//
you're reading...

Datablog

The changing relationship of wage inequality, growth and public spending

Since the 2007 financial crisis the general narrative given by politicians and key economic variables is that of an country that is unequal, in intermittent recession with a government intent on cutting spending. Understanding how economic growth or public spending has contributed to changes in wage inequality can provide a good framework to argue for or against current policy. A brief analysis of the three variables makes three points clear:

  1. Under Thatcher’s government the British economy underwent a significant change that has not been seen since.
  2. After 1990 the strong relationship between public spending and inequality no longer holds true.
  3. Economic growth consistently increased with inequality throughout the 1980’s

The use of a scatter diagram with a line between each sequential year shows the path the economy has taken over time, displaying economic transitions to great effect (as done by Alberto Cairo here and Hannah Fairfield here). Below I have used the same methodology to compare Britain’s economic inequality to growth, and then to public spending, highlighting the different periods of government.

I have used the Gini coefficient to measure inequality, where a value of 0 would represent complete wage equality, and 100 representing one individual holding all of the economies wealth. These figures are based on disposable income, which should give a sense of the difference in lifestyle.

Britain’s economic path measured by economic growth and wage inequality from 1977 to 2010

 The line represents the path of the economy over time. Higher wage inequality leads to a higher position on the chart, while increases in GDP is presented by movements to the right. Election years have circular markers, and the line color represents the different Labour/Conservative periods. Hover over points for more information, the line color indicates party, leader changes represented by images of the leaders, election years are circular.

For me what is immediately clear is that economic growth is not shared evenly across the economy evenly, otherwise there would be a horizontal, flat line. The shift made in under Thatcher’s government becomes clear, and the disparity in wages has remained since. While this graph doesn’t prove this, a possible cause is the boom in the financial sector and decline of the mining and shipping industries benefitted the wealthy more than the poor but leading to overall economic growth. This said it could just as easily be argued, as it was often argued by Thatcher, that Britain became more of a meritocracy which inherently leads to greater inequality.

What struck me as particularly interesting was the lack of any change in equality after New Labours election in 1997. In the 20 year period from John Major’s government to the 2010 election there was little shift in inequality yet a significant shifts in public spending. Despite being on the left of the political spectrum Labour saw wage equality change little more than John Major in its first two terms. Even though New Labour was very different from the party it was under Callaghan it still seems counter intuitive that higher government spending did not increase wage equality. The graph below, comparing public spending to the Gini coefficient, suggests exactly that.

 

Related before 1990, unrelated after 1990.

 This chart does does the same as the previous, swapping GDP for public spending as a percentage of GDP. The further to the right the line reaches, the larger the role of the state, the higher the line reaches the higher the inequality. 

Under Thatchers government there appears to be a strong negative relationship between public spending and wage inequality, indicating a bigger government decreases wage discrepancy.  That relationship does not hold from 1990’s onwards there is little correlation with inequality. If you exclude the 2007 financial crisis where bailouts significantly increased expenditure, public spending shifted between 34.5% and 43.7% of GDP. After Thatcher the size of the government changed just as much with no equivalent impact on wage equality.

To test the strength in the relationship between GDP, Public Spending and the Gini coefficient I performed a couple of regressions, essentially the same as two graphs above except it finds a line of best fit as opposed to drawing a line between each sequential years datapoint. This is a more rigourous method of testing how the relationship between the three variables changed before and after Thatcher left office.

It is worth pointing out that in both time periods the relationship between public spending and the Gini coefficient is as you would expect it, as spending decreases inequality increases. Yet the impact that a change in spending has on inequality dropped dramatically after 1990. Before 1990 a percentage increase in public spending as a proportion of GDP would decrease the Gini coefficient by 0.733, significant given that over the last 30 years the Gini coefficient has only varied by 10. After 1990 a percentage change in public spending would only decrease the Gini coefficient by 0.132. In fact after 1990 the relationship is so inconsistent it should perhaps be considered insignificant, this is suggested in part by its Rsquare value a measure between 0 and 1 which values the goodness of fit. I have put the relationship in full below:

Pre 1990 Gini-Public spending relationship:
Gini = -0.7332*(Public Spending as % of GDP) + 62.748
Rsquare = 0.691

Post 1990 Gini-Public spending relationship:
Gini = -0.1325*(Public Spending as % of GDP) + 39.941
Rsquare = 0.185

So why does the relationship change? There could be multiple answers. One may be that the public spending performed under Major, Brown and Blair were not effective at addressing inequality, or that it is easier to create an unequal economy by decreasing the size of the state than it is to do the reverse. You could use the old adage that correlation does not mean causation; while Thatcher significantly shrank the size of the state and oversaw a rise in inequality perhaps these were effects of the economy fundamental structure shifting as opposed to one causing the other. In which case the whole concept that an public spending has any effect on equality would be false.

the impact that a change in spending has on inequality dropped dramatically after 1990

Comparing GDP and the Gini coefficient tells a similar story, two variables that had a very strong relationship prior to 1990 under the Thatcher government broke down subsequently. In comparison the public spending, GDP has a lesser influence on inequality, for every percentage increase in GDP prior to 1990 the Gini coefficent would increase by 0.2. Inequality increasing with GDP shows that growth in the economy over this period was not shared equally. This appears to makes sense- bankers in London found 1980’s much more lucrative than coal miners across the country. This is a particularly strong relationship, with an R square value nearing to one.

After 1990 not only does the relationship collapse, it also reverses, an increase in GDP by 1 percent would decrease the Gini coefficient by -0.01 (barely significant). In other words economic growth was shared much more equally than it was in the 1980’s, while you wouldn’t statistically reject the direction of the relationship it has a much weaker fit (its standard error 0.876 so the coefficient is not reject, however its R square is 0.07).

Why did it change? I think its probably fair to say that Labour did do more to ensure that growth was shared more equally, but there were also industries that boomed which benefited all of the UK, namely technology. After 1990 the relationship between economic growth and equality was inconsistent, which would seem to support the analysis that growth in some time periods are shared more evenly than others. Here is the relationship between GDP and the Gini coefficient in full:

Pre 1990 GDP-Gini
Gini = 0.0337(GDP) + 5.192
Rsquare = 0.934

 
Post 1990 GDP-Gini
Gini = -0.0012(GDP) + 35.923
Rsquare = 0.0781

As a final note its probably worth giving some sense of what a change in the Gini index looks like- saying that it has risen from 26 to 36 doesn’t mean very much to very many people. The graph on the below shows how wealth has been shared in the UK over the last 30 years dividing the populations into five groups, from the wealthiest 20% (at the top) to the poorest 20% (at the bottom). The bigger their area gets the more disposable income they have.

The most significant shift came under Thatcher, which has been highlighted, it becomes clear that the wealthiest 20% had a significant increase in wealth, while the lowest 80%, fairly equally, lost a share of wealth. I should warn that this shouldn’t be seen a indication that the poorest had a lower disposable income but a lower share of the total. Given the change in GDP over the period, it could easily be that the rich got richer while the poor simply maintained the amount of wealth but a lower proportion of the economies wealth.

If you want to live in an economy which is more capitalist society, with more competition, where the top 20% own 43% as opposed to 36% of the economies disposable income then you’ll probably favour the changes that took place under Thatcher.

In terms of what Thatchers legacy and analysing her legacy, it really falls down to your own personal beliefs. If you want to live in an economy which is closer to a capitalist society, with more competition, where the top 20% own 43% as opposed to 36% of the economies disposable income then you’ll probably favour the changes that took place under Thatcher. If you think the jump from 43% to 36% was too much, believe the poor were and are too poor, then Thatcher becomes much less favourable.

What is perhaps more important to take away is the implications of past data to today- that a relationship held in a previous decade will not necessarily hold for the next. Neither increasing public spending or growing the economy are a certain paths to a more equal society, and anyone who provides that narrative leaves themselves open to being proven very wrong indeed. Perhaps a good question to ask is if it is more important to understand what the public sector spends money on, what parts of the economy grow, and which ones dwindle. If you really wanted to be a bit more forward thinking, perhaps its time that we place less weight on macroeconomic variables like GDP and public spending, instead developing better indicators to measure our economy.

Sources:

Data

  1. GDP: Office of National Statistics 
  2. Gini Coefficient: Office of National Statistics
  3. Public spending: Guardian
Note: all images used are free to use in the public domain. They can be found on the respective prime ministers wikipedia page along with their respective sources and copyright guidelines.

Data set in full
Year Gini coefficients 1977 to 2010/11 Annual GDP, £ billion Public Spending %GDP PM
1977 27.2 646.3 45.6 Callaghan
1978 26.6 667.5 45.1 Callaghan
1979 27.4 686.5 44.6 Callaghan
1980 28.6 672.7 47 Thatcher
1981 29.0 664.2 47.7 Thatcher
1982 28.6 679.2 48.1 Thatcher
1983 29.1 705.2 47.8 Thatcher
1984 28.4 726.0 47.5 Thatcher
1985 30.0 754.0 45 Thatcher
1986 31.6 786.5 43.6 Thatcher
1987 33.2 827.0 41.6 Thatcher
1988 35.1 873.1 38.9 Thatcher
1989 34.4 895.7 39.2 Thatcher
1990 36.8 912.1 39.4 Thatcher
1991 35.6 895.8 41.9 Major
1992 34.7 903.5 43.7 Major
1993 34.8 931.5 43 Major
1994 33.8 974.1 42.5 Major
1995 33.0 1005.1 41.8 Major
1996 34.4 1036.3 39.9 Major
1997 34.5 1076.3 38.2 Major
1998 35.4 1114.2 37.2 Blair
1999 35.8 1149.5 36.3 Blair
2000 35.0 1198.1 34.5 Blair
2001 36.2 1232.7 37.7 Blair
2002 33.8 1262.7 38.5 Blair
2003 34.0 1310.9 39.3 Blair
2004 32.8 1349.0 40.5 Blair
2005 33.9 1386.4 41.2 Blair
2006 34.7 1422.5 40.9 Blair
2007 34.2 1474.2 41 Blair
2008 34.4 1459.9 44.5 Brown
2009 33.5 1401.9 47.7 Brown
2010 34.0 1427.1 46.8 Brown