Speech by Luis de Guindos, Vice-President of the ECB, at the ECB conference “Inflation in a changing economic environment”
Frankfurt am Main, 24 September 2019
You have now been discussing issues related to inflation for two days, covering a wide variety of topics. Conferences of this kind are very helpful to us as central bankers, as they enable us to advance our thinking about policy-relevant questions, particularly if they are related to our core mandate. In concluding this conference, let me say a few words about economic slack in the euro area and the wage-inflation pass-through. I think these are interesting topics at the moment, because, as you know, euro area inflation has been persistently low for several years, and more recently we have observed some decoupling of wage inflation from price inflation.
Euro area consumer price inflation currently stands at 1% (August 2019, year-on-year HICP inflation) and is thus considerably below our inflation aim. Unfortunately, this low inflation has been a persistent phenomenon. Since November 2011, when euro area inflation peaked at 3% (year-on-year HICP inflation), the average rate has been less than 1.2% per annum.
Long-term inflation expectations are also low. Market-based indicators are stagnating at historically low levels. The five-year five-year inflation-linked swap rate has been hovering around 1.25% in recent weeks. Survey-based measures have also fallen recently, to 1.7%. While these numbers are higher than those for market-based indicators, we shouldn’t take comfort from them. International evidence shows that survey expectations can react very slowly to persistently low inflation (as was the case in Japan, for example).
Wage inflation, on the other hand, has behaved quite differently to consumer price inflation. Since 2017, wage inflation, as measured by compensation per employee, has overtaken consumer price inflation, peaking at 2.4% in the third quarter of 2018.
These developments raise important questions. Is the euro area economy hitting its capacity constraints? Or is there still economic slack that is holding headline inflation down? How will wage inflation pass through to price inflation?
How to measure economic slack
There are signs of a tight labour market. In addition to the wage dynamics I just mentioned, we have observed the unemployment rate moving close to its pre-crisis low, with participation rates being higher than ever.
At the same time, assessing the tightness in labour markets is not straightforward, for several reasons. First, these aggregate statistics hide substantial cross-country heterogeneity. Second, and more fundamentally, the natural rate of unemployment is unobservable.
Furthermore, tightness in the labour market provides only a partial picture of slack in the economy, as slack can vary across different parts of the economy. It is therefore risky to infer it from just a few economic variables. And for each measure of slack the same issue applies as for slack in the labour market: it cannot be observed directly, because the potential (or “trend”) levels of economic variables are unobservable and time-varying. Accordingly, there are various estimates of the aggregate slack and these estimates differ from one another. Some of them are also subject to substantial ex post revisions. For all of these reasons, we need to be cautious when assessing economic slack and look at many economic indicators to achieve a robust assessment.
The wage-inflation pass-through
Let us turn to the wage-inflation pass-through. Given the recent growth of wages, it is crucial to understand how this will pass through to inflation. But coming to a robust assessment is not straightforward. Both economic theory and the empirical evidence tell us that this pass-through depends on many different factors.
One consideration is that the structure of the economy is not constant over time. Over the past few decades, we have been witnessing gradual changes in the economic environment, many of which can be traced back to the ongoing IT revolution. These changes have important implications for inflation and pass-through, too.
One such change is the growth of e-commerce, or, as some call it, the “Amazon effect”. Many commentators argue that increased price transparency when shopping online has contributed to keeping inflation low in advanced economies. Whether the growth of e-commerce makes prices less responsive to wages is more debatable. Research suggests that the algorithmic pricing technologies of online retailers make retail prices more sensitive to aggregate, economy-wide shocks – which would imply that pass-through should eventually occur.
However, while the optimistic view is that e-commerce, and the IT revolution more generally, contributes to more competition and economic dynamism, there are also concerns about its more pernicious longer-term effects, such as: rising market concentration, a falling labour share and slower long-term growth.
One of the factors behind these negative effects is that IT makes it easier for highly productive “superstar” firms to expand into new markets, generating efficiency gains in the short run, but leading to more concentration in the long run.
A more general trend leading in the same direction is the rise of intangible inputs, such as brands, software, technologies and databases. Investment in these types of input is steadily increasing as a share of overall investment, and they are becoming increasingly important for firms’ success. The key feature of intangible inputs is that they can be deployed on a larger scale at almost zero additional cost. In other words, they primarily contribute to firms’ fixed costs. The increase in the role of fixed costs and the diminishing role of variable costs can have important implications for the structure of an industry: these changes contribute to an increase in concentration, a long-term decline in productivity (after an initial surge) and a fall in the labour share. The lesson for our discussion is that the rise of intangibles implies that marginal costs (such as wages) may be playing a decreasing role in price-setting decisions, thereby leading to a lower wage-inflation pass-through.
Some theories suggest that the wage-inflation pass-through might also fluctuate depending on the level of inflation. They start from the empirical observation that, at lower levels of inflation, price fluctuations are less persistent. In such an environment, firms are less inclined to match cost increases with price increases. Hence, low inflation leads to lower pass-through and a reduction in the expected persistence of cost and price changes.
Turning to the euro area data, we can observe that the recent decoupling of wages and prices is not unprecedented. We have seen a similar decoupling before, in 1999-2000 and in 2005-06. Formal econometric exercises confirm this observation and suggest that the wage-inflation pass-through is state-dependent. One finding is that the pass-through depends on the type of shock hitting the economy: it is stronger for demand shocks, and weaker for supply shocks.  Another finding that is particularly relevant now and confirms the theories I have just discussed, is that the pass-through is lower in times of low inflation. Other research has also found that the pass-through is lower in times of economic slack.
To sum up, the current low unemployment rate and the recent wage increases do not necessarily imply that higher inflation is around the corner. First, measuring economic slack is fraught with well-known difficulties. Second, both theory and empirical studies suggest that the wage-inflation pass-through varies over time and depends on many factors. Arguably, the structural changes under way in advanced economies imply that the pass-through may be on the decline. The bottom line is that monetary policy decisions need to be based on a wide array of economic indicators, not just the unemployment rate and wages.
I have touched upon some of the economic effects of the IT revolution. While there might be dangers ahead, I have no doubt that, with appropriate regulation, this revolution will contribute to growth and prosperity in Europe and the world. It will also change the way we work in central banks. We are actively trying to foster this change.