A Wage-Price Spiral in 2023?

Summary

• A new IMF working paper thoroughly examines past wage-price growth episodes in industrialised economies since the 1960s.
• Their results show accelerating wages rarely turn into wage-price spirals, downplaying the chance wage growth will further intensify current price rises.
• However, real wages usually rise to levels that are ‘consistent with observed inflation and tightening labour markets’, which arguably has not been the case this time around.
• With this year set to see the largest fall in real incomes since the late 1940s, household balance sheets will be tested in 2023.

Introduction

Price inflation tends to be followed by wage inflation, as real incomes catch up with the cost of living. But to what extent does accelerating wage growth feed back into price inflation? This phenomenon, known as a wage-price spiral, could be a concern for central banks, with wages now growing much faster than recent records.

Can we discount the possibility of a wage-price spiral in 2023? A new IMF working paper looks at 79 past episodes of accelerating wage and price inflation in 12 industrialised economies over the last 60 years, revealing a comprehensive overview of the risk. The results are mixed, suggesting no runaway inflation spiral but at the cost of persistently lower living standards.

Rising Nominal Wages

We have spoken in depth this year about how pandemic-induced labour market changes are influencing wage growth.

In February, Larry Summers estimated that around 7.3m people were missing from the US labour force – 1.9mn of which left because of Covid-19. This was leading to excess demand for workers, which he predicted would lead to 6.0% YoY wage growth throughout 2022 and 2023.

Six months later, a policy brief co-authored by Olivier Blanchard explained how the Beveridge Curve – the relationship between unemployment and vacancies – had dramatically shifted outwards since the pandemic. This has increased the natural rate of unemployment, meaning anything below 4.9% will still fuel wage inflation.

An NBER paper reinforced this point only a few weeks after but found that transitory pandemic-induced labour market frictions had already subsided. This, they argued, meant any first-order impacts of the pandemic on wages should have already fully materialised. And, barring any permanent changes in the workforce, which they did not find, the Beveridge Curve and labour market tightness should normalise.

However, even if the structure of the labour force returns to its pre-pandemic shape, there is evidence that price- and wage-setting behaviours may not if inflation persists. This is because firms start to factor higher inflation into their future prices, and workers become more aware of inflation when negotiating wages. In theory, this is the start of a wage-price spiral. But, in practice, does this actually happen?

Data and Methodology

The authors collect quarterly and annual data on inflation, energy prices, nominal wages, unemployment, GDP, and productivity for 38 advanced economies dating back to the 1960s.

To answer how often wage-price spirals have happened in the past and how the economy develops afterwards, they need to define what a wage-price spiral is. They settle on the following:

‘We identify a wage-price spiral as an episode where both price and nominal wage inflation (measured as year-over-year) increase successively for at least three out of four consecutive quarters.’

As such, they identify 79 such episodes. The first was in 1973 – a decade in which most wage-price spirals occurred (Chart 1) – and the last in 2017.

Wage-Price Spirals – The Evidence

The authors then look at the macroeconomic dynamics in the quarters following the first period, where the criteria for a wage-price spiral is met. They find that wage-price spirals are usually followed by a fall in the unemployment rate and do not lead to a further sustained acceleration in wages and prices (Chart 2).

More specifically, wage growth tends to decelerate and stabilise after the six to eight quarters that follow the identified episodes. It does, however, remain at a higher level than at the start of the episode.

There are some exceptions, such as in the US in 1973 – spurred by the first OPEC oil embargo of the 1970s. Prices continued to rise and, because of lacklustre growth in nominal wages, real wages fell.

What drives the wage-price spiral? Usually, inflation and labour market tightness explain all the fluctuations in nominal wages. For example, a 1pp increase in inflation is associated with a 0.6–0.7pp increase in nominal wages in the following period. Similarly, a 1pp drop unemployment gap (labour market tightening) is associated with a 1.1–1.5pp increase in nominal wage growth.

However, the authors find that other factors influence the wage-price spiral. This reinforces the idea that price- and wage-setting behaviours change when inflation accelerates quickly. For example, when negotiating their salaries, workers become more aware of inflation when prices are rising quickly. In particular, the authors find that the normal relationship between inflation, labour market tightness, and wages explain around only 60% of the rise in wages at the start of wage-price spirals.

The key message, though, is that whatever drives the increase in wages at the start of wage-price spirals subsides thereafter. Normality almost always resumes and the spiral uncoils, so it is very unlikely that wages have significant second-, third-, or fourth-order impacts on inflation, and vice versa. Good news!

A Covid-19-Specific Example

Could this time be different? Well, a notable feature of the most recent wage-price rise is one of negative real wage growth accompanied by labour market tightening. So, there are now four conditions:

1. Increasing YoY inflation.
2. Positive nominal wage growth.
3. Negative real wage growth.
4. Flat or falling unemployment.

Looking back, 22 episodes have occurred in which all four conditions have been met in at least three out of four consecutive quarters. But the wage-price acceleration did not persist (Chart 3).

Historically, under Covid-19-like conditions, nominal wages tend to increase for some time but inflation declines. Therefore, real wages usually increase alongside a relatively stable unemployment rate. However, the US provides a counterexample where, in 1979, inflation and unemployment rose, and real wages fell.

A Note on Expectations

Expectations play a large (perhaps exaggerated) role in price- and wage-setting dynamics. Earlier this year, we showed how the path of inflation depended crucially on whether these expectations are assumed to be adaptive or rational.

Under adaptive, the behaviours of firms and workers are determined by what has happened in the past. In the case of price inflation, this can lead to dramatically higher prices (Chart 4, red line).

Rational expectations, however, assume agents set prices based on their estimates of future inflation. If they think they will return to the central bank’s target rate, they will always set tomorrow’s prices lower than today’s (Chart 4, yellow line).

For wage-setting, the distinction between adaptive and rational expectations is less severe but still important (Chart 5). The IMF finds that wages moved 1.5pp higher under adaptive expectations in response to the pandemic-related shock.

The IMF’s models are based on rational expectations. But, even if they were not, there is no suggestion prices or wages have any chance of spiralling out of control. Given the results from the paper, this looks to be the most likely outcome.

Bottom Line

History tells us it is unlikely we will see a sustained wage-price spiral in 2023. In fact, in past contexts similar to ours, strong nominal wage growth has occurred simultaneously with receding inflation. So, even if wages accelerate in 2023, do not be too concerned about second-order impacts on inflation.

One anomaly of the pandemic recovery relative to similar episodes, however, is real wages. Typically, nominal wage growth exceeds inflation, driving real wages higher. We are still some way from positive real wage growth if we look at headline consumer price inflation (CPI). So this either means higher nominal wage growth for some time or a quick reduction in inflation. If it is the former, and higher nominal wage growth becomes entrenched, could there be a wage-price spiral in 2024?