Return to Blog

Inequality

I am still working on this. Comments via the Blog would be very welcome.
DateChange
10 Oct 2010Initial version
17 Nov 2010Major reordering to put conclusions and discussion on index page and all graphs on linked pages.
30 Nov 2010Start to introduce corruption as acting together with inequality as a barrier to getting wealthy.
2 Oct 2011Added page Corruption and Inequality are the joint enemies of wealth

Conclusion

I present a simple and robust interpretation  of reputable data which supports the conclusions in The Spirit Level and shows that:

The logical outcome should be a policy of deliberate and measured reduction in inequality, but what would the effect of this be? I present some estimates of the changes that would occur at different income levels. 

Here are some proposals for: 

What to do?

Campaign for a target for a reduction of inequality.

A modest initial target might be a 2% reduction of the Gini coefficient.[1]

Half the population will see a modest gain and for nearly all the rest the change will be small. Only the richest 3% will see a modest loss, with a tiny proportion, one in a thousand, of already very well off people will be made 10% to 30% less well off. Whatever the target, 85% of the population will be better off; This ought to be a vote winner, surely?

The aim would be to reach this target in, say, two years. This will give time for people to adjust to the change and a new target would be set for a further 2 point reduction in the next two years and then again for the two years after that.

The hope is that a modest medium term goal of a Gini coefficient of 30% could attract broad, all-party support.

Why?

How?

The aim should be to establish a broad consensus for a steady reduction of inequality.

There are a number of policies that might be pursued, eg:

However aim of this campaign is only to set a target and not to direct the means.

What would be the effect?

This table, which is based on an estimation the consequences of reducing inequality , shows what proportion of the population will loose and gain:

 Gini 36% > 34%Gini 36% > 30%
Significant gain, >10%0%44%
Modest gain, 3% - 10%52%35%
Insignificant change, less than ±3%45%12%
Modest loss, 3% - 10%3%6%
Significant loss, 10% - 30%0.1%3%
Major loss, > 30%00.1% 

The poorest half of the population are going to be better off but not dramatically so. The middle income earners can be reassured that it will make very little difference. Even the very richest can to reassured that they are still going to be very well off and not reduced to poverty. (0.1% of the population have an income greater than about £400,000 pa.)

An essential requirement for this campaign is the ready availability of up to date figures for inequality from a respected independent source.

Discussion

A long time ago I formed the opinion that excessive inequality is a Bad Thing and I thought it likely that there would be some sort of correlation between Bad Things and inequality. It seemed to me that, in extreme cases, very unequal countries would be divided between the vast majority, suffering the effects of poverty and a tiny minority, the rich, for whom falling into poverty is unthinkable. The prospect of poverty is so horrific for the rich that they will embrace anything that supports their situation, no matter how immoral or irrational. In particular they would favour actions that increased inequality even further. Thus inequality can be both a cause and an effect of Bad Things.

I felt that inequality was a factor that should be present in political dialogue as a quantitative measure that could and should be controlled, similar to the situation with inflation. Although commentators made statements like "X% of the population has Y% of something, Isn't it awful?" I saw little joined-up discussion. In particular, I was once dismayed when, in lunch-time conversations with some strongly left-wing colleagues, my suggestion that inequality was an important factor was rubbished. This attitude seems to be alive today. 

In 2007 using statistics published by the UN on the web, I produced a note: Does Inequality Matter? , for a discussion group

In 2010 year The Spirit Level  was published and has met with a considerable response and, at last, the issue of inequality seems to have entered the wider political consciousness. The conclusions of this book, that more equal countries usually do better, is based on evidence from the analysis of reputable data. Critics of this book have questioned the robustness of the authors analysis. The problem, as I see it, is that the statistical technique, regression analysis, used by the authors is good for teasing apart the the relationships between various factors, which is what you want if you are trying to understand the complexities of the real world but sometimes difficult to support in the face of hostile criticism.

I have looked at similar data in a different way  using boundaries rather than regression lines. This allows us to make making statements of the form: "no country has done better than this" rather than "there is a linkage between these two variables". No countries for which relevant data is available was excluded. I suggest that the conclusions are more robust than those from regression analysis but maybe less subtle.

Conclusions so far from the analysis of boundaries :

Inequality
  No country has become more well off without also being more equal and less corrupt. 
Infant mortality
 
1The lowest rates of infant mortality have been achieved by richer countries.
2Lower rates of infant mortality have been achieved by more equal countries although this is less strong effect than income.
Unhealthy life expectancy
 
1The lowest unhealthy life expectancies are found in countries that are more equal and less corrupt.
2The lowest unhealthy life expectancies are found in countries with a middling income (about $US10,000 p capita) with worse results for both the lower and the higher incomes.
Happiness
 
1The happiest counties are more equal and less corrupt.
2Low income has little effect in preventing countries from being happy.
Unhappiness
 
 The unhappiest countries are less unhappy if they are richer or more equal.
Suicide
 
1The lowest suicide rates are found in the poorest countries
2The lowest suicide rates do not depend on inequality
Life expectancy
 
1 No country has achieved a higher life expectancy without a higher income
2Life expectancy does not seem to be affected by inequality

The last conclusion which will disappoint some but perhaps we should not be surprised; to be 70 today means not dying in any of the last 70 years and for many of the countries that score well, inequality has fluctuated considerably in this period. (For some of us the war meant a paradoxically healthy childhood.)

There are remarkably few countries that don't fit this analysis and there is either a convincing reason or a high likelihood of inaccurate data.

[1] The Gini coefficient  is a scale of inequality with 0% representing complete equality, ie where everyone has the same, and 100% for the ultimate inequality where one person has everything. Actual values range from 25% for Japan and some of the Nordic countries to 80% There are other scales, such as the ratio between the incomes of the top and bottom 10% of the population and the Pareto index. The conclusions would be the same for them all. The Gini coefficient seems to be the one preferred by economists.