Rishi Sunak is a thoroughly modern UK chancellor. He devised this week’s summer statement from Number 11, sipping from a thermostatically-controlled tea flask while on official video calls. But the most important innovation was that his Covid-19 stimulus policies were calibrated using near real-time data on spending and footfall in shops, pubs and restaurants.
Evidence from these show that reopening non-essential shops in mid-June, and the hospitality sector last weekend, generated a burst of spending, but that people also remained cautious. With spending levels depressed, Mr Sunak had the comfort he needed to boost demand by reducing value added tax rate for the tourism sector.
But the chancellor’s actions also raise a question: will these faster, innovative indicators now always come first for economists, consigning official statistics to economic history?
The Bank of England’s chief economist is convinced they will. Andrew Haldane said last week a new suite of faster indicators, “has significantly shifted the technological frontier when monitoring the economy”. Using his selection of faster data, he said the UK downturn had been far shallower than the BoE predicted in May, and the recovery quicker.
Before we fall headlong for the hype, we should be careful not to be too easily seduced by these innovations in economic monitoring.
Faster indicators have given useful signals during the pandemic largely because the economic swings have been so large. When official figures show UK gross domestic product fell 25 per cent between February and April, all real time data will show an economy in freefall. A better test will come if faster data can accurately measure the magnitude of normal economic swings, which are roughly 100 times smaller.
A good corrective is to revisit the suite of real time indicators that existed around the 2016 EU referendum. With hindsight we know that immediate economic conditions were unaffected by the vote. Yet economic barometers the Financial Times and others published in the weeks afterwards pointed to considerable falls in output. Whether that was because humans picked the wrong indicators, or because the faster indicators were useless, they gave a false steer.
Even in this crisis, fast indicators have been far from perfect. Mr Haldane says they now refute the BoE’s May forecast of a 27 per cent drop in GDP between the last quarter of 2019 and the second quarter of 2020. But that forecast was itself based on May’s faster indicators, which we now know gave an incorrect view of the economy.
These caveats do not mean faster data should be ignored, but that we need to know its weaknesses while marvelling at its speed.
One problem is access and ownership. So far, the private sector has been innovative and open with its information, whether it be Google’s mobility data, OpenTable’s aggregate restaurant booking data or spending information from companies such as Fable Data. But there is no guarantee this will continue.
In contrast, the public sector has been poor in providing access to data that taxpayers fund. Whether it was the Highways Agency refusing to release data from its motorway cameras, or the Department for Work and Pensions having its knuckles wrapped by the statistical regulator for its release of universal credit information, officials have preferred to keep relevant economic indicators to themselves.
At the same time as Mr Haldane lauds daily spending data the BoE derives from its clearing house automated payment system, the central bank assiduously denies public access. These are major weaknesses. For now, when economic swings are large, faster indicators are the height of fashion. But in normal times, data openness and transparency matter more. This fashion will be fleeting.
Get alerts on Coronavirus economic impact when a new story is published