks,

and general socio-economic, cultural and environmental conditions The detailed was described by WHO in Figure 1 as below

Figure 1: Social determinants of health and health equity

2.1.3 Inequality and Inequity

Theterminequalityinhealthisdifferentthantheterminequityinhealth

Actually, inequalities in health are based upon observed differences on disparities

on health.Health inequalies are differences in health outcomes and their determinants between segments of the population, as defined by social, demographic, environmental, and geographic attributes

On the other hand, inequities in health are based on ethical judgments about the fairness of the differences.Health inequity refers to those inequalities in health that are deemed to be unfair or stemming from some form of injusticeor“the absence of potentially avoidabledifferences in health between groups of people who are more and less advantaged socially” (PAHE, 2013)

An example of health inequality is the higher incidence of illness among the elder people as compared withyoung people Thisisanunavoidablephenomenon(dueto

biologicalorigin)anddoesnotimplyamoraljudgment.However,ifthereexistsahigher incidence of illness among the poor elderly as compared with that amongthenon- poor elderly, thenthisreferstoasavoidableinequalityorinequity(determinedbysocio- economicfactors, etc) The distinction between health inequality and health inequity

Figure 3: Process to analyze inequity

Source: PAHE, 2013

The purpose of decreasing inequity in health care finance is to reduce the gap between the public expenditures and personal expenditures for health care To carry

on this aim we must manage and control three perspectives like Figure 4

Figure 4: Three dimensions of health coverage

Source: PAHE, 2013

2.1.4 Vertical equity and Horizontal equity

Vertical equity: persons or families of unequal ability to pay making appropriately dissimilar payments for health care, and

Horizontal equity: persons or families of the same ability to pay making the same contribution Horizontal equity also can be defined in terms of the extent to which those of equal ability to pay actually end up making equal payments, regardless of, for example, gender, marital status, trade union membership, place of residence, etc

In this study, the author analyzes inequality and inequity in health care finance mostly based on vertical equity perspective

2.1.5 Ability to pay – ATP

In a developing-country context, like Viet Nam, given the lack of organized labor markets and the high variability of incomes over time, household consumption (or at least expenditure) is generally considered to be a better measure of welfare and ability to pay than income With the objective of this thesis is simply to assess the degree of proportionality between health payments and some measure of living standards, then household expenditures gross can be used [20] Therefore, in this study ability to pay is typically total household consumption, including all payments toward health care

2.3 1 Concentration index

Wagstaff (1991) published a paper on the measurement of inequalities in health

Theprimaryobjectivesofthispaperwere:(1)toprovideacriticalreviewofthevarious measures of inequality in health, (2)toidentifywhichmeasuresarebestsuitedtomeasurehealthinequality.This paper

identified the three measures of inequality, namely: (a) the range, (b) the Gini coefficient (Lorenz curve), and (c) the concentration index (concentration curve) Theconcentrationindex(CI)providesameasureofthemagnitudeofinequality.Itis defined as twice the area between the concentration curve and the line of

equality(Figure 5) The index has amagnitudebetweenminusoneandplusone,andtakesthevalueofzerowhenthereisno

socioeconomicinequality.Theconventionisthattheindextakesanegativevaluewhenthe concentrationcurveliesabovethelineofequalityandittakesapositivevaluewhenthe concentration lies below the line of equality The absolute value of CI measures the magnitude of socio-economic inequality, the larger the absolute value of CI, the greater the disparity

Conventionally, a concentration index of less than 0.2 indicates a low magnitude

of inequality A concentrationindexofbetween0.2and0.39impliesamoderatemagnitudeofinequality.A

concentration index of between 0.4 and 0.6denotes a highmagnitude of inequality; it marksthe threshold at which inequalityshould be treatedas a matter of urgency.A concentration index of 0.6 or higher reflects a very high magnitude of inequality

(See Table 2) (forthcoming PAHE, 2013)

Table 2: The magnitude of inequality based on the value of CIs

Kakwani (1997)clarifiedtherelationshipbetweentwowidelyusedindicesof health inequality namely: the relative index of inequality (RII) and the concentration index (CI) and explained why these are superior to the other indices used in the literature For example, the

CIissensitivetosocioeconomicdimensionofinequalitiesinhealthbecauseitsvaluelies between-1to1.ApositiveCIrepresentsthepro-richandanegativeCIrepresentspro-poor

inequality in health

Figure 5: Lorenz curve for prepayment income and concentration

curve for health care payment

Source: Handbook of Health Economics

L pre (p) is the Lorenz curve for pre-payment income

L pay (p) is the payment concentration curve, which plots the cumulative proportion

of the population [ranked according to pre-payment income as with L pre (p)] against the cumulative proportion of health care payments

The degree of progressivity can therefore be assessed by looking at the size of the area between L pre (p) and L pay (p) If G pre is the Gini coefficient for pre-payment income, and C pay is the concentration index for payments,

Kakwani's index of progressivity, K or π K , is defined as : π K = C pay – G pre

Many questions must be answered when analyzing the inequity health care finance, every problem contributes its role Example,

 Who pays for health care?

 To what extent are payments toward health care related to ability to pay?

 Is the relationship proportional?

 Or is it progressive - do health care payments account for an increasing proportion of ability to pay (ATP) as the latter rises?

 Or, is there a regressive relationship, in the sense that payments comprise a decreasing share of ATP?

Which standards used to calculate and analyze for answering these questions, here are some suggestions

The Kakwani index (Kakwani, 1977) is the most widely used summary measure

of progressivity in both the tax and the health finance literatures (O’Donnell, Wagstaff, 1992; Wagstaff, 1999) It is twice the area between a payment concentration curve and the Lorenz curve and is calculated as π K = C pay – G pre , where C pay is the concentration index for health payments and G pre is the Gini coefficient of the ATP variable The value of π K ranges from –2 to 1

 A negative number indicates regressivity; L pay (p)lies inside L pre (p)

 A positive number indicates progressivity; L pay (p) lies outside L pre (p)

In the case of proportionality, the concentration lies on top of the Lorenz curve and the index is zero But note that the index could also be zero if the curves were to cross and positive and negative differences between them cancel

Given this, it is important to use the Kakwani index, or any summary measure of progressivity, as a supplement to, and not a replacement of, the more general graphical analysis

The rule of thumb was to consider only the concentration index for economic inequality of equal or greater than 0.2 (Moderate, severe or extreme inequality)for decomposition analysis

Wagstaff, Doorslaer, and Watanabe (2003) demonstrate that the health concentration index can be decomposed into the contributions of individual factors

to income-related health inequality, in which each contribution is the product of the sensitivity of heath with respect to that factor and the degree of income-related inequality in that factor For any linear additive regression model of health (y), such

regressors, where the weight for x k is the elasticity of y with respect to x k ( 𝜖𝜖 𝑘𝑘 =

𝛽𝛽 𝑘𝑘 𝑥𝑥̅ 𝜇𝜇 𝑘𝑘 ) The residual component - captured by the last term-reflects the

income-related inequality in health that is not explained by systematic variation in the regressors by income, which should approach zero for a well-specified model

The main aim of this methodis to unravel the causes of health sector inequalities, and their change over time Inequalities are caused by inequalities in the determinants of the variable of interest, and the decomposition in Equation (b) allows one to assess the relative importance of these different inequalities in generating inequalities in the variable of interest

O’Donnell, Doorslaer, Wagstaff, Lindelow (2005) wrote a handbook Analyzing Health Equity Using Household Survey Data, to provide researchers and analysts with a step-by-step practical guide to the measurement of a variety of aspects of health equity And, stimulate yet more analysis in the field of health equity, especially in developing countries Lead to a more comprehensive monitoring of trends in the health fair, a better understanding of the causes of these inequalities, more extensive evaluation of the impact of development programs on equity medical part, and the policies and programs more effective to reduce inequalities in the health sector In their book, they use many methods to evaluate the inequity, but

in summary they use two indices: Concentration index and Kakwani index to

evaluate

To understanding the definition of equity, Culyer and Wagstaff (1993) have

objectivesistoclarifythemeaningofthetwodefinitionsofequitywhichseemleastclear:

“distributionaccordingto need” and “equalityof access” Authors also concludethatthe

principlesof“distributionaccordingtoneed”and“equalityofaccess”have

been,andcontinuetobe,interpretedinanumberofdifferentways,andthatthevariousinterp retationsaremutuallyincompatible

To compare the inequity in a developed and a developing country, Wagstaff and Doorslaer (1994) useddatasets VHLSS 1998 (Viet Nam) and NPHS 1994 (Canada)in the paper which outlines a framework for comparing empirically overall health inequality and socioeconomic health inequality The framework, which is developed for both individual-level data and grouped data, is illustrated using data

on malnutrition amongst Vietnamese children and on health utility amongst Canadian adults In both cases, socioeconomic inequalities account for around 25%

of overall inequality

To examine which indices used in analyzing inequity, Kakwani, Wagstaff, Doorslaer (1997) used the dataset of Dutch HIS 1980/81 which clarifies the relationship between two widely used indices of health inequality and explains why these are superior to other indices used in the literature It also develops asymptotic estimators for their variances and clarifies the role that demographic standardization plays in the analysis of socioeconomic inequalities in health

To present evidenceonincome-relatedinequalitiesinself-assessedhealthin nine industrialized countries, Doorslaer,Wagstaff and partners (1997) used the datasets

of Sweden, Switzerland, UK,US, Germany among1980s-1990s in their study.Health interview survey data were used to construct concentration curves ofself- assessedhealth,measuredasalatentvariable.Inequalitiesinhealth

favoredthehigherincomegroupsandwerestatisticallysignificantinallcountries

Inequalities were particularly high intheUnited StatesandtheUnitedKingdom Amongst otherEuropeans,Sweden,FinlandandtheformerEastGermanyhadthelowest inequality.Across countries, a strongassociation was found betweeninequalitiesinhealth andinequalitiesinincome

To answer the question How is the inequity in Asia? O’Donnell, Doorslaer and partners (2005) studied the inequalities which described the structure and the distribution of health care financing in 13 territories that account for 55% of the Asian population Survey data on household payments are combined with Health Accounts data on aggregate expenditures by source to estimate distributions of total health financing In all territories, high-income households contribute more than low-income households to the financing of health care In general, the better off contribute more as a proportion of ability to pay in low and lower-middle income territories The distribution of out-of-pocket (OOP) payments also depends on the level of development In high-income economies with widespread insurance coverage, OOP payments absorb a larger fraction of the resources of low-income households In poor economies, it is the better off that spend relatively more OOP This contradicts much of the literature and suggests the poor simply cannot afford

to pay for health care in low-income economies Among the high-income territories, Hong Kong is the one example of progressive financing arising from reliance on taxation, as opposed to social insurance, and an ability to shield those on low- incomes from OOP payments Thailand has a similar financing structure and achieves a similar distributional outcome

To check which factors mostly affected to health care inequity, many writers used decomposition method for their researches, I remind that only factors that CIs are equal and greater than 0.2, mean that moderate and severe inquality, then for decomposition analysis Below are some studies on the world

Wagstaff, Doorslaer, Watanabe (2003) researched the decomposingthe causes of health sector inequalities with an application to malnutrition inequalities in Vietnam, they used VHLSS 1993 and 1998 for their study Inequalities across the income distribution in a variable y can be decomposed into their causes, and changes in inequality in y can be decomposed into the effects of changes in the means and inequalities in the determinants of y, and changes in the effects of the determinants

of y Inequalities in height-for-age in Vietnam in 1993 and 1998 are largely accounted for by inequalities in consumption and in unobserved commune-level influences Rising inequalities are largely accounted for by increases in average consumption and its protective effect, and rising inequality and general improvements at the commune level

To compare inequality decomposition from Vietnam and other countries, Wagstaff (2005) also researchedinequality decomposition and geographic targeting with applications to China and Vietnam.Inthis research they used dataset VHLSS

1998 The study answer the question How far are income-related inequalities in the health sector due to gaps between poor and less poor areas, rather than due to differences between poor and less poor people within areas? This note sets out a method for answering this question, and illustrates it with two empirical examples The disproportionate accrual of health subsidies to Vietnam’s better-off is found to

be largely due to the fact that richer provinces have larger per capita subsidies, while pro-rich inequalities in health insurance coverage in rural China are found to

be largely due to the fact that better-off villages have been more successful at preventing the collapse of their insurance schemes

As a similar research in Asian country, Chai Ping Yu, Whynes,Sach (2008) have studied health care finance in Malaysia, they used datasets HE 92/93 for this study.The primary purpose of this paper was to comprehensively assess the equity

of healthcare financing in Malaysia, which represents a new country context for the quantitative techniques used The paper evaluated each of the five financing sources (direct taxes, indirect taxes, contributions to Employee Provident Fund and Social Security Organization, private insurance and out-of-pocket payments) independently, and subsequently by combined the financing sources to evaluate the whole financing system Results showed that Malaysia's predominantly tax- financed system was slightly progressive with a Kakwani's progressivity index of 0.186

Tran Van Tien, Hoang Thi Phuong, Inke Mathauer and Nguyen Thi Kim Phuong (2011) drawed the general picture about equity in Viet Nam through the report “A Health Financing Review of Viet Nam with a focus on social health insurance” This report describes the findings of an assessment of the current health financing system in Viet Nam The report provides a detailed analysis of Viet Nam’s health financing system by assessing the system’s institutional design and organizational practice in relation to the key health financing functions of resource collection, pooling and purchasing and how these affect the performance of the system On this basis, it is possible to identify appropriate changes ininstitutional design and organizational practice that contribute to progress towards universal coverage

With all international and domestic studies, researching for equity especially

in health care finance becomes imperatively, Viet Nam is a developing country, these issues even more important The author want to answer the question whether inequity in health care finance in Vietnam deserve to the issue to examine And how the level of that problem compared to other countries? To supply to policy-maker a look to resolve the present problem

CHAPTER 3: METHODOLOGY

3.1 Analytical framework

The assessment on equity in financing health care draws on established techniques from the public finance literature The starting point of assessment is the notion that health care financed according to ATP is considered equitable To judge whether the health payment undermines or contributes to the equitable financing goal, one has to assess how closely in practice health payment is linked to ATP Progressivity measures the deviation from proportionality in the relationship between health payment and ATP It reveals the extent of inequality in paying for health care services between households of unequal ATP A health payment is progressive (regressive) if it accounts for an increasing (decreasing) proportion of ATP as ATP rises A progressive (regressive) system means thathouseholds with greater ATP are paying more (less) proportionally in financing health care Health care financing systems are proportional if households with different ATP are spending the same proportion of ATP in financing health care

Figure 6: Framework of analysing inequity

Inequality (CI index)& Inequity (Katwani index) in Vietnam?

Health expenses Food Expense Non-food Expense

ATP OOP

IP OP

3.2 Model

In this research, five basis approaches were used, namely 1) OLS and Quantile Regression;

2) Calculate Concentration indices;

3) Calculate Katwani indices (Ks);

4) Draw Concentration curves;

5) Decomposition of the concentration index for economic inequality

Table 3: Summary formulae analyzing inequity SUMMARY:

Step 1: OLS and Quantile regression to check which factors affect to y (ATP)

Examine the regression function y i =x i β+ ε i with

y i : dependent variable,

x i : vector of independent variales;

ε i : error terms; the estimation of regression function 𝑦𝑦 � = 𝑥𝑥 𝑖𝑖 𝑖𝑖 𝛽𝛽̂ + 𝑒𝑒 𝑖𝑖

Follow OLS method (OLS – Ordinary Least Square), sample regression function was estimated so that total of square of error is minimize, means that

The conditional quantile regression of dependent variable Y follow X at quantile

𝜏𝜏 ∈ (0,1) is the function 𝑄𝑄 𝜏𝜏 (𝑦𝑦 𝜏𝜏 ) = 𝑥𝑥 𝑖𝑖 𝛽𝛽 �, within 𝛽𝛽 𝜏𝜏 � is choiced as if total of error 𝜏𝜏 different of quantile τ is minimize Means that:

𝛽𝛽̂ = arg min�𝜏𝜏 ∑ 𝑦𝑦 𝑖𝑖 ≥𝑋𝑋 𝑖𝑖 𝜷𝜷 𝜏𝜏 (𝑦𝑦 𝑖𝑖 − 𝑥𝑥 𝑖𝑖 𝛽𝛽 𝜏𝜏 ) + (𝜏𝜏 − 1) ∑ 𝑦𝑦 𝑖𝑖 <𝑋𝑋 𝑖𝑖 𝜷𝜷 𝜏𝜏 (𝑦𝑦 𝑖𝑖 − 𝑥𝑥 𝑖𝑖 𝛽𝛽 𝜏𝜏 ) �, With 𝑦𝑦 𝑖𝑖 = 𝛼𝛼 + ∑ 𝛽𝛽 𝑗𝑗 𝑗𝑗 𝑖𝑖𝑛𝑛𝑖𝑖𝑖𝑖𝑚𝑚𝑒𝑒 𝑗𝑗 + ∑ 𝛾𝛾 𝑗𝑗 𝑗𝑗 𝑥𝑥 𝑗𝑗𝑖𝑖 + 𝜀𝜀

Step 2: Calculate Concentration indices (CIs)

Convenient formula for the concentration index defines it in terms of the covariance between the health variable and the fractional rank in the living standards distribution (Jenkins, 1988; Kakwani, 1980; Lerman and Yitzhaki, 1989),

𝐶𝐶 = 2 𝜇𝜇 𝑖𝑖𝑖𝑖𝑐𝑐(𝑦𝑦, 𝑟𝑟) (2)

The concentration index (C) can be computed very easily from microdata by using the “convenient covariance” formula If the sample is not self-weighted, weights should be applied in computation of the covariance, the mean of the health variable, and the fractional rank Given the relationship between covariance and ordinary least squares (OLS) regression, an equivalent estimate of the concentration index can be obtained from a “convenient regression” of a transformation of the health variable of interest on the fractional rank in the living standards distribution (Kakwani, Wagstaff, and Doorslaer, 1997) Specifically

2𝜎𝜎 𝑟𝑟 2 � 𝑦𝑦 𝑖𝑖

𝜇𝜇 � = 𝛼𝛼 + 𝛽𝛽𝑟𝑟 𝑖𝑖 + 𝜀𝜀 𝑖𝑖 , (3)

And where 𝜎𝜎 𝑟𝑟 2 is the variance of the fractional rank The OLS estimate of β is an estimate of the concentration index equivalent to that obtained from equation (3) The weighted fractional rank is defined as follows:

𝑟𝑟 𝑖𝑖 = 𝑛𝑛 1 ∑ 𝑖𝑖−1 𝑗𝑗 =1 𝑤𝑤 𝑗𝑗 + 1 2 𝑤𝑤 𝑖𝑖 (4)

where w i is the sample weight scaled to sum to 1, observations are sorting in ascending order of living standards, and w 0 = 0 The variance 𝜎𝜎 𝑟𝑟 2 and 𝜇𝜇of the fractional rank, depends only on the sample size and so has no sampling variability,

so we change (3) to model :

𝑦𝑦 𝑖𝑖 = 𝛼𝛼 + 𝛽𝛽𝑟𝑟 𝑖𝑖 + 𝜀𝜀 𝑖𝑖 (5) the estimate of the concentration index is given by 𝛽𝛽̂ = � 2𝜎𝜎 𝑟𝑟 2

𝜇𝜇 � 𝛽𝛽̂ 1 (6)or 𝛽𝛽̂ = � 2𝜎𝜎 𝑟𝑟 2

𝛼𝛼� 1 + 𝛽𝛽�1 2 � 𝛽𝛽̂ 1 (7)

Step 3: Calculate Kakwani indices (Ks)

Because a Kakwani index is the difference between a concentration index and a Gini index, both of which can be computed by the convenient regression method , its value can be computed directly from one convenient regression of the following form:

The OLS estimate of β =(𝛽𝛽 2 − 𝛽𝛽 1 ) is an estimate of the Kakwani index

Step 4:Draw concentration curves

Lorenz dominance analysis is the most general way of detecting departures from proportionality and identifying their factors in the distribution of ability to pay This study applies the Kakwani index (Kakwani, 1977), which is the most widely used summary measure of progressivity in both thetax and the health finance literatures (Wagstaff and others, 1992, 1999; O’Donnell and others, 2005)

The Kakwani index is defined as twice the area between a payment concentration curve and the Lorenz curve and is calculated as π K = C pay – G pre , where C pay is the concentration index for health payments and G pre is the Gini coefficient of the ATP variable The value of π K ranges from –2 to 1 A negative number indicates regressivity; L pay (p) lies inside L pre (p) A positive number indicates progressivity;

L pay (p) lies outsideL pre (p) See more detailed in 2.4

Step 5: Decomposition inequality of ATP

Wagstaff, Doorslaer, and Watanabe (2003) demonstrate that the health concentration index can be decomposed into the contributions of individual factors

to income-related health inequality, in which each contribution is the product of the sensitivity of heath with respect to that factor and the degree of income-related inequality in that factor For any linear additive regression model of health (y), such

as

The surveys collected information by means of household and community level questionnaires Information on households include basic demography, education, health, income, expenditure, etc Especially, expenditure and income per capita are collected using detailed questions The surveys also contain information on health insurance of household members, the number of annual outpatient and inpatient visits to hospitals and clinics, and out-of-pocket expenses for outpatient and inpatient services However, detailed information on out-of-pocket payments for health care is not available It means that the out-of-pocket pocket payments defined

in VHLSSs include not only the treatment fees, but also all costs related to treatments, such as bonus for doctors, service charge for additional medicine requirement, equipment, transport, etc (Nguyen, 2012)

3.4 Variables

In many researches used the household survey data of World Bank , that writers often use the socio-ecomomic varibles to analyze the problems So, I also use these variables in my thesis when analyzing inequity in health care finance in Vietnam These Explanatoryvariables ofhouseholdswhich comparisons were made include: