Fax machines, record stores, analogue cameras, Liverpool Football Club winning the league! All of these things were common in previous decades but have subsequently faded away. Ditto the relationship between unemployment and wage inflation, known as the Phillips curve? With US unemployment at record lows and wages benign, has this gone the way of electronic whistling, HMV, Kodak, and Kenny Dalglish?
I’m not convinced the Phillips curve has real-ly disappeared, though. Instead, I think people are underestimating the role of inflation expectations – a critique economists Edmund Phelps and Milton Friedman both stressed 50 years ago!
Phillips’ 1958 paper highlighted that UK wage growth was higher when unemployment was low, and vice versa, during 1861-1957. But within a decade, Phelps and Friedman independently argued that the relationship would break down because it was ‘real’ rather than ‘nominal’ wages that workers cared about, i.e. wages adjusted for inflation.
If governments held unemployment down for too long, boosting inflation, the whole curve would subsequently shift up – that is to say, wage growth would be higher for a given level of unemployment. Workers would anticipate higher inflation and demand compensation for it as well as an additional premium reflecting how tight the labour market was.
This is what happened in the 1970s after oil shocks de-anchored inflation expectations. Inflation remained stubbornly high even as unemployment shot up as wage-price spirals developed. The negative drag on inflation from economic slack was offset by the rise in inflation expectations.
I’ve previously argued the opposite happened in 2014: a sharp fall in oil prices reduced inflation expectations. While a temporary fall in inflation expectations was inevitable as oil prices plummeted from $100 to $30, I was surprised that inflation expectations failed to recover even as actual inflation and energy prices bounced back. Instead, consumers’ inflation expectations remain at a 53-year low.
I think this resolves one of today’s key macro puzzles. If we look at consumers’ perceived ‘real wages’ (wages less inflation expectations), they are growing as quickly as in previous economic booms. But low inflation expectations are self-reinforcing by depressing nominal variables.
So I don’t buy into the argument that the natural rate of unemployment keeps magically changing. Instead, I think that lower inflation expectations have shifted the entire curve down – the opposite of the 1970s.
If inflation expectations are lower and more firmly anchored than in the past, then it should mean bigger economic shocks are needed to lift inflation back up. A given fall in unemployment or rise in the oil price should have less of an impact on wages today than in the 1970s when inflation expectations were unanchored. This helps explain why the Federal Reserve is wincing at the first sign of a downturn.
In future blogs I will show how inflation expectations are driven by experience, explore cross-country analysis, and demonstrate how profit margins can be more volatile in a world of anchored inflation expectations.