Why is it hard to compare the economic costs of lockdown across countries?

The urge to compare the UK’s economic performance with other countries is natural in times of disruption. Globally recognised measures like GDP seem to allow this. But all is not as it seems: the UK’s service economy is one reason why the costs of Covid-19 may be especially high.

In the first half of 2020, previously normal activities became impossible almost overnight. This was due either to public policy measures (national lockdowns) or to the fear of infection, which led to a widespread voluntary withdrawal from social settings in the UK even before lockdown began. As a result, this year has seen a decline in UK economic activity that is historically unprecedented in terms of both its depth and speed.

The experience of lockdown and fear is common in almost all economies affected by the pandemic. Yet there is a surprising amount of variation in the economic impact of coronavirus across different countries. This raises important questions: can we make sensible comparisons of the economic cost of lockdown across countries; and what should we keep in mind when drawing such international comparisons?

How has the economic impact of Covid-19 varied across countries?

The most common way to compare countries is to look at GDP. Figure 1 presents growth rates of GDP in both the first quarter of 2020 (January to March) and the second quarter (April to June) for the UK and several other advanced economies.

The chart makes for grim reading for the UK. In the second quarter of 2020, UK GDP fell by 20.4%, the largest measured fall in GDP across major economies and by some distance the largest drop ever recorded. (Quarterly data adjusted for inflation go back to the Second World War, but even the data unadjusted for inflation back to 1920 have never seen a bigger decline in a single quarter).

The next largest single-quarter decline in national GDP was recorded in Spain, where GDP fell by 17.8% (also in the second quarter). By simply comparing growth rates, the UK appears clearly to be among the worst affected countries, with the initial decline in activity almost twice as large as those suffered by Germany and the United States.

Figure 1: International comparison of GDP growth rates in 2020

Figure showing international comparison of GDP growth rates in 2020

Are there reasons to be cautious about international GDP comparisons?

Just how informative are international comparisons of GDP when assessing the economic costs of Covid-19? There are (at least) three difficulties that need to be borne in mind when comparing its cross-country economic impact using GDP data:

Timing: countries differed in the timing (and severity) of the public health restrictions that they implemented.
Pre-Covid-19 growth: countries varied in their expected GDP growth rates for 2020.
Measurement: countries define and measure ‘economic activity’ in slightly different ways.

What is the role of the timing of public health restrictions?

Although almost all major economies implemented restrictions on activity to help prevent the spread of Covid-19, there was variation in terms of both speed – how soon restrictions were implemented – and duration – how long they were kept in place before being relaxed.

In general, the timing of the implementation (and relaxation) of public health restrictions will greatly affect measured GDP in the corresponding quarter, and therefore the corresponding growth rate between quarters. This relates to a technical (but important) point about GDP discussed here: GDP is what is known as a ‘flow’ variable. This means that at the start of each new quarter, the measure of quarterly GDP resets to zero and we start to count again. Focusing on headline growth rates (as in Figure 1) can therefore be misleading about the underlying state of the economy.

For example, in both Italy and Spain, two other European countries that experienced very high levels of Covid-19 transmission during the initial outbreak, public health restrictions were implemented earlier than in the UK, where they were not introduced until late March and so did not affect measured first quarter output as much. Therefore, the economic effects of the pandemic are largely captured by GDP in the second quarter for the UK, whereas in Italy and Spain, the effects are more evenly distributed over two consecutive quarters. This is reflected in Figure 1.

A more reliable indicator of the trend in economic activity is the underlying level of GDP. Moreover, to make the comparison fairer, we can standardise the level of GDP at the pre-pandemic peak, using the final quarter of 2019.

This comparison is presented in Figure 2. Looking now at the level of GDP, for a limited set of the countries from Figure 1, we can see that Spain appears to have suffered more than the UK. Compared with other advanced economies, Spain suffered very weak economic activity in both the first and second quarters. Nevertheless, it is clear that the UK has suffered a very large economic cost due to the pandemic by international standards.

Figure 2: Comparing the level of GDP across the UK, Ireland, Spain and France

Figure comparing the level of GDP across the UK, Ireland, Spain & France

What is the role of varying growth expectations?

Without Covid-19, most economies were expected to grow during 2020 – that is, GDP was expected to increase. But the amount that they would grow varied substantially across countries.

To get a better understanding of the ‘true’ economic cost of the pandemic, we should compare GDP outcomes with the expected path of GDP for each country. This is because for countries that were expected to have high growth during 2020, we understate the extent of the economic losses of the pandemic when we compare to the end of 2019. In high-growth economies, GDP would have been a long way above the end of 2019 level without Covid-19.

According to forecasts by the International Monetary Fund (IMF) published in October 2019, the UK economy was expected to grow by 1.4% during 2020. So very roughly, we should have expected about 0.35% extra level after the first quarter, and 0.7% extra after the first half of 2020. In contrast, Ireland as a higher growth economy should have grown 2.5% extra GDP over the first half of 2020. The equivalent for Spain is 1.35% extra GDP, while for Germany, which was expected to grow less quickly, the problem is smaller (0.9% extra GDP in the first half).

To take these expected differences in growth rates into account, we can instead compare the deviations in GDP from its expected path across countries. Even taking account of the fairly moderate growth expectations for the UK relative to some other advanced economies, the deviation in UK GDP from its expected path is among the largest – again, second only to Spain.

(Other adjustments for different levels of population could also be done in a similar fashion using output per capita as the measure. This would matter if the population had varied significantly as a result of the pandemic).

How – and why – do different countries measure GDP in different ways?

As different countries measure ‘economic activity’ in slightly different ways, international GDP comparisons are never ‘comparing apples with apples’. While this is always true, it is particularly acute in a pandemic. This is because one of the major sources of measurement differences is how different countries measure the contribution of public services to GDP.

In the UK, the Office for National Statistics (ONS) measures the contribution of health and education to GDP using indicators such as the number of people treated and the number of pupils in school.

Although hospitalisations increased hugely during the first wave, many other routine medical services provided by the NHS were paused due to the need to redeploy staff and other resources. Moreover, all schools in the UK were closed as part of the public health restrictions. As a result, the measured contribution of public services to UK GDP fell significantly, acting as a large negative drag on the overall GDP estimate.

Figure 3: Comparing government consumption across the UK, Ireland, Spain and France

Figure showing government consumption

In contrast, several other European countries define the contribution of public services as the amount of money spent in providing them, such as investment or staff pay. Given the historically high levels of public spending across many economies in responding to the pandemic, defining public services in this way potentially contributes positively towards measured overall economic activity.

This is illustrated in Figure 3, which shows overall government consumption in the UK and France falling sharply during the pandemic, whereas in other European countries, this measure actually increases.

Finally, it is always worth noting that measuring the total economic activity in an economy accurately in real-time is incredibly difficult. As such, we should interpret initial GDP estimates with some caution given that they are subject to future revisions.

Why might the economic cost of Covid-19 in the UK be particularly large?

Although there are clearly some difficulties associated with comparing economic performance across countries, it is clear from the comparisons above that the UK has suffered a very large economic cost due to coronavirus. And even if we could compare reliably across countries, there are two key reasons why we might expect still to see different effects of Covid-19 across different economies. The first is the very different sectoral make-up of national economies; the second is differences in the circulation of Covid-19.

Sectoral make-up of economies

The contribution of different economic sectors to measured GDP varies hugely across countries. Given that public health restrictions have a varied impact on the ability of different sectors to operate normally, it is important to keep in mind that countries differ markedly in their sectoral make-up when comparing the economic impact of the virus internationally.

For example, the experience of the pandemic has illustrated the importance of ‘social consumption’ to economic activity. This refers to services provided by the hospitality industry, restaurants and pubs, hairdressers, nail salons, etc.

A 2010 study by an interdisciplinary group of researchers on the possible macroeconomic impact of a pandemic on the UK highlighted that social consumption constitutes around a third of overall consumption in the UK. Hence economic activity in the UK was perhaps particularly vulnerable to the introduction of public health restrictions compared with other advanced economies.

Figure 4 illustrates this point further by plotting monthly UK GDP by sector during 2020. The provision of accommodation and food services in the UK fell to around 10% of their February 2020 levels during lockdown. There was also a large fall in other social activities, such as arts, entertainment and recreation. Overall, measured UK GDP in the first two quarters of 2020 has been significantly weighed down by this large fall in social consumption activities.

Figure 4: The sectoral evolution of GDP highlights very different lockdown experiences

Figure showing the sectoral evolution of GDP

In contrast, some sectors actually saw an increase in demand due to the pandemic – such as the manufacture of pharmaceuticals in the UK (see Figure 4). In countries where economic activity is weighted towards sectors that are more resilient to (or even directly benefit from) a pandemic, we would naturally expect GDP to decline less.

For example, Ireland initially saw overall growth in total industrial output at the outset of the pandemic because of the contribution from its large pharmaceuticals and medical equipment production sectors, whereas the contribution of this sector is relatively smaller in the UK.

Another aspect worth bearing in mind when making international comparisons is that while the nature of the pandemic is global, countries differ in their exposure to the slowdown in other countries. Moreover, the nature of the pandemic is such that it places restrictions on the movement of people more than the movement of goods (which are also affected). This may make countries like the UK, where international trade in services is an important part of the economy, even more exposed.

Circulation of Covid-19

One other reason why a country might suffer larger economic losses is that coronavirus has remained prevalent there for longer. It appears that there is no trade-off between the economic cost of Covid-19 and public health outcomes. Indeed, if anything, countries that have suffered high mortality rates from Covid-19 tend to have experienced higher economic costs (at least measured using GDP and so subject to the concerns already raised).

There are many possible reasons for this. Given the exponential growth of the virus in the early days of its prevalence, a delay to lockdown is likely to have contributed to greater transmission, requiring that the eventual policy actions taken are even more severe. Greater control of the virus may also give rise to more confidence among businesses and consumers to resume something closer to normal economic activity more quickly.

Relative to the size of its population, it is clear that the UK has suffered a particularly high cost in terms of public health. Since the beginning of March, the UK’s relatively high circulation of Covid-19 has led to a very high mortality rate compared with other countries that imposed restrictions earlier.

Moreover, compared with other countries, these restrictions were in many cases not as severe in terms of the impact on individual freedoms. While this may have delayed the initial onset of economic costs, the overall costs may end up being larger as a result.

Is GDP the only statistic for assessing the economic costs of Covid-19?

Despite GDP’s many deficiencies (see Coyle, 2015) – and a growing number of alternative statistics – it remains the leading indicator of the state of the UK economy. But GDP is a very specific measure of economic activity. While it may often capture general swings in economic activity, it is by no means the only indicator we should care about for judging either the health of the economy or the ‘true’ costs of the Covid-19 pandemic.

Many goods and services have been provided during the pandemic that are not traded, and therefore do not count towards measured GDP. Some examples are the increase in childcare done from home as a result of schools closing, as well as other activities such as increased levels of household maintenance.

Likewise, GDP will not take into account the many other ‘costs’ of the pandemic – for example, the large increase in reported mental health cases resulting from separation from family, friends and social settings in general.

So even if we could compare GDP perfectly across different countries, we should remain wary of such comparisons given that there are many forms of activities that contribute ‘value’ that GDP does not capture, as well as many other forms of ‘cost’ that are difficult to quantify.