UK Markets closed

In the age of hyper-partisan politics and the gig economy, is economic data broken?

Oscar Williams-Grut
Senior City Correspondent, Yahoo Finance UK
Changing working habits: A biker working for the Food delivery service Deliveroo cycles off to deliver an order on July 3, 2018 in Saint-Ouen, outside Paris. Photo: GERARD JULIEN/AFP/Getty Images

Economic data is becoming increasingly dislocated from reality, according to a leading economist.

Paul Donovan, the global chief economist of UBS (UBS) wealth management, told journalists last week that both official and unofficial economic data — ranging from unemployment numbers to surveys of business sentiment — were becoming less and less reliable as gauges of the underlying economy.

“Structural changes are simply changing what data means nowadays,” Donovan said.

If correct, Donovan’s analysis means stock market movements may be being driven by faulty information. It also means that investment funds, which people depend on for their pensions and for income, are at risk of being driven by feelings rather than fundamentals.

The Fourth Industrial Revolution

Donovan’s argument rests on two ideas. Firstly, that changes in the way the economy works means traditional data measurements are no longer fit for purpose. Secondly, that survey data is becoming increasingly unreliable.

On the first point, Donovan says that the so-called “Fourth Industrial Revolution” — the growth of robotics, AI, and technology-driven businesses — is changing the way the economy works. Take, for example, new “gig economy” platforms like Uber and Deliveroo that are driving an increase in self-employment. Two thirds of the most recent rise in employment in the UK was driven by self-employment.

“When you get a rapid growth in self employment this creates an awful lot of confusion in terms of economic data because, for example, a lot of self employed people will pay themselves a constant wage, so wage growth is zero,” Donovan said. “But then they will pay themselves a bonus or dividend when they get employment as contract workers or whatever.

“You end up then with wage data perhaps giving us less of an indication of the pressures that exist in the labour market.”

‘You’re going back to 18th century economics’

These changes also have implications for unemployment data.

“The unemployment number does what it says it intends to do,” Donovan said. “But what we’ve got to recognise now is we’re not in the economy of the 1950s where you clock on at 9 and you clock off at 5. That’s just not the way the world is working anywhere anymore.

“In many ways you’re going back to 18th century economics where people hold multiple jobs, they’re self-employed in a variety of ways and they do lots of different things. Some of the jobs they’re paid for, some of them are voluntary — it becomes a lot more complicated.”

Changes in technology usage and working habits are also messing up official data, Donovan said. The fact that people are increasingly using their own phones and laptops for work means businesses are having to invest less. Flexible working also allows them to save on office costs.

“You look at the headline number and think: ‘poor capital spending, this is terrible,’ and you don’t think ‘actually increased efficiency in the economy, this is great,'” Donovan said. “These confusions are coming in.”

‘Nobody fills in surveys anymore’

Donovan’s second point is that “surveys are becoming less and less reliable as a general rule because basically nobody fills in surveys anymore.”

“If you fill in a survey almost by definition you’re weird. That means you’re getting less and less representation.”

The response rate for economic surveys in the US has fallen from around 80% 20 years ago to around 50% today, Donovan said.

“If you look across pretty much all survey evidence, including business sentiment surveys and so forth, the correlation of these surveys with reality has been declining over the last 10 years or so.”

Don’t worry: Paul Donovan, Global Economist at UBS wealth management. Photo: Matthew Lloyd/Getty Images for ReSource 2012

Samuel Tombs, the chief UK economist for Pantheon Macroeconomics, alluded to this in his recent note on January’s IHS Markit PMI numbers for the service sector, which are a popular measure of future growth.

“Since 2004, the rate of GDP growth implied by the PMIs has been wide of the mark by an average of 0.3 percentage points,” Tombs wrote. “They were far too downbeat immediately after the referendum, suggesting that they tend to overstate the impact of political uncertainty on the economy. They also pointed to downturns in activity in 1998, 2001 and 2003 that failed to materialise.”

Donovan added that people “tend to introduce bias in their responses” to surveys. He highlighted a change in Republican consumer confidence after the 2016 election of Donald Trump as president, but not in Democrat voter confidence.

“Was there some kind of miraculous improvement in the economy if you’re a Republican in 2016? There was not,” he said. “What there was of course was a sudden change in politics and then that change is reflected in the data. That gives a misleading picture as far as the markets are concerned.”

‘A lot of confusion’

All of this means the “ability to accurately capture what’s going on in the economy is becoming more and more constrained,” Donovan said.

“What we’ve got here is a lot of confusion and this is translating in practical terms into more and more revisions of data.”

He highlighted GDP growth in the first quarter of 2015, which was originally pegged at 0.2%, but eventually came in at 3.2% after multiple revisions both higher and lower.

“The problem with financial markets is markets tend to react to the initial release, not the subsequent revisions,” Donovan said. “We continue to have this problem with perception and reality.”

Simon French, the chief economist for Panmure Gordon, told Yahoo Finance UK the problem of perception and reality was a constant issue in economics.

“We are never going to measure a complex, interconnected, and constantly evolving economy correctly in real-time and policymakers/investors have always had to deal with imperfect data,” French said.

He said survey data can often be “too erratic for over interpretation,” but disagreed with Donovan’s overall diagnosis, saying, “I am unclear that it has got any worse, and at the margin I think it has got better.”

Donovan recommended investors focus on five-year correlations with data, which he said would likely give number-crunchers enough time to properly revise the data until it’s relatively accurate.

He said that despite the current “noise” about the possibility of recession this year, the global economy “is doing just fine.”

————

Oscar Williams-Grut covers banking, fintech, and finance for Yahoo Finance UK. Follow him on Twitter at @OscarWGrut.

Read more:

Shares in trading platforms tank as new European rules bite into profits

Just Eat vs. Cat Rock: Inside the investor battle that’s made takeaway business a deal target

Crypto customers lost £110m after a founder died — it won’t be the last investor wipeout

UBS chief economist: ‘Equities are not actually very important for the global economy’

Swedish startup lands €56m from big banks to take advantage of ‘open banking’