What will the shape of the recovery tell us about the best policy response?

As lockdown restrictions ease, economic activity in the UK is beginning to recover. Will the economy bounce back to where it was before the onset of the Covid-19 crisis? Or will we be left dealing with the economic fallout of coronavirus for the foreseeable future?

There is a description of economic cycles which says that the economy goes up like an escalator but down like an elevator (the American word for a lift). The point is that the economic cycle looks nothing like the smooth waves you may have seen used to depict them, but rather the evolution of the economy is typically a process of gradual increase (‘economic expansion’) punctuated by sharp, but short-lived, decline (‘economic recession’).

We know that economic activity in the UK, and globally, has been affected by the Covid-19 pandemic. What has been particularly striking is the speed and depth of the decline in economic activity compared even with past major recessions. This is not surprising given that both policy measures (lockdown) and reaction to the fear of infection led to a sudden cessation of many activities that form an important part of measured economic activity.

But it also raises questions as to whether the recovery, when it comes, will be as quick. Could we go down in the elevator and back up in the elevator? This article considers what shape the recovery might take, and why reasonable people seem to hold quite different views about this.

What is economic output?

One specific way that economic activity is measured involves calculating Gross Domestic Product (GDP). GDP adds up the market value of all the final goods and services produced in a country within a period of time. So quarterly GDP represents the value of goods and services produced in that three-month period.

Two important distinctions should be borne in mind. First, GDP is a flow, not a stock. This means that each new quarter (or year) the measure resets to zero and we start to count again. It also means that the exact timing of when the goods and services are produced within a quarter does not matter for the overall quarterly level. And there is a lot of variation within any given quarter; for example, far fewer goods are produced at night, or at the weekend, than are produced in normal working hours during Monday to Friday.

The second distinction is between the actual amount of activity – the level of GDP – and its rate of change – GDP growth or what is more casually called economic growth. Much of the discussion of economic growth focuses on the growth rate, yet the underlying level is the more important barometer of economic activity; the growth rate can bounce around due to the noisy nature of economic transactions.

How is GDP affected by Covid-19?

Following the government-mandated lockdown from 23 March, UK businesses were restricted from producing the amount of goods and services they usually would, and so measured GDP declined. According to the first estimates, the level of GDP in the second quarter of 2020 was approximately 25% lower than in the first quarter. This is a huge drop in GDP in a single quarter, but it is also strangely desirable given the importance of lockdown as a public health measure.

Of course, GDP is not the final word on the health of the economy. Home production (childcare, DIY, cleaning or other activities done within the household) may have increased, but because these activities are not valued in the market (where we can measure prices and so values), they do not contribute to GDP. Despite GDP’s many deficiencies (Coyle, 2015), and a growing number of alternative statistics, it is the foremost measure of the state of the UK economy.

Related question: What are the key sources of data for measuring the economy in a crisis?

What will the recovery look like?

Forecasting is always difficult, even in normal times. The unusual nature of the economic slowdown in 2020 makes it even more difficult as it is a particularly uncertain time.

Part of the challenge is that uncertainty itself can affect the performance of the economy through reduced investment by firms and the reluctance of households to spend (as discussed in this article on the damaging effects of uncertainty). This means that some of the uncertainty in forecasting how GDP will grow is in assessing how much uncertainty there is and when it will abate.

Economists who have discussed the post-Covid-19 recovery have tended to describe the patterns of recovery in terms of four broad shapes that the recovery might take. In this article, we describe these and explain what beliefs might lead someone to think one of these paths of recovery is more likely than others.

What we hope to make clear is that different people can reasonably disagree on these judgements – and they do. This does not suggest a weakness of economics as a profession, but rather highlights the highly uncertain environment in which we have to form predictions, with the Covid-19 pandemic being even more uncertain than usual.

To highlight the differences, consider an economy that would have grown very gradually (and steadily) if the Covid-19 pandemic had not occurred. The level of GDP in this hypothetical economy is 100 in the first quarter of 2020 (January to March).

The pandemic hits, however, so that starting from March, but particularly in April and May, economic activity declines sharply, such that the level of GDP is around 25% lower in the second quarter of 2020. In the last month of the second quarter, June, the economy slowly starts to recover as lockdown measures are eased.

The four shapes of recovery that we might see beyond the second quarter (from July onwards) are as follows:

A wave-shaped recovery

Lockdown restricted our ability to spend and produce. The most optimistic response would be one where, once lockdown is ended, we go back to the same activities we would otherwise have done, and also make up for the lost working and spending from lockdown.

In Figure 1, we assume that the increase in consumption after lockdown is spread out over the next 18 months, although we do assume some front-loading of the spending. Once GDP has made up the gap, we continue as we would have done before the disruption of coronavirus.

Figure 1: A wave-shaped recovery

A graph showing an illustrative path for GDP in a wave-shaped recovery

This would only be a plausible scenario if we made the following four assumptions:

That there is no loss of income or permanent business closure resulting from lockdown.
That the ‘pent-up’ income that has been sitting idle during lockdown is spent, rather than saved, when restrictions ease.
That the fall in economic activity is only due to supply restrictions: consumers are not spending and businesses not investing only because there are restrictions on doing so. Once those restrictions are lifted, consumers will spend the income they saved during lockdown on the goods and services that they would have wanted to buy but were unable to do so. This rules out any hesitancy of consumers to return to economic activity owing to uncertainty about the future of the economy, fear of contracting Covid-19, or having developed new habits of economic behaviour that replace old market spending (such as home baking of bread).
That there is no long-term ‘scarring’ of the economy from the downturn – in other words, that the lockdown period of unemployment for both productive capital and labour does not lead to a permanent increase in the unemployment of either.

A V-shaped or U-shaped recovery

In this scenario, we do not actually make up the lost activity, but we do return quickly to our pre-Covid-19 activities. The lockdown period ends up resulting in a one-off loss of output, but otherwise it is back to business as usual.

Figure 2: A V- or U-shaped recovery

A chart showing an illustrative path for GDP in a U-shaped recovery

If we relax assumptions 1 and 2 from the previous scenario, but keep 3 and 4, then a V-shaped recovery is plausible. Relaxing those assumptions means that consumers and businesses may have lost some income during lockdown, and so are unlikely to ‘over-spend’ once lockdown finishes. Or perhaps the form of spending that was forgone during lockdown is one that is unlikely to be made up afterwards – for example, it would be surprising if following lockdown, consumers decide to have the additional haircuts that they missed (although they might spend extra on a fancier haircut).

That said, assumptions 3 and 4 are still vital. Spending must return to its pre-crisis level, and so cannot be impeded by uncertainty about the economy, fear of coronavirus or permanent scarring of the economy.

A swoosh-shaped recovery

In this scenario, activity takes much longer to return to its pre-crisis trend.

Figure 3: A swoosh-shaped recovery

A chart showing an illustrative path for GDP in a swoosh-shaped recovery

If we relax assumptions 1, 2 and 3, then this is the likely outcome, where demand factors weigh on the economy. Consumers and businesses are timid in their post-lockdown spending owing to a range of factors, including uncertainty about the state of the economy, fears of engaging in economic activity that requires face-to-face interaction and a sluggish global economy.

Related question: Is the Covid-19 recession caused by supply or demand factors?

Incomplete recovery in the next three years – L-shaped

This scenario is similar to the swoosh but builds in a permanent loss of output, as a consequence of relaxing assumption 4. This would be a likely outcome if, for example, social distancing becomes the norm, which could make certain industries permanently weaker. It would also be likely if periods of unemployment for workers led to a permanent loss of skills and periods of idleness for machines led to capital scrapping.

Figure 4: An L-shaped recovery

A chart showing an illustrative path for GDP in an L-shaped recovery

What will determine the shape of the recovery?

The scenarios and assumptions presented above are extremely simplified, and neglect many other factors that economists will consider. But they highlight how it is possible to look at the same data available today and extrapolate to very different potential futures.

One issue that will be important in determining which of the scenarios outlined above takes place is the distributional effects of the current crisis. Assumptions 1 and 2 – those that allow us to take the most optimistic view of the recovery – essentially rely on consumers saving during lockdown, and then spending those savings afterwards.

This kind of behaviour is not an unreasonable assumption for people in some parts of the income distribution. But evidence shows that those who have managed to save during lockdown are the rich, who are unlikely then to increase their spending following the easing of restrictions (Kenway et al, 2020).

Assumption 3 also seems unlikely. Evidence shows a significant role for the demand side in the evolution of the UK economy under Covid-19. Moreover, data show that consumer spending began to fall before the imposition of lockdown in the UK, suggesting that consumer confidence may remain low for some time to come.

Whether or not there are long-term consequences of the current downturn is more difficult to discern (assumption 4). Long-term consequences that result from things tending to stay as they are (‘hysteresis effects’) are relatively well documented for ‘normal’ recessions (Ball, 2009). But this recession is not the same as others.

In previous downturns, recovery was slowed by the fact that households’ savings were used up, and companies may have become heavily indebted trying to sustain their businesses in spite of low demand. That need may be reduced in the current crisis, conditional on sufficient government support during lockdown. Nevertheless, there is historical evidence that pandemics can have longer-run effects on the economy (Jordà et al, 2020).

So how do the data look now?

Monthly GDP estimates from the Office for National Statistics (ONS) suggest that GDP in May 2020 was 25% lower than three months earlier. There has been no official GDP estimate concerning the post-lockdown period, so it is difficult to make any predictions about the shape of the recovery from the GDP data.
PMI survey data (a monthly business survey of purchasing managers that is highly correlated with GDP) for July showed that the economy may have started to expand, but not enough to offset the lockdown contraction.
Other high frequency data are showing more promising signs. The Bank of England uses financial transactions data such as use of debit cards in shops to measure sales (CHAPS payments) and it shows aggregate spending in July back to 10% below baseline, compared with 40% below during the height of lockdown (Haskel, 2020). Google mobility data has shown visits to cafes and restaurants are 35% below baseline, compared with nearly 80% only weeks ago (Financial Times, 2020).
Retail sales data from the ONS shows that the total volume of sales has already rebounded to the pre-crisis level, although the composition has changed and online sales make up a larger portion of sales now.

To demonstrate the key issues, we gave simple and starkly different scenarios above; in reality, the recovery could be a mix of these (yielding shapes for which we could not yet find names).

For example, initial demand might appear high as consumers catch up on haircuts and DIY purchases to continue projects started in lockdown. But after an initial bounce back, demand could settle at a lower level than the pre-crisis average. This suggests that even as official data are released, policy-makers will need to continue to monitor the higher-frequency data to see to what extent any economic rebound is sustained.

The other clear omission from what we have discussed so far is what happens with a resurgence of Covid-19 and a second (perhaps even a third) lockdown.

It is not clear whether to think about further lockdowns as having the same effects as the first ones. On one hand, they may not be so damaging: the most vulnerable sectors will have already been hit, so output may not be able to fall any further in those sectors. On the other hand, if the capacity of firms and households to weather the economic storm has been used up by earlier lockdowns and/or government support is less generous, the effects could be much worse. The overall effect, as with so much right now, is subject to enormous uncertainty.

Related question: Why is uncertainty so damaging for the economy?

Where can I find out more?

From lockdown to recovery – the economic effects of Covid-19: Jonathan Haskel, a member of the Bank of England’s Monetary Policy Committee (MPC), gives his perspective on the UK economy following lockdown.

The second quarter: Bank of England chief economist Andy Haldane gives his view on the likely shape of a recovery.

Covid-19 and the economy: what are the lessons so far? MPC member Silvana Tenreyro gives her outlook on the UK economy.

Prospects for the UK economy: Cyrille Lenoël, Rory Macqueen and Garry Young provide NIESR’s latest UK economic forecast.