Icon Our Work

The ADB’s Social Protection Index: what does it really tell us about social security provision in Asia?


Development Pathways

The recent release of the Asian Development Bank’s Social Protection Index (SPI) has generated much interest; it even merited a story in the Economist. However, little attention has been paid to the SPI’s definition of social protection. Yet, it is this definition – and the schemes included and excluded in the SPI – that determines the scores allotted to countries.

The SPI is, evidently, not a Social Security Index. In other words, it is not restricted to the system of regular and predictable transfers that directly address income insecurity (which is how most people intuitively understand social protection). Instead it employs a somewhat broad and idiosyncratic definition of social protection, incorporating other public services such as health and disaster relief (along with others). By including these other services, it provides scores for countries that potentially distort their real levels of commitment to social protection.

With regard to health spending, key questions are: a) why is health spending included; and, b) why is this done on an inconsistent basis?

The SPI seems to incorporate health into its scoring system because it falls into the common trap of being fooled by the term “insurance.” As with many people working on social protection, the ADB appears to believe that, because the word “insurance” is used, then health insurance must be “social protection.” Yet, health insurance is only one means of financing the health sector. So, if health insurance is included in social protection, we also need to incorporate tax-financed health systems, since tax is merely an alternative means of financing the health sector.

Now, the SPI does include some elements of tax-financed health provision but only when “targeted” to specific groups such as “the poor.” It appears, therefore, to fall into another classic trap, which is related to its understanding of “targeting.” It fails to recognise that “free health for all” is only differentiated from schemes such as “fee waivers for the poor” by the targeting methodology used. Free health uses universal targeting while fee waivers are poverty targeted. So, if poverty targeted health provision is included then, logically, it is necessary to include health spending using universal targeting. Indeed, in countries that provide free universal health care, the entire health budget should – according to the ADB’s logic – contribute to the SPI score.

To illustrate these points, we can examine the British and German health systems. Under the SPI, the UK would receive almost a zero score for health – despite an impressive tax-financed health system that provides universal coverage – while Germany would score very highly because it uses social insurance to finance much of its health system. Similarly, the UK would score much worse on “health” social protection than countries providing fee waivers for a small proportion of the population but which charge fees to the majority (including large numbers of poor families). Neither of these cases makes sense.

Another issue is the inclusion of disaster relief in the SPI. While there are some similarities between disaster relief and social security – in that they both provide transfers – disaster relief rarely provides regular and predictable transfers (which is a defining characteristic of social security). Although most social protection experts would recognise disaster relief as closely linked to social protection, they would not go as far as recognising it as social protection.

Health spending and disaster relief both contribute significantly to the SPI score in some countries, thereby distorting inter-country comparisons. It would be much more helpful if the SPI were restricted to the social security sector. Health spending and access could then be measured in a specific “Health Index”, education spending and access in an “Education Index,” and so on. Multiple indices would give a more comprehensive understanding of social policy within developing countries. But, as it stands, the SPI is just a bit too vague and inconsistent and, as a result, offers a somewhat distorted picture of Asian countries’ commitments to social security. 

Author’s Note

In this blog I deliberately do not get into the issue that many so-called health insurance schemes are, in fact, financed by taxes, such as Indonesia’s Jamkesmas scheme. This would be just too challenging!

I also won’t discuss the fact that the SPI does not include social care services, except marginally under “child welfare.” Yet, many countries – as well as UNICEF – would count social care services as a core element of broader social protection.