The Millennium Development Goals (MDGs) embody the key dimensions of human development – poverty, hunger, education, health – expressed as a set of time-bound targets. They include halving income-poverty and hunger; achieving universal primary education and gender equality; reducing under-five mortality by two-thirds and maternal mortality by three-quarters; reversing the spread of HIV/AIDS; and halving the proportion of people without access to safe water. These targets are to be achieved by 2015, the comparison point being 1990.
It is often said that global targets are easily set but seldom met. The real question for the goals is whether they are feasible and how we measure whether we are achieving them. Progress to date in areas such as education, health, nutrition and income in more than 130 developing countries is difficult to summarise.
The 1990s saw many success stories, including a sharp increase in primary education enrollment in Guinea and Malawi; a halt to the rise in incidences of HIV/AIDS in Senegal, Thailand and Uganda; a sharp fall in child mortality in Bangladesh and the Gambia; improved nutrition in Indonesia, Mexico and Tunisia; and a reduction in income-poverty in China.
But for each success story, there is another tale of setbacks. The under-five mortality rate increased in Cambodia, Kenya, Malawi and Zambia – an unprecedented trend after decades of steady decline. The primary school enrolment ratio dropped in Cameroon, Lesotho, Mozambique and Tanzania. The gender gap in primary education widened in Eritrea, Ethiopia and Namibia. Instead of decreasing, malnutrition increased in Burkina Faso and Yemen. Access to water became more difficult for millions of people; Bangladesh faced a major problem with arsenic water poisoning. And countless countries saw their HIV prevalence rate double, triple, quadruple, even increase ten-fold – severely undermining the feasibility of most development goals, not just in health.
Overall, none of the agreed targets for the year 2000 was met at the global level. If the 1980s were called the “lost decade for development”, the 1990s should go down in history as the “decade of broken promises”. And if current trends prevail, only one millennium development goal will be reached at the global level by 2015: that of safe water.
But even if the goals appear feasible at the global level, it does not necessarily imply that they will be feasible everywhere, in all locations. Averages are commonly used at each level to measure progress, but while these give a good overall picture, they can be misleading. Average household income, for example, may exist as a useful concept in the mind of an economist, but bears little relation to the reality faced by the millions of poor women who have little or no control over how it is spent. Nor do average national indicators take into account which sections of the population are actually doing better than before.
Unfortunately, the poor have benefited proportionately little from ‘average’ progress to date, as evidenced by widening disparities in terms of income, education and mortality in many developing countries.
There are different ways of reaching a global or national development target. At one extreme, it can be achieved by improving the situation of the already better-off segments of society – a top-down approach. At the other extreme, a target can be achieved by improving the situation of the worse-off population, which is the bottom-up approach, and of course many combinations are possible in between. But the evidence suggests that most countries come closer to following the top-down rather than the bottom-up approach. And frequently when average national measures show progress it is the disadvantaged groups that are most often by-passed by the average progress.
Take life expectancy or education. Data from more than 40 demographic and health surveys show that a child from a poor family is invariably more likely to die before age five than his or her counterpart from a rich family. Similarly, children from poor families are less likely to complete primary education than children from rich ones. Data for 12 countries in Latin America show that over 90% of children in the highest 10% income earners of the population complete primary education. The share falls to two-thirds for children in the middle-income range and drops below 40% for the poorest children.
Demographic and health surveys for 1994 and 1997 in Bangladesh also show that improvements in access to basic education chiefly benefited children from better-off families, while children from poor families saw little or no improvement. And the poor also suffer first when things go wrong. In Peru, where access to primary education worsened in the 1990s, only the poor bore the consequences; the non-poor were not affected. Improvements seem slow at best to trickle down, but setbacks often tumble down like avalanches.
The trend in Zimbabwe offers a particularly stark example of the dangers of “average” measures of progress. Between 1988 and 1999, the national average under-five mortality rate decreased by a modest four percentage points, but under-five mortality in the poorest section of the population actually increased. By 1999, children in the poorest quintile had an under-five mortality rate four times higher than that for their counterparts in the richest quintile.
In sum, averages are deceiving. We already know that, so why do we keep applying it? Some countries appear to be on track for reaching a particular target, based on average progress; yet the situation for disadvantaged groups in those countries is stagnant or deteriorating. So not only was global progress towards the millennium development goals inadequate in the 1990s, much of it bypassed the poor. For them, being “average” still means dying too young or surviving to remain illiterate and excluded. Development goals should be more targeted than that. So should we. When it comes to poverty reduction, we really have to perform much better than average.
©OECD Observer No. 233, August 2002