Fixation with inequality distorts truth

In: Daniel In The News

9 Dec 2011

This is my latest book review for spiked.

I have a mildly embarrassing confession to make. I love statistics. Whenever I start to examine a social or economic question, I gather all the available data to see what it tells me.

Such statistics provide a valuable means to go beyond personal impressions. People I happen to know may tend to hold a particular set of views or behave in a particular way, but they may be untypical. Statistics can often help to provide a more balanced take on social developments and public opinion.

Naturally, I am aware that the use of statistics can be problematic. Politicians, in particular, often sling them around like children chucking snowballs. Rather than trying to establish the truth on a particular question, their concern is often simply point-scoring. There are also more prosaic difficulties in collecting data and constructing statistical measures. Nevertheless, such limitations do not mean that statistics cannot be a valuable aid to enquiry. They just need to be handled with skill and care.

With my stat-geek predilections in mind, I should be expected to like Bankrupt Britain. Daniel Dorling and Bethan Thomas, academics at the University of Sheffield, have produced what they call an atlas of social change. Their goal is to show statistically, along with accompanying graphics, how Britain is changing across several dimensions: financial, residential, moral, emotional and environmental. In principle, it is an excellent idea.

Unfortunately, the execution of the project is fundamentally flawed. The authors are so anxious to validate their central thesis – that inequality is corrupting Britain in all the dimensions they discuss – that their reading of the evidence is often questionable and sometimes bizarre. Their lack of a sense of history or in-depth economic knowledge can perhaps be forgiven. But their perspective is so blinkered that even what should be their strong point, the treatment of statistics, often fails to convince.

Paradoxically, a large part of the problem follows from what is often called an evidence-based approach. From this perspective, the facts speak for themselves. It then follows that the role of the experts is to interpret the data, diagnose the problems and prescribe solutions for social ills. This approach to studying social problems parallels the rise of technocratic leaders in Western governments. In both cases, the challenges facing society tend to be viewed in technical terms and experts are seen as best placed to solve them.

In reality, facts never speak for themselves. They are always interpreted in a broader intellectual framework. Yet Dorling and Thomas are so blinkered they often do not seem aware of the possibility of alternative interpretations of the data. Having strong views is fine – spiked is full of them – but to make a coherent case, it is necessary to be keenly aware of contestable points.

Bankrupt Britain does contain some insights. The chapter on Britain as ‘financially bankrupt’ is the best. It spells out with detailed statistics and vivid graphics different dimensions of inequality and poverty in Britain. The section on pensioners is particularly poignant as it is an important and often-forgotten topic. Evidently almost three out of every five pensioners living in poverty are homeowners. Despite having paid off their mortgages, their incomes are low. Typically, the older pensioners get, the worse their financial plight tends to become.

In contrast, the chapter on political bankruptcy reveals a breathtaking lack of any sense of historical perspective. The authors in all seriousness draw parallels between the political situation in 1918 and the present. They compare Ireland back then, on the verge of a bloody war of independence against British occupiers, with the shoulder-shrugging nationalist sentiment in Scotland today. In addition, they suggest parallels between the Suffragette struggle for women’s votes with the recent scandal around MPs’ expenses.

Such comparisons miss the fundamental political difference between now and then. Back in 1918, only a year after the Russian Revolution, the world was full of a sense of possibilities of alternative ways to organise society. It was widely believed that it was possible to create a better world. In contrast, the current period is one of deep social pessimism. The extreme cynicism about politics and corruption scandals reflects a strong sense that progress is impossible. Yet the authors of Bankrupt Britain are so immersed in their statistics on voting patterns that they are blind to the bigger picture.

The chapter on moral bankruptcy is so naive that, despite the seriousness of the subject matter, it is hard to avoid laughing. Much is made of the fact that Britain’s rate of bullying among children is the highest in the rich world. Yet if the authors had pondered the definition of bullying in a British Medical Journal study they quote at length, they might have seen the absurdity of their conclusions. It includes, among other things, name-calling, writing unkind notes, being unfriendly and being excluded from discussions or activities. It is hard to imagine how any child could go through a school year without occasionally suffering such everyday indignities. The amazing thing about the associated graphic the authors provide is that the highest rate of bullying they could find was 37.9 per cent. It is hard to imagine how any local authority could fail to reach 100 per cent with bullying defined so broadly.

If the authors were less blinkered they would realise that the truly interesting question here is why the British authorities are so obsessed with bullying. Many experiences that would previously have been seen as a normal part of growing up have become pathologised. Meanwhile, the small minority of cases involving serious harm to children are at greater risk of being missed because of the obsession with finding bullying everywhere.

In their conclusion, the authors take their environmentalist perspective, along with their scapegoating of bankers, to appropriately absurd conclusions. Dorling and Thomas construct an index of four different variables to measure local environmental conditions: benzene air pollution, crime rates, electricity consumption and malicious calls to fire brigades (as a measure of anti-social behaviour). Next, they weight each indicator to allow them to construct an index: benzene levels make up 18 per cent of the index figure, reported crime is 32 per cent, electricity consumption is 46 per cent and false alarms is four per cent. They concede that these proportions are arbitrary, but they are mathematically necessary to construct the index. Finally, they calculate which district deteriorated most between 2007/8 and 2008/9 and – ‘hey presto!’ – the answer is the City of London. The irrefutable power of statistics has apparently proved that It Was The Bankers Wot Done It.

There is so much wrong with this procedure from a statistical perspective it is difficult to know where to start. If you insist on constructing such an index why, apart from expediency, choose those particular weightings? And why compare those particular years?

Closer examination only reveals more flaws. The main reason the City scores badly is because of a rapid rise in electricity consumption. Yet rising consumption is generally correlated with more prosperity rather than less. Electricity consumption would generally be expected to rise in times of economic growth and fall during recessions. Indeed, Britain’s electricity consumption fell between 2007 and 2009 according to official statistics from the Department of Energy and Climate Change.

The City was also penalised by the authors for having a high crime rate relative to its resident population. Yet, as Dorling and Thomas acknowledge, the City has a tiny number of residents compared with the large numbers that commute into the area to work. The data therefore gives a misleading impression of the real trend.

Statistics never tell the whole story but they can be valuable investigative tool as long as they are used carefully. If the assumptions or the raw statistics are flawed, the results will be misleading. As with computers, the rule of GIGO – garbage in, garbage out – applies. Bankrupt Britain is a case study in how the misuse of statistics tends simply to reinforce prejudices rather than help illuminate the truth.