Paste your Google Webmaster Tools verification code here

Trends in Inequality: Truth and Myth

Concern about inequalities of income and wealth is now a fashionable topic. It featured strongly in the gathering of the world’s top brass at Davos earlier this year. Much of the popular coverage of the topic gives the impression that not only is inequality at record highs, but that it is confined to the wicked Anglo-Saxon economies. A recent paper published by authors linked to the George Soros-funded Institute for New Economic Thinking shows very decisively that neither of these points is true.

Tony Atkinson, former Warden of Nuffield College Oxford, and Salvatore Morelli examine trends in both income and wealth inequalities in 25 countries since 1900. Their work is a very impressive piece of scholarship, requiring detailed examination of a wide range of data sources. Even so, there are often historical gaps where no information is available. Regrettably, too, the study includes neither Russia nor China, socialist societies characterised by massive inequalities. In the old Soviet Union, for example, in the early 1950s the diet on offer to the general population was no better than that in the labour camps, whilst the Party elite luxuriated.

It is certainly the case that inequalities have increased in the UK and the US in recent decades. A crude but widely used measure of inequalities in inco

me and wealth is the so-called Gini coefficient. Theoretically, this can range between 0 and 100. If it is zero, everyone has literally the same amount. If it is 100, one person has all the income or wealth and no-one else has anything. Obviously, these theoretical extremes can never be observed in practice. The key point is that the higher is the GIni, the greater is inequality. Looking at the distribution of income, in Britain in the 1950s and 1960s, the Gini was approximately 30.

Atkinson and Morelli estimate that in America the Gini coefficient was 7 percentage points higher in 2012 than it was in 1980. For the UK, the increase was even higher, at 10 percentage points, though they note that much of this increase took place during the 1980s. Contrary to popular perception, it has not risen sharply since then. The authors also have a measure of ‘relative poverty’, which is even more interesting. The relative poverty rate in 1990 was twice that of 1977, but “overall the poverty rate has been falling since the 1990s”. In America, it has been “constant since about 1970”, although with cyclical variations around this level.Economics_Gini_coefficient

Strikingly, in the Scandinavian countries, often held up as exemplars of ‘fair’ societies, inequality has risen. In Norway, the Gini coefficient for income is 4 percentage points higher than it was in 1986. In Finland, it is 6 percentage points up on its 1990 level. And in Sweden, since the early 1980s, the Gini has increased by no less than 10 percentage points, just the same as the UK.

Atkinson and Morelli do not offer explanations of movements in inequality, but they provide an excellent example of the value of meticulous empirical analysis in economics.

As published in City AM on Wednesday 19th March 2014

Image Source: Volterra 2014

Read More

How Bad Has It Been? 2008-2013 in Historical Perspective

The end of a year is a good time to take stock. For the first time since 2007, prospects for the UK for the forthcoming year look unequivocally good. But looking back, just how bad have the last few years been across the developed world as a whole? And how do they compare with previous recessions in a historical context? To keep you out of suspense, the answer to both these questions is ‘pretty bad’. In one key respect, it has been awful.

A good piece of news is that, in one sense, the length of the recession in 2008/09 was nothing remarkable. Using the data on GDP on the OECD’s website, the evidence from 20 developed countries across the world shows that during the first half of 2008, output had begun to fall, albeit slowly. This accelerated sharply following the collapse of Lehman’s in September of that year. The only countries to escape a recession at all are Australia and Poland.

By the end of 2009, GDP had begun to increase again across the globe. So on a widely used definition of a recession in economics, it was over. It had lasted a maximum of two years. Looking at all recessions since the late 19th century, this was entirely in line with experience. Ninety per cent of all recessions last for only one or two years.

The size of the recession, in contrast, was rather brutal. In the Euro zone, for example, from peak to trough the fall in GDP was over 6 per cent. In the OECD as a whole, it was 5 per cent. Historically, only a quarter of all recessions exhibit a fall in output of 5 per cent or more. In the Great Recession of the 1930s, in a number of economies, including America and Germany, GDP dropped by over 20 per cent. So whilst not as devastating as the 1930s, this time around, the fall in output was a big one.

The really bad aspect of the recent recession is the length of time it has taken for GDP to regain its previous peak level, before the crash. Historically, excluding the years of the world wars, only 20 per cent of all recessions last more than two years, using this definition rather than the conventional one, where it ends once output stops falling. Just 13 per cent persist for more than three years, and only 6 per cent for more than five. The all-time record, as it were, is the 10 years it took the US to regain its 1929 level of GDP.

America and Canada regained the previous peak level of GDP after some three years. The same is true of Germany and many in its immediate zone of influence, Austria, Belgium, Netherlands, Sweden and Switzerland. Even in these countries, the recession was long. But in 9 out of the 20 countries, output remains below its peak six years afterwards. The UK should get back next year, but the prospects for Italy, Spain, and Portugal look dire.

As published in City AM on Wednesday 18th December 2013

Read More

Rising Residential Segregation, but Less Racial Prejudice: How Can This Be?

Britain is becoming more sharply divided on ethnic lines, according to a study just published by the think-tank Demos. During the past decade, more than 600,000 white people have moved out of London to areas which are more than 90 per cent white. The effect is strongest amongst white Britons with children, with a fall of almost 20 per cent in the number of them living in London.

The Demos project is chaired by Trevor Phillips, the former head of the Equalities Commission. At face value, the numbers suggest that racial prejudice is alive and well. It is not just whites who are seeking out white areas. According to Phillips, ethnic minorities are becoming more tightly clustered in areas where their own personal minority is well represented. But Phillips insists that the overall evidence does not show increased prejudice at all.  In fact, personal prejudice is declining. How can this be? Prejudice is allegedly falling, yet we are becoming more segregated along ethnic lines in terms of where we live.

The answer to this seeming paradox was provided over forty years ago by the brilliant American academic Thomas Schelling. Schelling, based at the University of Maryland, has carried out highly original work in areas such as national security and nuclear strategy, using a game-theoretic perspective. Along the way, he picked up the Nobel Prize in economics in 2005.

Schelling imagined a checkers (draughts) board, with many more squares than a standard one, on which an equal number of black and white checkers are placed at random. A small percentage of squares are left empty. The rules of Schelling’s game are very simple. One of the checkers is chosen to see if it wants to change location. Each square is surrounded by eight other squares. So including its own square, it looks at nine squares in all to see who else lives in its neighbourhood. Suppose the checker is white. Provided that in total there are five white checkers in the nine squares, it is happy to stay put. It does not mind living with black checkers on the other four. But if it is 4-5 rather than 5-4, it moves to one of the empty squares.

The players in this game have a very mild preference for living amongst players like them. They are happy to live with a large minority of the other colour. What happens as the game progresses? Each checker can decide, one at a time whether or not to move. Then they all get another chance to move if they want, given the new pattern of location. Remarkably, the board segregates very rapidly into dense blocks in which the checkers are all the same colour. It appears as if the players have a very strong preference to live near players of the same kind, which is not the case at all.

Schelling’s game is of course highly abstract, but it has profound practical insights. Increased tolerance and increased residential segregation need not be incompatible at all.

As Published in City Am on Wednesday 27th November 2013

Read More

Political map of London is like America: strong geographic segregation

Thomas Schelling is a brilliant American polymath, who deservedly won the Nobel Prize in economics in 2005.  One of his most remarkable insights is about segregation in cities, which he published as long ago as 1971.

The residential pattern of American cities tends to be pretty sharply divided on ethnic grounds.  The population of many areas is often overwhelmingly drawn from a single ethnic group.  There are white neighbourhoods, black neighbourhoods, as so forth, in which there are very few members of other ethnic groups.

An obvious implication of this seems to be that there is strong racial prejudice in the US, that many people actively prefer to live amongst people of their own ethnicity.

Schelling showed that strong segregation at the level of the city as a whole can arise even when individuals have only a very weak preference in favour of being surrounded by people of their own ethnic group.  There is a big, often highly mathematical, scientific literature on this in the four decades since Schelling made his discovery.  But his basic finding still remains valid.  Very weak individual preferences often translate into apparently very strong ones at the city-wide level.

His work was all the more remarkable given that personal computers had not been invented. Schelling obtained his results using coins on graph paper.  He placed pennies and nickels in different patterns on the “board” and then moving them one by one if they were in an “unhappy” situation.

The first preferences in the 2012 London Mayoral election are a remarkable example of city-wide segregation. They give the impression of a city which is sharply divided in its political allegiances and outlook.

Of course, many factors determine electoral outcomes. And the chart is not implying that people select their place of residence according to its political preferences. But, certainly, there is a self-reinforcing aspect to this process, which Schelling did not incorporate into his original model. The underlying maths of these models came much later.

The overall culture of a locality is an important determinant of an individual’s political preference. Many wealthy left-wingers, for example, choose to live in the Labour areas immediately north of the river in the centre of the city. They do so because of the local culture, which in turn reinforces their own opinions and voting habits. So we have a Schelling-type process underpinning the electoral map, overlaid with the kind of self-reinforcing feedbacks which pervade modern life.

A very American outcome in Europe’s leading city!

Read More