In my last post on the Global Peace Index (GPI) I implied that potential users (such as myself) who are primarily interested in the internal peacefulness of different societies would be more interested in a sub-index excluding the effects of militarization. It so happens that a sub-index of the Legatum Prosperity Index, the Safety and Security Sub-index (LSS), may provide a good measure of internal peacefulness.
Indicators covered in the LSS include standard measures of violence such as homicides and assault, refugees and displace persons, flight by professionals (brain drain), civil war and ethnic violence, survey information relating to incidence of theft and perceptions of whether people feel safe walking alone at night. The safety and security indicators covered in the GPI seem to be broadly similar.
The LSS and GPI are compared in the chart below. The position of each blue diamond reflects the scores on both index for each country. The pink diamonds are predicted scores of the GPI resulting from a linear regression.
It is apparent from the chart that both indexes present a similar picture of the peacefulness of the vast majority of countries. The two main exceptions are Israel and the United States – which have a higher level of internal peacefulness than others with similar GPI scores.
The creators of the Global Peace Index (GPI), the Institute for Economics and Peace (aka Vision of Humanity), claim that their index represents the first attempt that has been made to rank the nations of the world by their peacefulness. The GPI is apparently the brainchild of Australian entrepreneur, Steve Killelea, but has been developed in conjunction with the Economist Intelligence Unit and with the guidance of an international team of academics and peace experts.
As noted in the 2009 GPI report, peace is notoriously difficult to define. The approach adopted was to define peace as ‘absence of violence’ and to use metrics that combine both internal and external factors to rank144 countries. The Index is composed of 24 indicators covering three broad categories: measures of ongoing domestic and international conflict, measures of safety and security in society and measures of militarization.
The third category seems to me to raise a problem because it may sometimes be possible for us to promote peace more effectively by sending potential aggressors a credible message that if they mess with us they will suffer hellish consequences, rather than by disarming to convince them that we do not pose a threat. Instead of basing the index on the assumption that unilateral disarmament would promote peace I think it would have been preferable to leave this question open for further research.
For the benefit of potential users (such as myself) who are primarily interested in the internal peacefulness of different societies it would be desirable for the creators of the GPI to publish sub-indexes which exclude the effects of militarization (and international conflicts).
In view of the way the GPI has been constructed it is not surprising that a country like New Zealand, which does not have any reason to feel threatened by any other country, should be given the highest ranking as a peaceful country. The United States could be expected to have a somewhat lower ranking because of its greater militarization. I was surprised, however, that the GPI ranking of the U.S. is as low as 83 out of 144.
The relatively low ranking of the U.S is not entirely attributable to militarization. The percentage of the population in jail in the U.S. is apparently higher than in any of the other countries included in the index and the rate of homicide is higher than in New Zealand (and many other countries including Australia and the Britain).
The chart below indicates that it is possible for a country to have a relatively low GPI ranking while remaining relatively peaceful. For example, despite its relatively low ranking, the GPI score of the U.S. is much closer to that of New Zealand than to that of Iraq.
One final thought might be worth noting. This measure of the peacefulness of nations does not take account of the extent that different governments pursue policies that induce some of their residents to send their capital to other jurisdictions to avoid confiscatory taxation. This omission may be important for users who are interested primarily in the internal peacefulness of different societies.
Before attempting to answer this question I should recap some previous posts about the good society.
Is there such a thing as a good society? Yes. A good society is one that has institutions that are good for its members.
What are the characteristics of a good society? I suggest three characteristics:
• Its institutions enable its members to live together in peace.
• Its institutions provide its members with opportunities to flourish – to obtain the things that it is good for humans to have.
• Its institutions provide members with a degree of security against threats to individual flourishing.
Are the institutions of a good society the same as those of the “great society”? I suggest that there is a lot of overlap between the institutions of the good society and the liberal principles of the great society, or open society, as discussed by Friedrich Hayek.
Why is it wrong to coerce people for their own good? Richard Kraut raises this question in his book, “What is good and why”. My answer is that it is wrong to coerce people for their own good because this is not consistent with living in peace with them.
Should we expect the rules of the good society to be good for everyone? Richard Kraut questions societal rules such as those discouraging theft and interference with other people because they are not always good for everyone. The point I make is that the rules for the good of all – rules that nearly everyone would agree to behind a veil of ignorance about their own particular interests and vulnerabilities - are not necessarily good for everyone.
So, I think I am on the way to identifying the characteristics and institutions of a good society. Having reached this point the thought has crossed my mind that it might not be too difficult to identify indexes that others have constructed that measure the various characteristics and institutions of a good society. If I were to combine those indexes into a good society index that would make it easy to classify various societies as bad, not so bad, OK, good, better, best. (I might even be able to think of a better classification.) A composite index of this kind would make the statistical feast more easily digestible and might even grab public attention (and perhaps even make me rich and famous).
I see a couple of problems with this idea, apart from the possibility that the world might already have just about enough composite indexes measuring broadly similar things (and fame would probably not be good for me). First, as I noted in my article ‘Gross National Happiness’, recently published in Asian-Pacific Economic Literature (abstract; draft), there is the problem of finding appropriate weights for each item included in a composite index. It is not really good enough to just have a set of weights that reflect my own values. Even if I could find a Nobel prize winner, or two, prepared to endorse a particular set of weights that would not really help much because people do not get Nobel prizes because their values are universally accepted. Ideally, every user of the index would be able to use a set of weights reflecting their own values, but then we would have a multitude of indexes rather than a single index.
Other options include use of equal weights or weights that come out of regression analyses that use the sub-indexes to explain per capita GDP levels or average life satisfaction in different countries. Equal weights are intrinsically mindless and the second option would not be much better because the ‘good society’ index resulting from it would not amount to much more than an index of average of per capita GDP or average life satisfaction.
The second problem is that composite indexes tend to hide the relationships between sub-indexes. It seems to me to be more useful to make observations about such things as differences in life expectancy in countries with differing levels of economic freedom than to be able to classify particular societies as being good, bad or indifferent using a single index number.
I have decided that a good society index would not be a good idea. In later posts I will attempt to identify the suite of indicators that would be most relevant to assessing to what extent particular societies might qualify as good societies.
Prosperity. However, my initial reaction when I first looked at the index was that it was more like an index of the quality of life or well-being than of prosperity. My perception of prosperity was too narrow. One of the purposes of the index is apparently to encourage people ‘to take a holistic view of prosperity’. The authors might have a point. Dictionary definitions of prosperity don’t focus solely on economics – they suggest that prosperous is synonymous with flourishing, successful and thriving.
The index has nine components or sub-indexes:
• Economic fundamentals
• Entrepreneurship and innovation
• Democratic institutions
• Education
• Health
• Safety and security
• Governance
• Personal freedom
• Social capital.
Is this a comprehensive list of factors affecting human flourishing? One apparent omission is environmental quality. Although this is considered to some degree in the indicators used for health, I get the impression that the authors’ view of prosperity does not place much value on the enjoyment that people obtain from the natural environment. I wonder whether that view is backed up by research findings.
The main problem I have with this report is that I am not sure how much substance lies behind it. The presentation is incredibly smooth, mainly because it is completely uninterrupted by references to research papers or data sources. The preface suggests that information on data sources is available on the prosperity web site. All I could find on that site, however, apart from the acknowledgement of input from research consultancy, Oxford Analytica, is a statement that the index draws on the Gallup World Poll and other data and includes only factors for which a statistical link with material wealth or life satisfaction can be shown.
The Legatum Institute claims to be a think tank. It seems to me that no organisation claiming to be a think tank should publish research findings without making sure that the underlying research is open to public scrutiny.
However, leaving aside my doubts, this report is worth looking at just to view the chart showing that countries with high prosperity tend to have high performance on all sub-indexes and those with low prosperity have low performance on nearly all sub-indexes.
Addendum:
Hamish Banks of Legatum has responded as follows:
"Thank you for writing about the Prosperity Index; may I refer you to the Prosperity Index website where you may download both the 55 page Methodology and 85 page Technical Appendix, which should answer your questions on the the data sources? (http://prosperity.com/report.aspx)
So far as possible we have been transparent with the sources and in some cases the data is publicly available; in others, like the Gallup World Poll, it is proprietary and not ours to publish.
You are right in surmising that the non-inclusion of environmental quality is based on research findings: we did not simply choose the factors that appealed to us, but regressed more than 200 variables to the 79 for which there was a clear and strong correlation with prosperity - environmental quality qualified only, as you note, in the general health indicators."
