How the UK Employment Market Has Baffled Economists

Ryan Bourne

Each month, as Government ministers celebrate invariably strong UK employment numbers, some members of the public let rip on Twitter.

These individuals cannot accept the unemployment rate is at a 45-year low of 3.8pc. They pour opprobrium on those citing the current 76.1pc employment rate as the highest since modern records began.

For the truthers, the official statistics are misleading or wrong. By counting someone as “employed” if they work one hour per week or more, official stats are said to be concealing the true parlous state of the labour market from the public.

Richard Clegg, an outgoing stats bod of the Office for National Statistics, has explained in vain that only a tiny fraction of those employed work such limited hours. The trend is falling too. Those working fewer than six hours per week fell from 1.5pc of those employed in May 2010 to 1.4pc today; the share working six to 15 hours dropped from 7pc to 6.2pc.

Yet the truthers don’t care. Their views are impervious to evidence and indifferent to hearing that “nothing has changed” in terms of measurement.

Our priors, it seems, are stubborn. That’s particularly bad news for economists, since the labour market has dumbfounded most of them since the crash. Cast your minds back to the coalition government’s early days, when unemployment still stood at a lofty 7.8pc. It’s easy to forget just how wrong most economists’ pronouncements about job prospects have been.

Plenty (including me) believed the financial crisis was a profound structural shock to the economy. Though unemployment would no doubt improve, we thought it was unlikely to fall back to its pre-crash level for a long time, not least because of skills mismatches between redundant workers and new demands. The banking crash had shrunk the capacity of the economy, we thought, and, with it, prospects for employment.

In economic terms, we considered some new unemployment “structural” rather than “cyclical”. It would likely remain stubbornly high as the economy recovered.

The Bank of England agreed. When it issued its forward guidance for interest rates in 2013, it estimated the “natural rate” of unemployment at around 6.5pc to 7pc. According to its model then, the low unemployment rate we enjoy today was not possible, at least not sustainably. It could only be achieved temporarily, through suffering large spikes in inflation.

This view, self-evidently, was wrong. The proportion of unemployed workers out of work for more than a year is indeed higher now than pre-crash – 26.6pc versus just under 25pc. But the unemployment fall means the share of the population who are long-term unemployed is actually down. An impaired financial sector maybe did worsen employment prospects for a few years, and certainly did affect productivity. But the labour market rebound implies the spike in unemployment was largely cyclical and not structural. We gloom merchants were incorrect.

Yet if we were mistaken, so were those Keynesian economists who warned that public spending cuts would produce a “lost generation” of workers. One of the precise reasons “pro-stimulus” economists wanted higher government spending was because they thought weak demand would lead to unemployed workers’ skills wasting away.

This would create what economists describe as a “hysteresis” effect — skills atrophy making it difficult for the unemployed to find willing employers. They foresaw an Eighties repeat, where many workers, including laid off public sector employees, might struggle to find productive employment again.

Former Monetary Policy Committee member Danny Blanchflower predicted unemployment might rise dramatically following spending cuts. Contrary to his musings, between 2010 and 2015 private sector employment growth exceeded public sector job losses by a factor of seven.

Keynesian economists swiftly pivoted to a speculative thesis with no historical grounding that cuts resulted in “productivity hysteresis” — that the weak demand resulting from public spending restraint had worsened productivity, resulting in the slow wage growth we’ve seen. As yet nobody has explained how (excluding investment spending) higher Government outlays for benefit payments or most Government departments would have made the economy so substantially more productive.

Whereas most people thought getting unemployment down would be the economic issue dominating political discourse, instead discussion now centres on how low unemployment can fall and how to improve productivity to deliver real wage growth.

In a new book, Not Working, Danny Blanchflower argues that there is still substantial labour market slack today. Many part-time and zero hour contract workers want more hours, he claims, and falling home ownership has improved the propensity for workers to move. His conclusion is that policymakers should put the “pedal to the metal” — using government spending and monetary policy to “run the economy hot” until we see the white of inflation’s eyes.

Yet wages now ticking up as unemployment falls suggests we are reasonably close to “full employment” already. What’s more, the longer we are into the recovery, the more the poor productivity performance looks like a supply-side, rather than demand-side, issue that “more stimulus” won’t solve.

Though Blanchflower and co were right that unemployment after the crash was cyclical, the past decade surely suggests that a flexible labour market is reasonably self-correcting. After wages and prices adjusted, unemployment fell significantly, even after the huge shock of the Great Recession. That suggests deficit spending is far less important to employment prospects than stimulus proponents envisaged.

Strong productivity growth, not pump priming, is what we need to deliver rising wages and better living standards. Regrettably, economists of all stripes can be as entrenched in their views as the truthers on Twitter, even as the evidence changes.

Ryan Bourne is the R Evan Scharf Chair for the Public Understanding of Economics at the Cato Institute