Official statistics have become useless
The pandemic has exposed the shortcomings in how we measure everything, from inflation to GDP to jobs, says Helen Thomas
Statistics haven’t quite kept pace with the world they’re meant to measure. Designed for an economy of assembly lines and cheque books, many of today’s headline figures feel increasingly out of step with a digital, service-driven, side-hustling reality. GDP can’t tell you what Tiktok’s doing to productivity and inflation stats seem to live in a parallel universe where no one buys coffee. Yet these numbers are still rolled out like gospel, ready to be cherry-picked, spun and weaponised by anyone with a point to prove. Unfortunately financial markets rely on them as if they were unbiased indicators in the cockpit, showing altitude and cruising speed. But because they always have, doesn’t mean they always will. If investors forget this, the ground may be approaching them much faster than they think.
The pandemic has exposed the shortcomings in our historical measures of data. The US ISM Manufacturing data is widely valued for its long track record of closely tracking the ups and downs of the US business cycle. This didn’t work when the global economy suddenly shut down and reopened in 2020. Within one year the indicator plunged to a 12-year low then reversed to print a 40-year high. This is because it’s a survey, asking people for their sense of whether activity has gone up or down in the past month. When they were frightened out of their minds by the arrival of a disturbing new virus, they reported disaster; when they were relieved after the arrival of vaccines, they reported triumph. As Rudyard Kipling warned, it’s best if you can treat those two impostors just the same.
When people were frightened out of their minds by the arrival of a disturbing new virus, they reported disaster; when they were relieved after the arrival of vaccines, they reported triumph
Meanwhile the technological revolution has undermined both the collection and dissemination of data. The UK hasn’t had a reliable labour force survey since it was withdrawn in October 2023 and its replacement might not be ready until 2027. This is due to a declining response rate to the survey, falling as low as 17.4 per cent. This led to the survey swinging wildly based on just a few responses. The process itself was already struggling prior to the pandemic but sending a letter to households and then following up with a phone or in-person interview just doesn’t work in the modern world. Does any self-respecting millennial even pick up a phone call any more?
It’s the same story in the US. Response rates to almost all the establishment surveys have fallen significantly over the last ten years:
Even when government agencies get hold of the data, they have had difficulties in releasing it. The Bureau of Labor Statistics (BLS) has had a particularly fraught time with this. On 21st August 2024, they failed to deliver the annual revision to the jobs data on time; when some financial firms rang to find out why, they received the information earlier than its eventual public release half an hour later. On 15th May 2024, the BLS admitted that it “inadvertently” posted crucial CPI data early. In February 2024, an economist at the BLS sent an email explaining how the calculation of housing costs had impacted CPI to what they called “super users”, sparking outrage that some financial institutions might be receiving more information than others.
Lies, damned lies and statistics
If people don’t trust the statistics are calculated properly and disseminated fairly, trust withers. Some might have felt it was overdue to replace the person at the top of the BLS. President Trump evidently did, having now fired the Commissioner of Labor Statistics. The UK Office of National Statistics still only has an head after the National Statistician resigned.
New management will hopefully usher in reform, not just of how the data is collected and disseminated, but what its purpose is. The pilots of the economy need to have the latest technology on the flight deck. After all, central bankers were largely flying blind through the pandemic. The last five years can be summed up as two phases: panic and cut rates due to plague, followed by panic and raise rates due to war.
Meanwhile the real economy is being shaped by technological revolution, an aging population and indebted governments. Scarcity of resources is leading to the reimposition of physical borders just as technology leaks intangible ideas across the globe. This is not a world where a basis point on unemployment or inflation makes the marginal difference.
If we want a steady glide path to a safe landing, statistics need to reflect the terrain where we are, not where we came from.
Helen Thomas is CEO and founder of Blonde Money