Per capita GDP: where’s my 27 grand?

Photo: William Warby @ Photo: William Warby @

The Matchstick Man is an economic illiterate who likes to ask awkward questions. This month: what is per capita GDP and how much do I get?

Last week I read that we’d passed yet another milestone in our recovery from the Great Recession which began in 2008 with the banking crisis and may or may not still be going on, depending on whom you speak to (economists can’t seem to agree on when a recession ends). According to the latest official figures, something called “per capita GDP” has finally returned to the level it was before the crisis began. Yes, the good times are well and truly back. We can all go back to spending money we haven’t got on things we don’t really need and boasting in the pub about how our house is worth more than Moldova or something.

Well, I don’t know about you, but I still feel a good deal worse off than I remember feeling seven years ago, when there were no beggars outside the Mini Mart and you could have a good night Up West for a fiver. So I thought I’d dig into this per capita GDP thing and find out what it means. You won’t be surprised to learn that things aren’t quite what we’re led to believe.


GDP, or “Gross Domestic Product”, tries to measure how big or prosperous an economy is. It is a measure of “national income” or output – the total value of everything produced or done in the country by everyone – firms, the government, charities and individuals. It even includes an estimate of earnings from illegal and black market activities like drug dealing and prostitution (best not ask how the stattos go about “estimating” that).

Statisticians measure GDP in three ways – the output measure, the income measure or the expenditure measure. As everything we produce has to be paid for, and those payments go to someone or something, the three measurements should add up to the same thing. Actually, there are some minor statistical differences that the Matchstick Man can’t be bothered with. It probably is worth knowing that the number crunchers do adjust the figures to take account of inflation, so the growth in GDP we read about means real growth, not just that the same things are more expensive.

So if GDP goes up, we’re better off right? Well, it turns out, not necessarily. We could just have more people. If GDP goes up by 5% and the population grows by 5% at the same time, on average we’re no better off. We’re producing more and earning more as a country, but we’ve got more people to share it between. The famous French economist Thomas Piketty, who has researched growth figures for many countries since about the time of the dinosaurs, estimates that around half of economic growth in the last 100 years just comes from population growth. It’s worth remembering that next time you hear a politician trying to claim credit for economic growth.

WTF is per capita GDP?

Because GDP on its own doesn’t tell us much, economists also look at Per Capita GDP (per capita means per head). This is simply GDP divided by the total population: our own individual slice of the national pie. Except no one actually gets a share of GDP as such. It’s a purely notional concept: if the total value of everything “produced” (or “earned” or “spent” – see above) was divvied up equally among every single one of us, this is what we’d get. As I found out, we actually get nothing like that.

In the second quarter of 2015 (April to June), GDP per head in the UK was estimated at £26,680. If that looks on the low side remember it’s the average for every single one of us, including children, retired people, students, people on the dole, lunatics, prisoners, criminals, and anyone else who’s just hanging about — whether they’re doing anything or not. When you think about it, 27 grand per man, woman and child is actually quite a lot.

The ONS says this figure was “broadly comparable” with the level of per capita GDP when the crisis began in September 2008. Total GDP itself passed the pre-crisis level a year ago, but per capita GDP has taken longer to catch up because a good chunk of the growth since the recovery began was simply down to having more people.

Where’s my money, man?

So, per capita GDP should be a good measure of how well-off we are generally. After all, that’s why the champagne corks were popping last week: the recession is really over this time because, on average, our incomes are back to what they were in 2008.

But is per capita GDP really an accurate measure of our incomes? Perhaps because they’re useful, estimates of actual household income are harder to find than headline GDP figures. But the Institute of Fiscal Studies, a sort of unofficial referee of economic arguments in the UK, reckons average household income was £555 per week (or £28,860 a year) in 2012-13. Other estimates come up with very similar figures – the Department of Work and Pensions has £27,500 in 2011-12, for example.

At first glance those look quite close to the per capita GDP figure of around £27k. But hang on, that’s per head! As the average UK household has 2.3 people, we should expect to see household income up above the £60k mark. Quite a lot of our share of the GDP pie seems to have gone missing.

One reason for this is that measures of “household income” exclude direct taxes, like income tax, capital gains tax etc – quite rightly since we never get to spend or save that money as a household. But the money collected (and spent) by the government is included in GDP.

Another even more important reason is that GDP or “national income” has little to do with income as you and I understand it. GDP includes all sorts of earnings that never come near households. The most important of these is company profits, which only count as household income when they are paid out as dividends to people who own shares. Profits retained by firms, or paid out to other firms, banks or institutional investors, obviously don’t count as household income, but they do count as part of GDP.

So although per capita GDP purports to show us our share of national income, it actually has nothing to do with the income we receive as individuals or households. We’re fooled by that “per head” bit. Economists seem to agree that GDP is best way we have to measure the size of the economy and to compare growth rates. But while per capita GDP is useful for stripping out the effects of population changes, as a measure of how well off we are as households and individuals it’s worse than useless.

Ways and means

Actually estimating average household income seems a much better way of gauging how well off we are. But what do we mean by average?

Most figures for things like average incomes or average earnings use a particular kind of average, what statisticians call the “arithmetic mean”. You just add up all the money and divide it by the number of people. But the resulting number is not necessarily typical or representative.

Why? One word: inequality. Imagine a very simple economy of just five people. Let’s say their annual incomes are as follows:

Joe Cleaner:    £12k
Becky Carer:   £15k
Dave Courier: £20k
Alison Nurse:  £28k
Grace Hedgefund-Manager: £150k

In our example economy, total income would be £225,000 and (using the arithmetic mean) average income would be £45,000. But it feels absurd to say £45k is an “average” or representative income, when four out of five people earn far less than that. In this example, “average” incomes tells us nothing at all about anyone’s income. No one earns anything like 45 grand. Grace’s very high earnings compared to the others have distorted the “average”.

Obviously, my model exaggerates things, but the UK economy is a bit like that, with the large bulk of people clustered around low to middle incomes, with a few very high earners at the top (see graph below). If we use the arithmetic mean, this pulls up “average” figures for things like incomes or GDP to an untypical level. It seems oxymoronic to say most people earn less than the average but, when you take the average to be the mean, it’s true.

Another “average” measure, the median, is worked out by ranking all the individual numbers in order and picking the middle one. In this case, Dave’s annual earnings of £20k give us a much better picture of most people’s material circumstances. It’s much more realistic to say that the “average income” in our hypothetical country is £20k, rather than £45k.

You can see this at work in the in the chart above with real figures for the UK. In 2011-12, the median household income was £427 per week, over £100 less than the mean, which is what most people think is meant by “average”.

So median figures for incomes and earnings tell us much more about how “typical” or “representative” families are doing, but get a lot less publicity than GDP. One reason for that might be that they tell a quite different story:

Not only does the typical household get far less than is implied by per capita GDP (or even estimates of “average” incomes), but when you look at real money going to real people, we’re still not back to the previous peak. And something else interesting emerges – household income continued rising in the early years of the recession – households seem to have been cushioned from its worst effects – and only went into reverse in 2010-11 when George Osborne’s started cutting everything that moved. No wonder he’d rather we looked at per capita GDP instead.