Thanks to Thomas Piketty, we’ve heard a lot this year about rising inequality. And with just over a year to go before the MDG ‘window’ closes, we’ve also heard a lot about the ‘post-2015 agenda’. In a paper with Leander Buisman that just came out in the World Bank Research Observer, we bring these two themes together and ask: “Were the poor left behind by the health MDGs?” Influenced perhaps by all the talk of rising income inequality, there are certainly plenty of pessimistic folks out there who think that health inequalities, too, are on the rise; that the better off are likely to have seen much faster improvements in MDG indicators than the poor.
Are the pessimists right?
To see whether the pessimists were right, we assembled the raw data from 235 household surveys – mostly Demographic and Health Surveys (DHSs), but some Multiple Indicator Cluster Surveys (MICSs) as well. We looked at progress on 12 MDG indicators: five health status indicators (stunting and underweight among young children, infant and under-five mortality, and HIV prevalence), and seven health intervention indicators (full immunization, measles immunization, 4+ antenatal visits, skilled birth attendance, contraceptive prevalence, condom use during risky sex, and malaria net use among children). To see whether progress has been faster among the poor or better off, we used a wealth proxy – a weighted average of indicators capturing ownership of household durables, type of roof, type of floor, etc. – and then ranked households from poorest to richest.
We found that on the MDG intervention indicators the pessimists have been too gloomy. Whether we compared rates of change for the poorest 40% and richest 60%, or looked at changes in a measure of relative inequality, we found that progress has, in the average country, been pro-poor. The pessimists have a stronger case when it comes to health outcomes. We found that, in the average country, progress hasn’t, in fact, been pro-poor. But only in the case of HIV/AIDS has it been pro-rich. In the case of the two child malnutrition measures, as well as infant and child mortality, it has been neither pro-rich nor pro-rich – there have been similar proportionate improvements among the poorest 40% and the richest 60%, and a zero change in the relative inequality measure.
Peering behind the mean
Averages always mask variations. And, in our case, quite big ones. Even on the intervention indicators which, for the most part, have changed in a distinctly pro-poor way, we found a sizable fraction of countries where progress has been pro-rich: 20% in the case of skilled birth attendance and antenatal care, and 25-30% in the case of immunization. In almost half of countries, relative inequality in child malnutrition and child mortality increased, often quite markedly.
We also found that countries that have achieved pro-poor progress on one indicator have often achieved pro-poor progress on others too: two thirds of the variation in ‘excess growth’ among the poorest 40% across indicator-country combinations is explained by some countries systematically doing better than others in achieving pro-poor progress across all indicators. There’s also a geographic pattern to the pro-poorness of progress on the health MDGs: in almost all countries in Asia, progress on underweight has been pro-rich, and in much of Africa inequalities in under-five mortality have been growing.
We also found that although for the most part relative inequality has not been rising, current inequalities are appreciable. In the latest round of surveys used in this study, we found that the richest 60% of children in the developing world are 1.2 times as likely to sleep under a bednet, 1.5 times as likely to be immunized, and 3.7 times as likely to have been delivered by a skilled birth attendant as the poorest 40% of children.
Pointers and a puzzle
Our results suggest at least two things. First, the authors of the UN’s 2013 high-level panel report on the post-2015 development agenda are right to suggest that the post-2015 agenda should be about monitoring progress not only of populations as a whole but also of subpopulations like the poorest 40%. Second, while these results are by no means an indictment of policies to date, they do suggest that more needs to be done to reach the poor. Our results raise some obvious questions. Do universalist policies do enough for the poor, or should there be greater reliance on targeting? Do ‘new’ initiatives such as conditional cash transfers and pay-for-performance (P4P) stand a better chance of reducing inequalities than ‘traditional’ approaches like pure input-financing?
In our results there’s also a puzzle. Why has progress on interventions been so much more pro-poor than progress on health outcomes? Inferior quality of health care among the poor is one possibility. Yes, the poor have seen faster progress on the quantity of care received, but if the quality of care they receive is systematically worse, then this may not be enough to close the gap in health outcomes. If true, this would imply a need to monitor not just the quantity of care received, but its quality too – for example, not just whether or not 4 or more antenatal visits occurred, but what happened during that visit. Another possible explanation is that the MDG intervention indicators cover mostly interventions delivered by health professionals. It’s possible that interventions delivered by caregivers – breastfeeding, supplementary feeding, handwashing, oral rehydration remedies for a child suffering from diarrhea – have progressed in a less pro-poor way than the monitored interventions. If true, this would suggest further broadening of the monitoring framework, and some rethinking of how to raise the levels of caregiver-provided interventions among the poor.