DAN O’BRIEN

Gross domestic product: fiddling with the most powerful metric in history

GROSS domestic product has been described as the most powerful metric in history. Since its invention (in depression-era America, where it was created with a view to better understanding what was happening in a collapsing economy), it has gone from strength to strength over decades.

But more recently, misgivings from various quarters are more frequently heard. One of the main criticisms of GDP is that output and production measures may not give a good guide to what is happening to people’s incomes.

Simon Kuznets, the economist who invented the measure, saw this from the beginning. He initially focused on national income in the 1930s — but in the 1940s the focus shifted to national production in order to assess how the economy was contributing to the war effort.

Despite Kuznets’ urging, the focus never shifted back to income after the war. As time passed, it came to be accepted as the best measure of economic performance despite its limitation — perhaps a little like the qwerty keyboard: not perfect but good enough, and too messy to change.

But there have always been critics, and the chorus of criticism has been growing internationally. Our ‘leprechaun economics’ episode last summer, when Irish GDP data made headlines in the financial press around the world after revisions led to growth in 2015 jumping to an unbelievable 26pc, has merely added to the debate.

A new study by Brian Nolan, Max Roser and Stefan Thewissen* has highlighted just how different per capita GDP can be from other more incomefocused measures. One of the authors, as it happens, is the Oxford-based Brian Nolan — probably Ireland’s leading authority on income and wealth equality issues.

Their study compares the growth of real GDP per capita and median household income in OECD countries. The ‘real’ means that they are adjusted for inflation — a pay rise is no good if prices rise as well. Household income here refers to earnings after taxes and transfers. And the ‘median’ represents households in the middle of the income distribution.

The two measures come from different sources: GDP from national accounts and household income from separate surveys. Unfortunately, longrun data from surveys are patchy, but the authors were able to get hold of data from the 1970/1980s to 2010 for most countries (alas, the time period for each often differs).

Their results show a widespread divergence between growth of GDP per head and the income of a typical household as measured in surveys.

In 23 out of 27 countries studied, GDP per capita grew faster than median household income (the exceptions were Czech Republic, Estonia, Norway and Switzerland).

On average, the annual growth rate for GDP was 2.17pc, but 1.6pc for median incomes. In other words, the latter measure shows that people have been getting richer by one quarter less than by what the GDP measure suggests.

That is a sizeable amount when you consider these are annual rates spread over three decades, as the Irish case shows. Real median household incomes rose by 106pc between 1987 and 2010, while per capita GDP grew by 150pc, as depicted in the first chart.

It should be noted, however, that the scale of difference varied a lot by country.

The big outlier was the United States. Despite decent economic growth since 1979, there has been a paltry increase in median wages. And this is a point that has often been raised in American public debate, most especially during the ongoing presidential election campaign. (Jason Furman, one of Barack Obama’s economic advisors, has stated that when new employment data comes out, the figures the US president is most interested in are those on wage growth.)

And there was in fact some positive American wage data released this month, showing that median American household incomes grew by 5pc in 2015 — the largest single increase since records began. But it would take years of strong wage growth to reverse the falling share of GDP going to wages.

There are numerous theories to explain why labour’s share of GDP has fallen or stagnated in most developed countries in recent decades, including falling union membership, globalisation and technology.

The authors of the study look at several factors for the divergence. One of the most cited reasons is inequality, which should rise if median incomes grow more slowly than average national income. This is indeed a factor in a number of countries, notably Canada, the US and the UK. Yet, the overall contribution of inequality was found to be modest on average.

Two of the most common factors could perhaps be described as statistical quirks. The first relates to how prices are used to adjust for inflation. Household income is usually deflated by the Consumer Price Index, which measures the price of goods and services consumed by ordinary households. Whereas GDP is adjusted by a GDP deflator, a measure that largely looks at prices for producers. Although it differs by country, on average the CPI grew faster than the GDP deflator. In other words, the prices for consumers have increased more than the prices for producers.

A second reason is a largely neglected one. In most countries, the size of a typical household has fallen. There are economies of scale if people live together, which are lessened when the numbers living in a given home are lower.

The findings raise as many questions as answers. Moreover, there are significant variations among countries in terms of the scale of divergence and the factors behind it. No grand theory emerges from the evidence.

Ireland’s growth rate from 1987 to 2010 for both measures was well above the OECD average, but the discrepancy of 1pc per annum was also among the highest. In the Irish case, the authors found that the single largest factor explaining the discrepancy was the decrease in household size.

The three authors call for a better integration between the national accounts and household surveys. An EUOECD working group is already looking at this, reflecting the international consensus that we need more than GDP to measure economic health and how people’s levels of prosperity are changing.