Uncertainty matters greatly for households and businesses when taking decisions. Visibility about the future path of earnings, employment, sales volume, profits, etc. declines when uncertainty increases.

This may trigger a wait-and-see attitude whereby postponed spending and investment decisions weigh on growth. The assessment of the potential consequences of uncertainty supposes that the latter can be appropriately identified and measured. The former refers to the fact that uncertainty can have many causes: economic, economic policy, political or even geopolitical. Depending on the nature of the uncertainty, the consequences may differ.

Measuring uncertainty is a challenge. A widely used metric is based on media coverage. This offers the advantage that specific sources of uncertainty can be monitored. In this respect, chart 1 shows a geopolitical risk index. The start of the war in Ukraine led to an exceptionally large increase in this index although the attacks on 11 September 2001 had an even bigger impact. Media coverage, after spiking, tends to decline swiftly but this does not mean that the uncertainty confronted by economic agents has disappeared. This calls for surveys in which people are asked how they feel about the future. Since May 2021, the European Commission has added an uncertainty question to its business and consumer surveys[1]. Participants indicate how difficult it is to make predictions about their future business/financial situation. This implies that fluctuations in the answers can have many causes.

Geopolitical risk index

As shown in chart 2, the spreading of Covid-19 infections in the spring of 2020 caused a huge increase in uncertainty of households and businesses. This was followed by a gradual decline but, one year later, uncertainty was still above its pre-pandemic level. In the summer of last year, it started to rise again and following the war in Ukraine, uncertainty jumped in March but hardly changed in April. Chart 3 provides greater detail. Interestingly, consumers’ uncertainty has been on a rising trend since summer last year, which probably reflects the impact of high and rising inflation. Uncertainty in retail trade shows the waves of lockdowns and mobility restrictions and, more recently, probably also the impact of inflation.


Table 1 compares the change in uncertainty triggered by the Covid-19 pandemic and the war in Ukraine[2]. Looking at the cumulative increase over two months, Covid-19 caused a far bigger increase of aggregate uncertainty than the war in Ukraine. Whereas the former saw rising uncertainty in March and April 2020, as the pandemic was spreading, the latter led to an increase in uncertainty in March, followed by a small decline in April. The effect of Covid-19 on uncertainty at the aggregate level as well as for consumers and business sectors, was similar for the European Union countries and the euro area. This also applies to uncertainty caused by the war in Ukraine, except for consumers’ uncertainty, which increased more in the euro area.


Uncertainty in the EU and the euro area

Finally table 2 looks at the change in uncertainty in following the war in Ukraine. All countries except Malta saw an increase in March but the dispersion is considerable. Uncertainty declined in most countries in April except for Denmark, Greece -where there was a big increase-, Spain, France, Hungary, the Netherlands and Slovakia. It was stable in Portugal. It will be important to monitor the development of uncertainty in the coming months at the level of consumers, businesses and individual countries. In the absence of a decline, one should expect that the impact shows up in spending and activity data.


Uncertainty by countries



William De Vijlder


The author thanks Elias Krief for the data analysis used in this text.


[1] The data cover the EU countries as well as key economic sectors (consumers, industry, services, retail trade, construction). Source: https://ec.europa.eu/info/sites/default/files/bcs_user_guide.pdf.

[2] The monthly changes are expressed in z-score to take into account possible differences in the mean and standard deviation of the various series.