Socioeconomic inequalities in informal payments for health care: An assessment of the ‘Robin Hood’ hypothesis in 33 African countries Hyacinthe T. Kankeu and Bruno Ventelou
Aix-Marseille School of Economics
Rationale (1)
Out-of-pocket expenses (OOPE) represent important shares of the total health expenditure in many African countries Country
OOPE as a % of total health expenditure, 2012
Sierra Leone
76
Sudan
74
Guinea
67
Nigeria
66
Cameroon
63
Mali
61
Egypt, Morocco
60
South Sudan, Côte d’Ivoire
56-57
Niger, Chad, Sao Tome and Principe
51-53
Source: WHO, 2014
IPs
Source: OECD, Eurostat, WHO, 2011
Rationale (2)
These OOPE often take the form of informal payments (e.g. bribes) initiated by the patients or the health staff
to access health care
to avoid queues
to receive care of better quality
to express gratitude
Rationale (3)
Existing literature on informal payments for health care mainly covers countries of Central and Eastern Europe (Cherecheş et al., 2013; Stepurko et al., 2010)
Equity: very little empirical evidence
Only one study by Szende and Culyer (2006) in Hungary: the system of informal payments is regressive
Some authors have found no significant effect of the income on the probability and amount of informal payments (e.g. Kankeu et al., 2014; Tomini and Maarse, 2011; Aarva et al., 2009)
Others have identified unclear associations (Özgen et al., 2010; Balabanova and McKee, 2002)
Aims
Ensor and Savelyeva (1998): informal payments may lead to a quasi redistribution, with physicians playing a ‘Robin Hood’ role, subsidizing the poor at the expense of the rich.
Our aim: test for this hypothesis in the context of African countries study
the socioeconomic gradient in demands for informal payments and in the actual payment of bribes in public health facilities
identify
the main factors associated with the observed inequalities
Data (1)
Afrobarometer surveys (rounds 3 and 5) :National representative samples of individuals 18+
Afrobarometer: an independent research project that produces a series of national public attitudes surveys on democracy and governance in Africa
Round 3 18 countries March 2005 - February 2006 N from 1048 to 2400 A total of 25,397 individuals
Round 5 33 countries October 2011 - June 2013 N from 1190 to 2407 A total of 51,605 individuals
This work includes only those who had contact with public health facilities in last 12 months
Data (2)
Outcomes variables:
Binary variable which indicates whether the individual has faced at least one demand for informal payments in the local public health facility during the last 12 months
Binary variable indicating whether the individual has paid a bribe at least once during the last 12 months
Socioeconomic variable
Lived Poverty Index: an aggregated measure of how frequently people actually go without basic necessities during the course of a year (Mattes et al., 2003) “Over the past year, how often, if ever, have you or your family gone without enough - food, water, cooking fuel, cash income –”
From 0 (constant) to 4 (never)
Methods –
Measuring the inequalities
Concentration Index (Kakwani et al., 1997): Value in [-1 ; 1]
Negative values indicate disproportionate concentration of the outcome variable (Y) among those with lowest value of the ranking variable (R)
The poor/rich gap in bribe payment rates:
The Odds Ratio:
Methods – Identifying contributors to inequalities
Decomposition of the concentration index (Wagstaff et al., 2003)
Blinder-Oaxaca type decomposition of the poor/rich gap in bribe payment rates (Fairlie, 2005, 1999)
Demand side factors: gender, age, education, employment status, place of residence, religion
Supply side factors: presence of a health facility in the enumeration area, lack of medicines, absent doctors and long waiting time at the local public health facility
Regional dummies (within the country)
Results– Incidence of informal payments
Results– Inequalities measured by the concentration index
Results– Decomposition of the concentration index
Results– Poor/Rich differences in bribe payment rates
Results–
Decomposition of the poor/Rich gap in bribe payment rates
Summary (1)
Demands for informal payments and the actual payment of bribes are disproportionally concentrated on the poor
The system of informal payments is highly regressive (not pro-poor)
No evidence – at the country level - of a redistribution from the rich to the poor (“Robin Hood”)
In line with the findings of Szende and Culyer (2006) in Hungary
Inequalities in the payment of bribes: highly associated to poor/rich differences in supply side factors like lack of medicines or other supplies, absence of doctors and long waiting time The
poorest face more frequently these problems which are well known to increase the risk of incurring informal payments
Summary (2)
The socioeconomic disadvantage itself is a main contributor to inequalities in bribe payment. poor/rich
differences in bribe payment also reflect directly the socioeconomic gradient.
The
richest may be more self-confident and more aware of what they are entitled to.
When
facing demands for informal payments in their local public health facility, the richest may have more alternatives.
The contribution of regional dummies, highlights the fact that the local context significantly affects poor/rich differences in bribe payment.
Limitations
No information on the amount paid and the number of visits.
No information to distinguish between primary and hospital care.
Self-reported data, subject to recall bias (but no reason to think that the interviewees systematically distorted their responses).
Conclusions
Results highlight the need for African health systems to better protect the worse-off from financial risk when seeking care.
Insurance schemes and other pre-paid mechanisms (e.g. vouchers) targeting the most vulnerable groups (compulsory health insurance inexistent in 12 of the 33 countries included).
Improve the working conditions in (public) health facilities as well as the quality of care provided to patients: ensure an appropriate funding to avoid shortages of medicines and supplies, to maintain and motivate staff. Performance-based financing can be a good approach.
Socio-cultural specificities of each region in each country should be taken into account.
Thank you!
© Tugela Riddley/IRIN
Concentration curve
back
Results–
Contributors to the poor/Rich gap in bribe payment rates
‘+’ (‘-’) : group differences in the variable are associated with an increase (decrease) of the inequalities disfavoring the poorest
*, ** and *** : the contribution is significant at 10%, 5% or 1% respectively
“Payments schemes for health care can be regarded as fair if payments are made solely according to the individual’s ability-to-pay and not according to their need for health care”. Van Doorslaer E, Wagstaff A, Rutten F, editors. Equity in the finance and delivery of health care: an international perspective. Oxford University Press; 1993.