More on the proposed PAW ('Punching above [GDP] Weight') index: On the relationship between HDI, IHDI, and PAW

In a commentary response to my earlier posting on the possible utility of an index comparing GDP to HDI ranking, which I called the PAW ('Punches above its Weight') index, Mariana Prado suggested that the PAW index might be improved if one were to use inequality-adjusted HDI (IHDI) rather than raw HDI:

 The HDI includes a measure of wealth (until recently they used GDP per capita, and now they use GNI per capita). As a consequence, the rank of a particular country in the HDI may reflect a high level of income, and relatively low levels of education and health. Therefore, my suggestion is to calculate the PAW as the difference between a country's rank in GDP (or GNI) per capita) minus its rank the HDI health and education indexes. If this is too complicated, perhaps an easy improvement would be to use inequality-adjusted HDI (IHDI) instead. The advantage of IHDI is that it reflects more accurately the percentage of the population that is actually benefiting from the existing levels of income, health and education in the country.

Actually, exploring the relationship between IHDI, HDI, and PAW revealed something very interesting:  namely, that the relationship between a country’s IHDI and HDI rankings seems to correlate significantly with that country's PAW index.  In other words, the greater the positive difference between a country’s IHDI and HDI rankings, the more likely that country will also have a high PAW (i.e., high positive difference between its HDI and GDP rankings).  Below are the top 15 countries listed in order of difference between IHDI rankings and HDI rankings (IHDI/HDI), followed by their PAW.

Country	IHDI/HDI	PAW
Moldova	18	22
Uzbekistan	17	19
Kyrgyzstan	17	21
Mongolia	15	18
Ukraine	14	34
Vietnam	14	13
Armenia	13	31
Azerbaijan	11	11
Belarus	10	11
Czech Republic	9	11
Serbia	9	9
Sri Lanka	9	39
Gabon	8	-60
Indonesia	8	17
Tajikistan	8	17
Jordan	5	43
Georgia	2	30

This leads me to suspect that what IHDI is adding to HDI may be similar to what PAW is adding to GDP.  This makes sense, conceptually.  The lower HDI conditions of the poorer and more impoverished part of a country's population are the low-hanging fruit of HDI development:  the relatively low level of HDI resources available to this population means that it is likely to be the easiest and most cost effective part of the HDI spectrum to improve.  All in all, it suggests that the way that a country punches above its GDP weight is by effecting a progressive (re)distribution of HDI resources to poorer populations. 

There are a number of significant implications stemming from this.  First, the PAW index probably should not use iHDI rather than HDI, because that would lead to double counting -- if PAW is ultimately measuring inequality in HDI resource distribution, and IHDI is also measuring inequality in resource distribution, than using IHDI rather than HDI in calculating PAW would be counting resource inequality twice.

Second, it would seem to confirm my hypothesis that insofar as development is concerned, strategic institutional design -- aka 'law and development' -- should properly focus on issues of resource (re)distribution and not on growth, as that is the part of the HDI basket that is most likely to be susceptible to institutional architecture.

This leads to a third implication, which has significant and very ambiguous implication for the law and development project.  Because it is unclear to my mind how much legal-institution reform can actually effectuate progressive redistribution of HDI resources.  I will address this in a separate post.