The waning power of the trade unions has been an important driver behind rising wealth inequality

On the beautiful lake where I live, the two pairs of common loons, to much excitement, have both started nesting. In the UK, another set of loons won an election without a manifesto or an economic policy and beat the Tories into third place, while the Lib Dems took fifth behind the Greens.

The local elections were devastatingly bad for both coalition parties. Yet the annihilation sent the Labour-hating commentariat led by Dan Hodges into apoplexies, countering that really the big loser was the Labour Party, which according to polls out last week in an important marginal is headed for a comfortable majority of 70 in 2015. Nothing more than sour grapes.

The usual suspects also became very excited about Chris Giles’ comments in the Financial Times where he picked at Thomas Piketty, claiming that his much-heralded data and conclusions regarding wealth and income inequality were deeply flawed. The Spectator’s Fraser Nelson claimed that proved that Piketty had “fiddled” his figures despite the fact that they were all posted online. Giles suggested “the conclusions of Capital in the 21st Century do not appear to be backed by the book’s own sources”.

Interestingly the FT did send Piketty an email at 6.37pm on Thursday 22 May requesting a response to questions about his 685-page book by 3pm the next day. It is highly unusual to rush to publication without giving an author adequate time to respond; less than 24 hours doesn’t cut it.

Piketty’s full response six days later, along with analysis by the economic consultant Howard Reed, support the view that inequality of wealth has been rising in recent years and Giles’ rush to judgement is without foundation.

It is well known that it is difficult to measure wealth inequality because of the importance of measuring the wealth of the wealthiest, who tend not to respond to surveys. The great expert on this is Arthur Kennickell from the Reserve Board of Governors, who has documented the difficulty: “Experience has shown that some groups of survey respondents tend to have higher rates of non-response than others. One such group is wealthy households … who appear generally to be far more difficult to contact and to persuade to participate.”

For this reason especially, analysing and valuing wealth involves a lot of judgement calls; the major question is how sensitive are the conclusions to such judgements and how appropriate they are.

Survey data on incomes and wages are also often a problem because of the possibility of identifying high-income individuals who have been promised anonymity. Statistical agencies often adopt procedures such as top-coding incomes above some number, which frequently means that top-end incomes are understated. Results can be sensitive to assumptions made to close these top-ends which frequently result in published debates among academics on the best course of action. This takes time and there is no place for “gotcha”.

An Associated Press/Equilar survey published last week dealt an immediate blow to the conclusion that income inequality and wealth inequality aren’t rising. The survey showed the head of an average large public US company earned a record $10.5m (£6.2m) in 2013, up nearly 9 per cent on 2012 and the fourth annual year of rises since the Great Recession. These chief executives now earn about 257 times the average salary – up from 181 times in 2009. House prices continue to climb, the FTSE 100 is up by a third since the Coalition took office, and income tax rates for the top earners were cut while the highly regressive VAT was raised. The poor don’t own stocks and shares or houses.

A new paper by Gregg, Machin and Fernández-Salgado* suggests increased sensitivity of real wages to unemployment, in part driven by the sustained decline of trade union bargaining power, which has also been an important driver behind rising income (and wealth) inequality.

New data from the Department for Business Innovation and Skills (BIS) can also shed light on the extent of the decline of unions. The first chart shows the decline in the proportion of workers who are members of unions in both the UK and the US; even though the proportion is higher in the UK the series both show steady declines. In 2013, 23 per cent of workers in the UK were union members compared with 11 per cent in the United States. The second chart shows that these declines have occurred across all regions, with the decline being most marked in the North-east, down from 43 per cent in 1995 to 31 per cent in 2013, and least in the South-east, down from 23 per cent to 20 per cent. Unionisation rates are highest in Wales (35 per cent), Northern Ireland (35 per cent) and Scotland (32 per cent).

The BIS also provided new data on the differences in the earnings of union and non-union workers. In 2013 the hourly earnings of union workers was £14.45 compared with £12.41 for non-union workers, giving a union wage premium of 16 per cent, which has been largely unchanged for the last 15 years or so. But the characteristics of union workers is not the same as those of non-union workers, because they work in different industries, locations and firm types and have dissimilar characteristics such as age, gender and schooling.

Once account is taken of these factors the union/non-union wage differential – which I actually wrote my PhD thesis about – in the UK is around 8 per cent, also broadly unchanged since around 2000. Fewer workers are now earning that premium, contributing to the long-term rise in both earnings and wealth inequality.

Paul Krugman put it well: “The point is that Giles is proving too much; if his attempted reworking of Piketty leads to the conclusion that nothing has happened to wealth inequality, what that really shows is that he’s doing something wrong.”

It certainly looks that way.

* Paul Gregg, Stephen Machin and Mariña Fernández-Salgado, “Real wages and unemployment in the big squeeze,” Economic Journal, May 2014