As many countries face their second or even third lockdowns (Germany, Austria, Israel), others (Switzerland, the United States) have done what they can to escape such repetitions. Much of the discussion about this drastic move involves its costs.

As Covid-19 became pandemic in early 2020, most countries opted for a similar set of measures. They implemented social distancing, slowed down the traffic of people in public spaces and eventually put a stop to large amounts of economic activity. While countries structured lockdowns in different ways, they all generally closed some parts of their economies and limited social activities for lengthy periods.

In some countries, for example in China or Italy, the lockdown entailed physically closing points of entrance and exit to certain regions. In others, like France and parts of the United Kingdom, the lockdown included curfews. In countries such as Germany, Austria or Switzerland, lockdowns meant closing all restaurants, entertainment locations and nonessential shops. These measures led to a considerable slowdown or even a halt in communal practices, be they social or commercial.

Countries have also used their lockdowns in different ways due to varying public discussions about their efficacy. Some question if lockdowns mitigate the pandemic or are compatible with human rights. In the economic realm, the main debate is over how much they cost.

Direct costs

Before continuing, it is worth adding a general methodological disclaimer. Here, we will address only the lockdown’s economic cost in the first half of 2020 since the best research and comparable data regards that time frame. Also, this is an exercise in approximation, as is all empirical economics. There are many more factors at play than can be covered by any model. The conclusions of these approximations are to be taken with a grain of salt.

There are two ways to assess the economic costs of the lockdown. One is calculating the fiscal (and possibly monetary) expenditures that governments (and central banks) use to soften the economic impact. Typically, this involves direct aid to affected firms, subsidized credit lines and labor market stabilizers like compensating workers for lost hours.

These costs are cash outflows that can be determined using standard economic accounting tools. This way, we find that in the European Union, for example, the lockdowns have cost around 540 billion euros. This is the sum of the programs set in motion via the European Stabilization Mechanism, the European Investment Bank and the European Commission.

Many objections can be raised to this method of calculating costs. Some of these outflows have not occurred yet. If they do not happen at all, they must be deducted from the total cost amount. Then, some of these programs were already in the political pipeline well before the pandemic hit. Once it did, leaders sped up the political process in which they were to be released. Finally, and more importantly, many of the measures are not costs as such – it is expected that they will be paid back or absorbed by other economic agents.

Relative costs and duration

Relative measures are more challenging in theory but better anchors in practice. They involve estimating how much gross domestic product (GDP) was lost during a lockdown. Methodologically, this is hard to determine. It is difficult to separate what would have been economic expansion without the pandemic, expansion with Covid-19 and expansion with the pandemic and lockdown. Nonetheless, with the appropriate simplifications and statistical tools, economists can sketch a rough picture.

Once the relative costs of the lockdown in terms of GDP loss have been established for different countries, one can compare the magnitude of these losses. Relative costs are a comparison in two steps: one between economic growth without the pandemic and economic growth with the lockdown, and the other between different countries.

However, this approach would neglect an essential aspect of the lockdowns: the countries implemented them separately. One important difference is that each nation had a different lockdown duration. A more realistic comparison would contrast the loss of GDP with the length of the lockdown. Using available data from the International Monetary Fund and cross-referencing it with information released by governments, the graph above plots each country as a data point of loss in percent of GDP and days in lockdown.

The message is straightforward: The longer the lockdown duration, the larger the GDP loss and, therefore, the economic cost. This might not surprise. What does surprise is how unrelated these two parameters are. If regressed against one another, there is only a meek correlation between them. For every additional day of lockdown, only 0.04 percent of GDP is lost. Additionally, it shows that duration is not very predictive of economic shrinkage. In other words, the link between lockdown duration and loss of GDP is weak.

Does that mean that lockdowns do not cost much? Could it be that the pandemic itself is what hurts, and the lockdown comes at little economic detriment? Not precisely. Lockdowns have not only differed in length, but also stringency. It seems like a better approach, then, is to analyze the loss in GDP growth against lockdowns’ strictness.

The International Monetary Fund has done precisely this. It regressed data on loss of GDP growth against the so-called stringency of the lockdowns. It is measured as an index averaging several sub-indicators – school closures, workplace closures, cancellations of public events, restrictions on gatherings, public transportation closures, stay-at-home requirements, restrictions on internal movement and controls on international travel – provided by the University of Oxford’s Coronavirus Government Response Tracker.

The graph above shows the relationship between countries’ loss in GDP in percent and the stringency of the lockdowns. Here, the trend line shows a stronger correlation. The more stringent the lockdown, the heavier the toll it takes on GDP. The trend line is only about 22 percent predictive of the data, which is rather weak. But this weakness occurs because of an interesting economic phenomenon. The data shows that the proportion of the loss in GDP grows as the stringency increases.

If only the countries with a stringency value over 45 are analyzed, the negative relationship more than doubles, from -0.22 to -0.48. In other words, the downside of stringent lockdowns is disproportionately large to their increase in strictness. The message is twofold. First, the more rigid the lockdown, the more economic cost it creates – and that cost is significant. Second, from a certain degree of stringency on, the economic costs of the lockdown grow to a greater degree than the strictness of the lockdown itself. In other words, lockdowns hurt the economy disproportionately more the stricter they get.

Three scenarios

This discussion has both economic and policy implications. From an economic point of view, lockdowns have costs. These costs seem more related to their strictness than to their duration.

In the best case, policymakers refrain from implementing lockdowns. This approach minimizes the impact of the pandemic on the economy. The data suggests that there would be growth if there were no lockdown.

Suppose further lockdowns are necessary. In the next best-case scenario, policymakers will limit their stringency to the point that the measures do not take a disproportionately large toll on the economy.

In the worst case, policymakers assume lockdowns have a proportionate negative impact on the economy. Their measures for mitigating the adverse economic effects of the lockdown are implemented based on this assumption.

In this scenario, policymakers devise ever-stricter lockdowns, neglecting to adjust their economic policies to incorporate the disproportionately large economic impact of this increase in stringency. This approach creates two economic gaps – one in loss of GDP and one in the mismanagement of mitigation measures. As many countries are approaching even more stringent lockdowns, this worst-case scenario may already be underway.