Syndicate content

What’s an ambitious but realistic target for human capital progress?

Zelalem Yilma Debebe's picture

Globally, 56 percent of children live in countries with Human Capital Index (HCI) scores below 0.5. As these countries gear up to improve their human capital outcomes, it is vital to set a target that is ambitious enough to prompt action and realistic enough to be achieved. One way to get at this is to examine the historical rate of progress that countries demonstrated to be possible.

Using time-series data between 2000 and 2017, we estimated countries' progress in the health components of HCI (fraction of children not stunted, child survival and adult survival) using a non-linear regression model. [1] Our measure of progress is the fraction of gap to the frontier that is eliminated every year- the frontier being 100 percent child and adult survival, and no stunting.,[2]

We address the following two questions:

  1. What is the typical progress in the health components of HCI observed globally?

On average, 84 countries for which we have sufficient data eliminated 2.7 percent of the gap to having zero stunting every year. [3] This is equivalent to saying these countries reduced 2.7 percent of their stunting rate every year. What this means in terms of absolute reduction in stunting rate depends on where the country is at a given point. For example, a country that has a stunting rate of 19 percent (the average in these 84 countries), could reduce its stunting rate by 0.5 percentage points in one year. For adult survival and child survival, the typical percentage of gap to the frontier eliminated every year—observed across 155/156 countries—is 2.0 and 4.0 percent, respectively. This means that, on average, these countries reduced 2 percent of their premature adult mortality rate and 4 percent of their child mortality rate every year.

Do these averages mask differences across income groups? For stunting reduction, middle income countries have progressed faster than both high income and low-income countries. Whereas in the case of child survival and adult survival, low-income countries appear to have progressed faster, on average. As our metric of progress is percentage of gap rather than the absolute change, this is not an artifact of the fact that low income countries started from a lower base. In fact, a scatter plot of our measure of progress against both initial levels or income shows no particular pattern (Figure 2 & 3). The absolute changes corresponding to our relative measures of progress would, however, be higher for countries starting from high stunting and mortality rates.

Figure 2. In the non-linear model we adopted, rate of progress does not depend on where countries started

Note: Rate of progress for countries with at least 3 data points.

Figure 3. The percentage of gap to the frontier eliminated does not systematically vary with income

Note: Rate of progress for countries with at least 3 data points.
  1. What would be an ambitious but realistic rate of progress that other countries demonstrated to be possible?

The best performing country in the dataset reduced premature adult mortality by 6.3 percent every year, which is more than three-fold the typical rate of progress across countries (Figure 4). All the countries in the top 5 percent of progress in adult survival rate are from the African continent (Tanzania, Zimbabwe, Zambia, Morocco, Malawi, Rwanda, Kenya and Botswana).

In terms of child survival, the best performer eliminated, on average, 9.8 percent of the gap to the frontier every year (i.e a 9.8 percent reduction in child mortality every year). Except Rwanda, all the other top 5 percent improvers are either from East Asia and the Pacific or Europe and Central Asia Regions (Mongolia, China, Hong Kong SAR China, Macao SAR China, Estonia, Montenegro, Kazakhstan and Rwanda).

While typically countries reduced their stunting rate by 2.7 percent every year, the best performer reduced it by 10 percent every year. As opposed to the survival measures, the best 5 percent improvers in stunting reduction come from a wide set of Regions (Algeria, China, Mongolia, Kazakhstan and Paraguay). It is worth noting that the best improvements are not driven by one outlier country as suggested by the narrow gap between the best and the average of the top 5% countries.

These impressive rates of progress can be used as targets towards which countries could strive. Being a low-income country does not inhibit countries from making similar progress. This is evident in that there is at least one low income or lower-middle income country in the top 5 percent improvers across the three dimensions.

[1] A non-linear regression model is used to summarize trends as all the three indicators are bounded from above (for example, child survival rates cannot exceed 100%). Recognizing these bounds is important when making predictions about the future based on this historical progress.  In contrast, simple linear extrapolations may lead to scenarios where projected values for the health indicators exceed their natural bounds.
[2] To maintain consistency with the HCI, we assume zero stunting rate as the frontier but a stunting rate of 2.5% is expected in a given population.
[3] "sufficient data" here means at least three data points between 2000 and 2017.


Submitted by Stanislas on

Thank you Zelalem for this great analysis of HCI health components trends through the world. we nourish big expectation to get another one on health components. That would be so useful to set target at country levels in comparison with national policies targets. Best regards.

Add new comment