The danger of building strong narratives on weak data

Receive free economic statistics updates

Do you remember the double-dip recession of 2012? This is a collectible. Two years after its slow climb from the nadir of the 2008-09 recession, the UK economy shrank by 0.2% in the last quarter of 2011 and the same situation occurred in the first three months of 2012, the Office for National Statistics announced. .but then it changed its mind.

By mid-2013, the ONS looked at the data again and concluded that output had been flat in the first quarter of 2012: the recession had receded.Revisions continue and as of today the Office for National Statistics’ best estimate of the UK economy increased by 0.8% Q1 2012 — an overall improvement of a full percentage point.

If this sounds like I’m being critical, I shouldn’t be. Measuring GDP is extremely complex. Around the world, national statistics offices are struggling to get the numbers right the first time.

Some people struggle more than others. When Ireland first reported GDP growth expectations for the first quarter of 2015, the forecast was 1.4%. A year later, with some pretty unique twists as it’s home to many of America’s big tech and pharmaceutical companies, pediclethe s Increased to a jaw-dropping 21.4%.

On average, five years after the Irish GDP quarterly growth forecast was first published, the latest revision to the figure is a full two percentage points below the original value. The figure for the UK is almost ten times smaller at 0.25 percentage points, making the ONS’s preliminary estimate one of the most accurate among developed countries, narrowly ahead of the US (0.26) and well ahead In countries such as Japan (0.46) and Norway (0.56).

But it’s not just the size of the revision that matters, but also its direction. Of the 24 developed countries that continue to report quarterly GDP revisions to the OECD, the UK’s preliminary estimate is the most pessimistic. UK quarterly growth figures are typically 0.15 percentage points higher than initially expected.Germans rose by an average of 0.07, the French by an average of 0.04, while the ever-optimistic Americans usually ended up revising their estimates down 0.11 percentage points.

The chart shows that the UK's quarterly GDP estimates are the most accurate and pessimistic among the developed world, with upward revisions tending to be larger over time than in other countries.

In other words, the next time you hear a set of quarterly growth figures, it’s not unreasonable to mentally add 0.15 to the UK and subtract 0.11 to the US.

This may sound like a nerdy detail, but it’s important because people graft strong narratives onto these very fragile data. Britain is the only G7 economy that has yet to rebound to pre-pandemic levels until it does. Ireland is boomingObviously, in addition to its actual individual per capita consumption (a better measure of living standards than GDP) Steady decline It fell from just above the Western European average in 2007 to 10% below last year.

Chart shows Ireland's GDP per capita has surpassed that of the rest of Western Europe, but its average living standards remain low

This phenomenon is not unique to economic data. Two years ago, progressives critical of the government’s response to the pandemic began calling Britain “plague island” because it reported the highest COVID-19 death rate in the developed world. But in hindsight, we know the UK is better than most countries at counting deaths.

When the dust settled, the UK was in the middle of the pack on pandemic death rates. It’s not just that the UK underperformed two years ago and then improved: the UK has been performing better than the initial data suggested.

The chart shows that early statistics suggested the UK had one of the worst records for loss of life and economic performance during the coronavirus pandemic, but using revised GDP data and more robust mortality data we now know the UK had one of the worst records for both loss of life and economic performance during the pandemic. All indicators are closer to the average

A rare example of cautious analysts refusing to let flimsy data shape the narrative occurred in the United States last year, when the National Bureau of Economic Research’s Business Cycle Dating Committee simply decided that despite two consecutive quarters of estimated gross domestic product contraction , but more specific indicators such as employment data show that this is only a short-lived phenomenon. There was no recession.

We can all learn from what the United States is doing. GDP data remains the go-to statistic for tracking progress within and between countries, but when these notoriously fuzzy numbers point in one direction and more reliable or tangible statistics point in another, perhaps we should be careful before jumping to conclusions. stop.

john.burn-murdoch@ft.com, @jburnmurdoch

Svlook