How wary should we be about higher wage growth?
With everyone trying to predict the path for inflation next year, there’s a lot of focus on the outlook for wages – a major determinant of inflation. While a wage-price spiral looks unlikely to develop, it’s still difficult to predict how quickly wage inflation will decline.
Wages as a gauge of underlying inflation momentum
The performance of the economy and the markets next year will to a considerable extent be determined by the trajectory of inflation. In particular, a faster-than-expected decline in inflation momentum will translate into a positive shock for household real income growth and business profit margins. All else being equal this is likely to mean a more positive growth outlook. In addition, it will enable central banks to bring their policy rates back towards neutral sooner and faster than expected, a development that is likely to reinforce the positive growth impulse. Of course, if high inflation proves more persistent than currently expected, all these effects will go into reverse: there will be an additional drag on household real income growth and business profit margins, the effect of which will be enhanced by central banks pushing policy rates even higher for much longer.
It is generally recognized that inflation momentum over the past two years has been strongly influenced by a series of very unusual cross-currents. These mostly came in the form of pandemic-induced shutdowns of certain sectors and subsequent reopenings, policy support that enabled households to build up a large buffer of excess savings and the Ukraine conflict, which triggered a large increase in commodity prices. The one thing all these cross-currents have in common is that they are in principle one-off effects. For instance, excess savings can (and certainly did) provide additional demand support, but only for as long as the pile is not depleted. Commodity prices cannot continue to rise forever, if only because if they did so it would destruct so much demand that the potential for further rises would vanish completely.
The big question for the inflation outlook, then, is where it will settle once these cross-currents have died down. The standard answer from a macroeconomic perspective is that it will depend on whether or not inflation expectations remain anchored. The deeper meaning behind this is that the central bank issues a credible threat to workers and businesses that it will punish them with higher unemployment and lower demand if they allow a power struggle for their respective shares of the economic pie to heat up to such an extent that a wage-price spiral develops.
Unsurprisingly in light of this, central banks and investors are focusing on wage growth as a signal of where inflation will settle down once the cross-currents have played themselves out. As a first approximation this makes a lot of sense. Firms set output prices as a mark-up over (marginal) costs. This means that if wage growth accelerates while productivity growth remains fixed, unit labour costs will rise, which firms will pass on through higher output prices.
There are at least two caveats to be made here. First, from a cyclical perspective, wage growth is also directly affected by the cross-currents we refer to above. This means that part of the acceleration in wage growth seen so far could reverse of its own accord. Second, from a longer-term perspective the relationship between real wage growth and productivity growth may be more of a two-way street than many pundits believe.
Shocks outside the labour market are only likely to affect wage growth temporarily
The sharp increase in headline inflation against a backdrop of unemployment rates that are close to or in some cases even below their pre-Covid levels has sparked renewed research interest in the risk of wage-price spirals. After all, nominal wages tend to remain fixed for quite some time (a year, sometimes more), whereas most output prices can be more frequently adjusted. Workers have therefore suffered an unexpected drop in real wages that they will seek to make good in future wage negotiations.
The fact that labour markets seem quite tight will support their bargaining power in this respect. If workers do indeed manage to recoup a significant part of their initial real wage loss, businesses might respond by raising their prices to protect their margins. This could in turn lead to a further acceleration of wage growth, and so on. This was more or less the story of the Great Inflation of the 1970s, when workers and businesses fought over how the real income loss induced by higher oil prices should be divided.
Fortunately, recent research by the IMF (summarized here: Wage-Price Spiral Risks Appear Contained Despite High Inflation (imf.org)) suggests that the risk of a repeat of the 1970s is limited if three conditions are met. First, the shock should come from outside the labour market. Second, inflation expectations must be well anchored. Third, central banks must take decisive action to signal that they will do whatever it takes to keep inflation expectations anchored.
The second and third conditions make perfect sense: anchored inflation expectations simply mean that workers and businesses see the central bank’s threat to punish them if they behave badly as credible. In this respect a determined pace of policy tightening will enhance the credibility of this threat by cooling down the economy and the labour market somewhat and thus damping the ability to raise wages and prices.
The first condition is a very interesting one and in fact also makes a lot of sense. An example of a shock inside the labour market is the introduction of automatic price compensation in wage contracts – a clause in the contract that stipulates the wage rate be automatically increased in response to higher-than-expected inflation – which would make a wage-price spiral much more likely. Another example is a substantial increase in workers’ bargaining power due to a change in labour market regulation or the rise of militant unions. In all these cases much stricter monetary policy would have to be maintained for much longer to prevent a repeat of the Great Inflation.
We have seen two examples of shocks outside the labour market over the past two years. The most recent was the sharp rise in commodity prices, which initially caused real wages to fall sharply. As new wage contracts are signed, workers partially recoup these losses, which shows up as an increase in measured wage growth. Still, provided the other two conditions are met, historical evidence suggest that the amount they recoup is one-off in nature and does not lead to a wage-price spiral. The fact that many wage agreements contain one-off lump-sum payments to amend for a loss of purchasing power lends credence to this notion.
The second shock is more relevant for the US than for Europe because the European policy response to lockdowns revolved around keeping workers in their jobs through furlough schemes. By contrast, the US policy response focused on providing direct income support, because of which the US saw a much bigger rise in unemployment. As such the Covid crisis caused a massive shake-up in the sector make-up of the US labour market. In the early phase of the pandemic employment fell substantially in high-contact services such as leisure and hospitality. A share of the workers who were laid off from these sectors found employment in the retail and transportation sectors, which benefitted from the surge in demand for goods. In order to entice them back into high-contact services, employers had to offer substantially higher wages. Another element playing into this is that fears of contracting Covid made workers less willing to apply for jobs in high-contact services, and elevated unemployment benefits enabled them to refrain from doing so. As a result, employment in these sectors is still below its pre-covid peak.
This massive shake-up of the labour market, which involved large flows into and out of the labour market as well as large flows between sectors, requires widespread changes in relative wage rates between sectors. Because nominal wages almost never decline in absolute terms, this means that the changes in relative wage rates mostly have to result from higher wage growth in expanding sectors rather than falling wages in contracting ones. This pushed overall wage growth higher. Nevertheless, as the labour market finds a new equilibrium, this source of wage acceleration should abate, and there is some tentative evidence that this may be happening. For instance, the proportion of sectors experiencing wage growth in excess of 4.5% in the Employment Cost Index fell from a peak of 80% to 35% in the latest data.
Could higher wage growth lead to higher productivity growth?
In the long run, once all shocks have died out, economic theory predicts that real wage growth will be determined by the growth rate of labour productivity. The central idea here is that in a competitive economy, each input is rewarded in accordance with its (marginal) contribution to output. An important real-world complication is that the economy is often characterized by various concentrations of power, such as monopolies in goods markets, because of which the balance of power between different actors is often an important determinant of how various production inputs are rewarded.
In this respect, workers’ bargaining power has fallen substantially over the past 40 years, as evidenced by the decline in the labour share of GDP in most developed economies. It is reasonable to assume that the trend could partially reverse going forward due to population ageing and deglobalization. This could cause somewhat higher average inflation in the future. However, the extent to which this happens will depend crucially on how such a redistribution of income in favour of workers affects productivity growth.
At the micro level this redistribution means higher real wages and correspondingly lower profit margins. It would be tempting to conclude that lower profit margins would reduce private investment growth, but this would be jumping to conclusions. Private investment growth is driven less by profit margins than by expectations of future profitability. Profit margins are one part of the equation, and aggregate demand growth is another. Because the marginal propensity to save out of wage income is much lower than the propensity to save out of profit income, the rise in real wages may trigger a persistent increase in underlying demand growth. On balance, this may cause an increase in profit expectations and thus a higher rate of private investment growth.
There are various other reasons why higher real wage growth may go hand in hand with higher productivity growth. The theory of efficiency wages holds that paying workers a higher wage than the competitive equilibrium price can raise their productivity by increasing the cost of shirking (which, if found out, leads to dismissal) and making them more attached to the firm. What’s more, the presence of a large pool of low-wage workers who can easily be hired and fired may prevent firms from investing sufficiently in labour-saving technology. After all, investing in additional capital requires a large upfront cost, and the return on such investments is always uncertain. It is therefore tempting to meet increased demand by hiring more low-wage workers.
What all of these reasons have in common is that they are more likely to hold when the labour share of GDP and real wages have a low starting point. Conversely, they are unlikely to hold when the labour share and real wages are high. In that case, raising them further is almost certain to squeeze profits to such an extent that investment growth will decline substantially.
This dependence on the initial state of the economy does not hold so much for the final reason why higher wage growth can go hand in hand with higher productivity growth. Ongoing structural shifts in the economy such as deglobalization and the energy transition are likely to require an ongoing high level of sectoral churn in the labour markets. The movement of labour towards more productive sectors is made much easier if policymakers allow the required changes in relative wage levels to proceed uninhibited, even if this comes at the expense of a somewhat higher economy-wide rate of wage growth.
Conclusion
Central banks and investors are rightfully focusing on wage growth as a gauge of the extent to which inflation is likely to decline next year. The combination of a large, prolonged inflation spike and very tight labour markets does indeed involve a non-negligible risk that attempts by workers to recoup purchasing power that they have lost will spill over into a wage-price spiral. Fortunately, empirical evidence suggest that the risk of this happening is likely to remain low as long as the shocks that triggered this process have their origin outside the labour market, inflation expectations are anchored and central banks take decisive action to ensure expectations remain anchored going forward.
It seems that all three of these boxes are ticked for most developed economies, but unfortunately that does not help us much in predicting how fast wage and price inflation will decline. On the one hand, labour markets could remain more overheated for longer, in which case central banks will have to become more restrictive before their actions exert enough downwards pressure on wage growth. On the other, we should also recognize that part of the acceleration in wage growth was a direct result of the various supply shocks that have hit the economy. These supply shocks abating could also cause wage growth to fall faster than expected.
