[First published on May8, 2006] In Part II, I used a people’s life expectancy from birth (LE) as an indicator of whether they were ruled by a mortacracy or not. But is LE enough?
Perhaps, in addition to LE, I should consider a wider measure of human development that takes into account LE’s social and economic context, and its high and low. We have this from the UN’s Human Development Report for 2005 It used a human development index (HDI) based on a people’s income, education and health. Its purpose is not to give a complete picture of human development, but to provide a measure of human well-being (see here for the indices involved, and their calculations). This is precisely what is impacted by mortacracies.
The report also provides a Life Expectancy Index (LEI), which among other indices goes into calculating the HDI. It is:
(a country’s life expectancy minus the world low) / (world high minus world low).
Thus, the lowest LEI would be 0, and the high would be 1.0. As to calculating the HDI, each of the indices that go into it is determined as is LEI above, and HDI is an average of them all. Thus, HDI also varies from a low of 0 to a high of 1.0. The 2003 HDI plotted against LEI is shown in the chart below (If it is unclear or does not show, click here)
Since LEI is a linear transformation of LE, the same curve would obtain even if LE were used in place of LEI.
The best fitting curvilinear function for the plot is the natural logarithmic one shown, with a correlation R^2=.82. It accounts for 82 percent of the variance between HDI and LEI. As the well being of a people increases as measure by HDI, there is an increasingly close relationship between this well being and their life expectancy. This is clear from the chart, where along the fitted log curve, the distribution of countries (blue dots) around the curve tightens into a cone at the highest level. I would argue that something is causing the wide distribution of countries at the low end of HDI and LEI, most likely the mortacracies.
Now, through inadequate health services, forced impoverishment, and extensive violence, thug regimes repress their subjects’ well being such that they die at an early age. That is, HDI and LEI should both be low. This can be determined by averaging them together (since one-third of HDI is calculated from the LEI, averaging LEI and HDI means that 50 percent of the average is owed to LEI). When this is done, the ten countries with the lowest averages are shown below (If the table is unclear or does not show, click here):
<img src=” http://www.hawaii.edu/powerkills/AVG.HDI_LEI.2003.GIF”
Clearly, a study of such countries would show corrupt, and in many cases tyrannical regimes, run by leaders who give to their relatives, tribesmen, henchmen and sycophants the best businesses, and the millions from exports and international aid they receive. Little is left over for the welfare of their people. Little is left over for the welfare of its people.
This raises the question as to the overall relationship of freedom to the average HDI & LEI. To answer I will use the Freedom House ratings for 2003 on the political rights (rated 1-7) and civil liberties (also rated 1-7) of all countries. When I add these two ratings together, the result ranges from a “2″ for the freest to a 14 for the most unfree. When I plot these ratings against the average HDI & LEI, I get the plot below (If it is does not show, click here):
<img src=” http://www.hawaii.edu/powerkills/HDI_LEI_FREE.GIF “
The linear fit is, as a freedomist would predict, inclined downward. That is, the greater the decrease in a people’s freedom, the greater the decrease in their well being. The correlation is r =.50 (r^2=.248), and although this is a good correlation, it accounts for only 25 percent of the variance, the cutoff for what I consider a meaningful social science correlation.
A study of the plot shows that the average HDI & LEI tends to rise at free and not free ends, but less at the latter, and thus creating a dip in the middle. This is a lopsided U-distribution (one side is lower than the other) and suggests a third degree polynomial regression would best fit the points. The best fitting one is shown in the plot. It increases the correlation considerably to R=.61 (R^2 = .367), or 37 percent of the variance.
This is fascinating. For taken at face value, the worst mortacracies are in the middle range between free countries and not free ones. How can this be? Research on democide shows well that the tendency of a regime to commit democide increases as the freedom of its people decreases. While this also shows for mortality (the dipping straight line), the relationship is not as tight as for democide.
I believe the reason for this is totalitarian control over the statistics submitted to the UN. From a variety of memoirs, media stories, UN reports and refugee reports, and those of human rights organizations, we know that life in Sudan, North Korea, Burma, Libya, Ethiopia, and other such countries is dismal, not only with widespread democide, but with high mortality as well. Yet, this is not shown in their average HDI & LEI. To see this, consider the worst of the worst dictatorships, the most totalitarian ones, as rated by Freedom House (If it is unclear or does not show, click here):
<img src=” http://www.hawaii.edu/powerkills/NOT_FREE.HDI.LE.GIF”
I am using LE, rather than LEI, since the former is simply how many years from birth that people live on the average, it is easier to understand. As can be seen, some of the HDI and LE are surprisingly high. For comparison, I provide the HDI and LE for different groups of countries, and for the world (If it is unclear or does not show, click here):
<img src=” http://www.hawaii.edu/powerkills/WORLD_HDI_LE.GIF”
That for the U.S. is .94 and 77.4, respectively.
So far, based on LE alone I have defined a potential group of mortacracies, which however included two liberal democracies. I have refined this by selecting the lowest average HDI & LEI, none of which were free. But the problem with this group is that it did not include what we know to be among the worst mortacracies, such as North Korea, Sudan, and Burma.
Perhaps another approach will work better, such as the change in LE over time, and I will analyze this in Part IV.