Wednesday, November 25, 2009

What institutions explain the peacefulness of societies?

I don’t think many people would dispute the idea that the observance of common rules plays an important role in making it possible for individuals to live together peacefully in society. Even if I were to quote something written by Friedrich Hayek that emphasizes the importance of that idea, I doubt whether that would provoke many people to dispute it. (For some suitable quotes from Hayek, see: Are the institutions of the “good society” the same as those of the “great society”? )

 
However, in this instance accepting the logic of a proposition is only one step along the road to understanding its practical significance. It might be useful to know to what extent the peacefulness of societies can be attributed to various factors associated with observance of common rules – factors such as political stability, the quality of legal institutions, civil liberties, and strong social networks and relationships. I report below an attempt to assess the relative importance of these factors using regression analysis.


The measure of the peacefulness of societies used in the analysis is the Legatum Institute’s safety and security sub-index (LSS), which is a component of its Prosperity Index. 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. (I compared the LSS and the Global Peace Index here.)


The measure of political stability is the World Bank’s political stability and absence of violence index which captures perceptions of the likelihood of governments being overthrown by unconstitutional or violent means.


The measure of the quality of legal institutions is the World Bank’s rule of law index. This index captures perceptions about contract enforcement, property rights, the police and the courts, as well as the likelihood of crime and violence. The appropriateness of this index for current purposes is discussed here.


The measure of civil liberties is the Freedom House civil liberties index. This index covers freedom of expression and belief, associational and organizational rights, rule of law and personal autonomy.


The measure of social networks and relationships is the Legatum Institute’s social capital sub-index (LSC), which is a component of its Prosperity Index. The coverage of the LSC includes perceptions relating to reliability of others, importance of friends, trustworthiness of others and membership of various types of community organizations.


To enable the estimated contribution of the explanatory variables to be compared more readily all indexes have been converted to the same form as the Legatum indexes i.e. with highest value equal to one and the lowest value equal to zero. Estimated coefficients (with standard errors in brackets) are as follows:


• Intercept:                              0.06 (0.03)
• Political stability:                 0.46 (0.06)
• Quality of legal institutions: 0.42 (0.06)
• Civil liberties:                     -0.02 (0.05)
• Social capital:                     -0.01 (0.04)          Adj. R squared = 0.82


The results confirm that political stability and quality of legal institutions play an important role in explaining the peacefulness of societies. They suggest that civil liberties and social capital do not play a significant independent role in this context, but it is possible that these factors may contribute positively to political stability and the quality of legal institutions.


A chart showing how well political stability and quality of legal institutions predict the peacefulness of societies is shown below.



It is apparent from the chart that actual peacefulness is never much greater than predicted peacefulness. This means that political instability and poor quality legal institutions virtually guarantee that a society will not be peaceful.

Saturday, November 21, 2009

Does the World Bank's rule of law index measure the quality of legal institutions?

The authors of ‘Governance Matters’, Daniel Kauffman, Aart Kraay and Massimo Mastruzzi, tell us that the World Bank’s rule of law index captures “perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police and the courts, as well as the likelihood of crime and violence” (Working Paper 4978, p 6).



On the basis of that description the index seems highly relevant to assessment of whether societies have institutions that enable their members to live in peace with one another. (For background on reasons why I am interested in such indexes see: Is a ‘good society’ index a good idea?)


I think it would be more appropriate to describe this index as a legal institutions index than as a rule of law (RoL) index. The rule of law is the ancient principle that no-one, not even the king, is above the law. It possible for a jurisdiction to have a relatively high score on the World Bank’s RoL index even though its legal foundations for rule of law may be somewhat tenuous e.g. Hong Kong. (Someone might be interested in a previous post on the question: Is rule of law an esoteric concept?)


As with the five other indexes in the World Bank’s suite of governance indicators the RoL index is based on perceptions based data reflecting the views of a diverse range of people, including tens of thousands of household and firm survey respondents and thousands of experts working for the private sector, NGOs, and public sector agencies. The aggregation method gives greater weight to indicators that are correlated with each other.


The RoL index seems to cover similar ground to the Legal structure and property rights sub-index (LSPR) of the Fraser Institute’s economic freedom index. Indicators incorporated in the LSPR cover: judicial independence, impartial courts, protection of property rights, military interference in legal and political processes, integrity of the legal system, contract enforcement and regulatory restrictions on sale of property.


The chart below shows how closely the World Bank’s RoL index and the Fraser Institute’s LSPR index correspond to each other. The blue diamonds represent actual indexes and the pink diamonds represent the predicted value of the LSPR index using linear regression.




In later posts relating to good society indicators I will use the World Bank’s RoL index as an index reflecting the quality of legal institutions.

Friday, November 20, 2009

How closely does the global peace index reflect internal safety and security?

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.

Wednesday, November 18, 2009

What does the global peace index measure?

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.

Sunday, November 15, 2009

Is a 'good society' index a good idea?

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.

Thursday, November 12, 2009

What does the Legatum Prosperity Index measure?

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."

Friday, November 6, 2009

How do preferences relate to well-being?

Readers of this blog will know that I am attracted to a developmental view of individual well-being. I have previously argued that if happiness is viewed as a positive emotional state (involving, for example, peace of mind, optimism, uncompression, exuberance, flow, joy and cheerfulness) then well-being cannot be the same thing as happiness. Well-being also involves other factors - including personal security and security of property, health and longevity, access to goods and services – as well as a positive emotional state. (See: What are the links between freedom and flourishing?) The kind of life that is best for all humans is a life of flourishing - one that follows a pattern of psychological and physical growth filled with enjoyment.

 
What does a developmental view of individual well-being have to say about the relationship between well-being and the satisfaction of preferences? Richard Kraut’s questions whether a more preferred outcome is necessarily a better outcome for an individual, even if it satisfies her properly informed and rational desires. His point is that it is possible to have properly informed and rational desires that are not good for us.

 
One example Kraut gives is a desire for fame:


“There is no way to criticize the desire for fame except to say that its object is not something that it is good to have. But it should not be at all surprising that human beings sometimes have desires whose objects fall into this category. Nature has not installed in us some wonderful mechanism that guarantees that what propels us forward and focuses our minds on certain courses of action will bring us to something that it is good for us to have. Somehow or other, we have to take steps to learn about what is good for us, and even to care about what is good for us; that is not a topic about which we inevitably come to have tacit knowledge merely by virtue of having desires” (“What is good and why”, p 185).


Daniel Hausman and Michael McPherson note that if well-being is not the same as preference satisfaction then the normative significance that economists attach to preference satisfaction would appear to be without foundation. These authors then offer a new partial defence of welfare economics based on the view that if people are more or less self-interested with respect to certain alternatives, then economists can use their preferences to make inferences concerning what they believe will benefit them. This means that if it is reasonable to suppose that individuals are good judges of what will benefit them, then economists can use people’s preferences as evidence concerning what in fact makes them better off (‘Preference satisfaction and welfare economics’, Economics and Philosophy: 25 (2009).


What happens if we are dealing with policy issues where evidence has accumulated that individuals are often poor judges of what will benefit them? Does this mean that paternalistic government intervention will produce better outcomes? No. The argument that imperfections in individual judgement justify government intervention is just as fallacious as the old argument that departures from perfect competition in markets justify government interventions to increase competition.


In order to consider whether intervention is warranted it is necessary to consider the consequences of alternative institutions, or rules of the game. The judgements that governments make on behalf of individuals may be worse than the judgements that people make on their own behalf. For example, compulsory superannuation in Australia forces some young people to begin to save for their retirement at a time of their life when they would probably be better off saving to buy a house.


In addition, it is important to consider the consequences of government interventions that displace personal responsibility. It is good for adult humans to accept responsibility for decisions affecting their own well-being because this contributes to their personal development and self-respect.