Lecture 19 - COVID-19's Economic Damage I
Table of Contents
Introduction: A Major Recession and Externalities
The Scale of the Recession
The COVID-19 pandemic triggered one of the most severe economic contractions in modern history. Unlike the 2007–09 Global Financial Crisis, which originated endogenously within the financial sector through credit market failures and balance-sheet deterioration, the COVID-19 shock was exogenous in origin, arising from a public health emergency that then reverberated through the real economy via multiple, mutually reinforcing channels. The novelty and speed of transmission made it extraordinarily difficult for policymakers, firms, and households to anchor expectations, producing elevated uncertainty that itself became an independent source of economic damage.
The left-hand panel of this slide illustrates the sharp sell-off across major global equity indices in early 2020. The FTSE MIB (Italy) and FTSE 100 (UK) fell by approximately 30 per cent from their January 2020 levels by mid-March, while the S&P 500 (USA) declined by a similar magnitude before recovering. Notably, the Shanghai Composite (China) fell less steeply, partly reflecting earlier shock absorption and a faster initial lockdown. The right-hand panel tracks UK GDP as a percentage deviation from its pre-pandemic (February 2020) level, sourced from the ONS Monthly GDP series. The first national lockdown produced a contraction of approximately 25 per cent, an unprecedented peacetime fall. Crucially, subsequent lockdowns in autumn 2020 and winter 2020/21 generated far shallower declines, consistent with the hypothesis that households and firms adapted behaviour, shifting to remote working, online commerce, and other substitutes that partially insulated economic activity from the public health restrictions.
The diminishing severity of successive lockdowns illustrates learning and adaptation effects: firms reorganised production processes; consumers shifted to digital substitutes; and furlough schemes stabilised household incomes. This asymmetry matters for modelling: the first shock was largely unanticipated and therefore maximally disruptive, whereas later shocks were partially incorporated into expectations.
If asked to compare the economic impact of successive lockdowns, emphasise adaptation mechanisms, the role of policy buffers (furlough, business loans), and the distinction between anticipated and unanticipated shocks. Reference the ONS GDP data as empirical support.
Economic Uncertainty and Durable Goods Expenditure
Economic uncertainty is not merely a background condition during crises; it is an active contractionary force. The World Uncertainty Index (WUI), constructed by Ahir, Bloom, and Furceri (2018) on the basis of the frequency of the word "uncertain" and its variants in country-level Economist Intelligence Unit reports, demonstrates that COVID-19 produced the largest spike in global uncertainty ever recorded in the index's history, surpassing even the Global Financial Crisis, 9/11, and the Iraq War.
The left-hand chart shows the normalised WUI from 1990 to approximately 2022. The index was on a rising trend from 2016, driven by Brexit, US–China trade tensions, and US political risk, before spiking dramatically at the onset of COVID-19. The right-hand chart documents the consequences for a canonical durable good: Chinese car sales collapsed by 92 per cent in the first two weeks of February 2020 relative to the year-prior period. This is an extreme illustration of the precautionary savings motive and the irreversibility effect on durable goods spending.
In standard permanent income models, a household facing an income shock will smooth consumption by drawing on savings or borrowing. However, durable goods purchases are inherently lumpy and reversible only at a cost. When uncertainty rises sharply, households with a high marginal propensity to consume (MPC) — typically those with limited liquid assets — respond by postponing large discretionary outlays such as vehicle purchases. This precautionary motive, formalised by Kimball (1990) and empirically studied by Carroll (1997), implies that uncertainty shocks can generate demand contractions that are disproportionately large relative to the underlying income loss.
Do not conflate the precautionary savings motive with simple income effects. Even if expected income is unchanged, a mean-preserving spread in the income distribution (i.e. increased uncertainty with no change in the mean) is sufficient to trigger precautionary saving if households have prudent preferences (positive third derivative of the utility function).
Restaurants, Travel, and Contact-Intensive Services
The pandemic shock was not uniformly distributed across sectors. It disproportionately affected contact-intensive services: industries where production requires physical proximity between producer and consumer and therefore cannot be relocated to remote or digital environments without fundamental restructuring.
The left panel of this slide shows year-on-year changes in restaurant reservations and walk-ins tracked via OpenTable data. From mid-February 2020, bookings in the United Kingdom, the United States, and globally collapsed towards zero by mid-March, with the UK reaching approximately a 100 per cent decline as lockdowns were imposed. The right panel reports the year-on-year change in US outbound flight bookings by region between 6 January and 8 March 2020. Asia Pacific bookings fell by 98.1 per cent, reflecting the earlier onset of the outbreak there, while bookings to Europe fell by 31.9 per cent. These figures underline the asymmetric geographic sequencing of the shock and the near-total shutdown of international aviation, a sector that itself generates significant upstream and downstream linkages across tourism, hospitality, and logistics networks.
The collapse in restaurant and travel data also illustrates why standard Keynesian demand-management tools face limitations in this type of crisis. A fiscal stimulus that puts money in consumers' pockets cannot easily restore demand for restaurant meals if the restaurants are legally closed or if consumers fear infection. The constraint here is not primarily a lack of purchasing power but a disruption to the feasibility of transacting. This is why some economists described the pandemic shock as a "wall between supply and demand."
Externalities Generated by the Pandemic
A central analytic lens for understanding the economics of COVID-19 is the concept of externalities: situations in which an individual's action imposes costs or confers benefits on third parties that are not reflected in market prices. The pandemic generated both negative and positive externalities simultaneously.
The right-hand satellite imagery from NASA/ESA Copernicus compares nitrogen dioxide (NO
On the negative externality side:
- Contagion as a negative consumption externality: an individual who chooses not to self-isolate when infectious imposes a risk of illness on others. This is a textbook externality that justifies government intervention (mandated quarantines, social distancing rules, lockdowns).
- Healthcare crowding-out: non-COVID-19 patients may be displaced from intensive care capacity, so the social cost of infection extends beyond the direct victim.
- Congestion in delivery services: a surge in demand for online grocery and food delivery — driven by voluntary or mandated social distancing among high-risk groups — can create congestion externalities in logistics networks, raising prices and reducing service quality for all users.
The public health economics literature models infection as a negative externality in the spirit of Pigou (1920). The socially optimal infection rate — accounting for the spillover cost imposed on others and on healthcare capacity — is lower than the privately optimal rate, justifying corrective intervention. The standard Pigouvian remedy is a tax or prohibition equivalent to the marginal external cost; in the context of a pandemic, this translates to lockdown mandates, mask requirements, or vaccination incentives.
If an exam question asks you to apply externality theory to COVID-19, structure your answer around: (1) definition of the externality; (2) why the market fails to internalise it; (3) the divergence between private and social optimum; (4) the corrective policy (Pigouvian tax/regulation); and (5) real-world complications (enforcement, distributional consequences, uncertainty about the magnitude of external costs).
- COVID-19 caused an unprecedented GDP contraction, deepest during the first lockdown; subsequent lockdowns had smaller impacts due to adaptation.
- Global stock markets fell sharply in early 2020, reflecting both fundamental reassessment of earnings and a surge in uncertainty.
- The WUI reached historically unprecedented levels, triggering precautionary savings and collapsing durable goods demand.
- Contact-intensive services (hospitality, aviation) were disproportionately affected.
- The pandemic created negative externalities (contagion, healthcare crowding-out) and positive externalities (pollution reduction), both of which are outside the market mechanism.
The Most Affected Sectors by COVID-19 and Production Networks
Sectoral Heterogeneity in Impact
Not all sectors suffered equally. According to the McKinsey Global Institute report of 9 March 2020, the industries facing the most severe and protracted disruption were: airlines; automobiles; energy equipment and services; hotels, restaurants, and leisure; and specialty retail. These sectors share several characteristics that made them acutely vulnerable: high labour-intensity in face-to-face settings; dependence on international supply chains; high fixed costs combined with near-zero revenue; and sensitivity to discretionary consumer spending.
This McKinsey matrix summarises the estimated global restart timeline and characteristics of disruption across six broad sectors. Tourism and hospitality faced the longest duration of impact, with recovery delayed to Q4 2020 at the earliest and potential for further negative demand shocks if the disease resurged. Aviation faced a similar timeline, with domestic travel recovering approximately twice as fast as international. The automotive sector faced compounding difficulties: pre-existing vulnerabilities (e.g., declining Chinese sales, US–China trade tensions) were amplified by supply chain disruptions and a sharp fall in consumer sentiment. The consumer electronics and semiconductor industry presented a more nuanced picture: the pandemic accelerated structural shifts towards digitalisation, with some product segments proving resilient or even experiencing increased demand, though supply chain disruptions in China created downstream bottlenecks for 5G rollout and high-tech manufacturing.
The key asymmetry in sectoral recovery timelines reflects the distinction between tradeable and non-tradeable goods and between discretionary and non-discretionary consumption. Non-tradeable services (a restaurant meal, a hotel stay) cannot be stored or imported; demand deferred is largely demand destroyed. Tradeable manufactured goods, by contrast, may see some pent-up demand release upon reopening, though this effect is partial and depends on the durability of the underlying good.
Production Networks: Upstream and Downstream Spillovers
The sectoral analysis above understates the true economy-wide impact of COVID-19 because it ignores input-output linkages through production networks. Modern economies are characterised by highly complex webs of vertical specialisation: each sector purchases intermediate inputs from supplier sectors and sells its output either to final consumers or to other industries as intermediate inputs.
This diagram illustrates the stylised production network for the automotive sector. The focal sector (automotive) sits at the centre. Downstream effects are shocks that propagate from upstream suppliers (e.g., steel producers, electricity producers) down through the supply chain to the automotive firm. Upstream effects are shocks that originate in downstream customer sectors (e.g., automobile dealers, car rental companies) and propagate backwards towards the focal sector and its suppliers. A productivity shock to the steel industry, for instance, constitutes a downstream effect on automotive: it raises intermediate input costs and constrains vehicle production. Conversely, a fall in government fleet procurement of vehicles constitutes an upstream demand shock that reverberates upstream to reduce orders placed by automotive firms on steel producers.
A critical finding from the empirical production-networks literature (see Carvalho et al., 2021) is that spillover effects — both upstream and downstream — are quantitatively large relative to the "own" effect of a shock confined to the originating sector. This means that naive partial-equilibrium estimates of sector-specific COVID-19 damage substantially understate the general-equilibrium aggregate impact.
The production network framework is formalised in the Leontief input-output model and its general-equilibrium extensions (Acemoglu et al., 2012). In these models, an exogenous productivity shock to sector
Do not confuse upstream and downstream effects. Downstream effects travel FROM suppliers TO the focal sector (i.e., an input supply shock). Upstream effects travel FROM customers BACK TO the focal sector (i.e., a demand shock originating downstream).
- COVID-19 disproportionately damaged contact-intensive, supply-chain-dependent, and discretionary-spending sectors: aviation, tourism, automotive, energy services.
- Recovery timelines varied by sector; domestic services recovered faster than international ones; electronics were partially shielded by digitalisation trends.
- Production networks amplify sectoral shocks: both downstream (supplier-to-firm) and upstream (customer-to-firm) spillovers are empirically large.
- Aggregate GDP impacts of COVID-19 substantially exceeded what simple sector-by-sector arithmetic would suggest, due to network amplification.
The Race Between Supply and Demand: Transmission Channels
Framing the Shock: Supply, Demand, or Both?
A fundamental question in the economics of COVID-19 is whether the pandemic constituted primarily a supply shock or a demand shock. The answer matters enormously for policy: a pure supply shock, for instance, cannot be remedied by fiscal stimulus because the constraint on output is not insufficient demand but insufficient productive capacity. Conversely, a pure demand shock with idle capacity calls for expansionary policy to restore aggregate demand to its potential-output level. COVID-19, as the lecture demonstrates, was neither one nor the other, but rather an unprecedented simultaneous disruption to both sides of the economy, with strong complementary feedback loops between them.
The lecture distinguishes between two types of sector-level shocks in empirical work:
- A supply shock, proxied by changes in sectoral Total Factor Productivity (TFP)
- A demand shock, captured by changes in sectoral government spending
Stage One: COVID-19 as a Supply Shock
At first glance, COVID-19 has the appearance of a supply shock. Quarantine requirements and social distancing mandates directly reduced labour supply by keeping workers away from their places of work. Global supply chains were disrupted as factories shut, ports reduced capacity, and logistics networks fragmented. In the standard AS-AD framework, this is modelled as a leftward shift of the Aggregate Supply curve from
- The Great Recession of 2007–09 originated as a financial sector supply shock (credit market seizure, bank balance-sheet deterioration).
- Wars and natural disasters generate supply shocks through destruction of physical capital and permanent workforce losses.
COVID-19's supply shock was distinctive in that it was: (a) temporary in anticipated duration; (b) policy-induced rather than arising from resource destruction; and (c) highly uncertain in its duration and ultimate magnitude.
A supply shock that reduces labour supply shifts AS leftward, raising P and reducing Q. Unlike a natural disaster, COVID-19 did not destroy capital stock — it rendered it temporarily inaccessible. This distinction has implications for recovery: once restrictions lifted, latent productive capacity could (in principle) be rapidly re-engaged, producing a V-shaped recovery in some sectors.
Stage Two: The Emergence of Demand Effects
The initial supply shock was rapidly compounded by a series of demand-side contractions. The AS-AD diagram on this slide shows aggregate demand shifting leftward from
- Uncertainty about the epidemiological trajectory: households and firms could not form reliable expectations about when restrictions would end, suppressing investment and consumption decisions.
- Uncertainty about the policy response: the scale and form of fiscal and monetary policy support was itself uncertain in the early weeks, reinforcing precautionary behaviour.
- Income losses among non-permanent workers: hospitality, retail, and manufacturing workers on zero-hours contracts or temporary employment had no income floor, directly reducing their consumption.
- Precautionary savings: even households that retained employment reduced consumption, raising their saving rate to buffer against potential future income losses.
- Firm-level investment postponement: capital expenditure was suspended as firms awaited clarity on the duration of the shock and faced liquidity constraints due to revenue collapse.
In Keynesian theory, a fall in aggregate demand is self-reinforcing through the multiplier mechanism: a reduction in consumption lowers income, which further reduces consumption. COVID-19 amplified this standard multiplier by adding the uncertainty channel, which elevated precautionary savings and compressed the marginal propensity to consume out of current income. The combination of reduced income and elevated uncertainty thus produced demand contractions far exceeding what either channel alone would generate.
Stage Three: The Supply-Demand Feedback Loop
The interaction between supply and demand deterioration became mutually reinforcing, creating a feedback loop. This slide introduces the mechanism by which demand contraction feeds back into further supply contraction:
- Firms — particularly those more dependent on operating cash flows — faced a dual shock: revenues collapsed (due to demand contraction) while fixed obligations (rent, wages, debt service) remained. Liquidity-constrained firms were forced into bankruptcy, permanently removing productive capacity from the economy.
- This represents a demand-driven supply destruction: falling demand caused firm exits, which in turn contracted aggregate supply further (AS shifts further left from
The slide also draws an important distinction from war or natural disaster scenarios. In wartime, government expenditure typically surges (war production, reconstruction), pushing aggregate demand upward even as some supply is destroyed. This divergence means that wartime output contractions can be partially or fully offset by demand stimulus, and there is a risk of inflation. In the COVID-19 case, both supply and demand fell together, with the feedback loop making it qualitatively different from most historical macroeconomic crises.
Think of this as a spiral: reduced demand → firm revenue collapse → insolvencies → capacity destruction → further supply contraction → further demand contraction → deeper insolvencies. Each iteration amplifies the shock. The social planner's role in interrupting this spiral is the economic rationale for emergency furlough schemes (which maintain household income and thus demand) and business interruption loans (which relieve cash-flow stress and prevent unnecessary firm exit).
Stage Four: Further Demand Feedback
A further demand feedback loop operates through the labour market. Workers who lose jobs due to business closures lose their income entirely and therefore reduce consumption, further depressing aggregate demand. This shifts
The Unified Picture: A "Wall Between Demand and Supply"
This slide synthesises the entire dynamic sequence into a single conceptual framework. The COVID-19 virus is characterised not merely as a large shock to real economic fundamentals, but as a shock to the frictionlessness of the market: it introduced a "wall between demand and supply" — a structural impediment to the matching of willing buyers and willing sellers that would ordinarily occur in a well-functioning market. The red shaded area in the diagram represents the cumulative destruction of economic surplus as supply and demand spiral downward through multiple rounds of feedback. This wedge between what producers could supply and what consumers could demand — under normal market conditions — represents the core economic harm of the pandemic.
The "wall between demand and supply" framing resonates with the theory of search and matching frictions (Diamond, Mortensen, Pissarides). In normal times, market frictions are present but limited. COVID-19 dramatically increased these frictions — physical distancing made it literally impossible for buyers and sellers to meet in many markets. The result is that transactions that would have been mutually beneficial at prevailing prices simply could not occur, generating a deadweight loss that is fundamentally different from the price-driven welfare losses analysed in standard competitive market failure models.
The supply-demand spiral is a high-value analytical framework. In essays or short answers, walk through the stages clearly: (1) initial supply shock (AS shifts left); (2) demand effects emerge (AD shifts left); (3) demand contraction feeds back into supply via firm insolvencies (AS shifts further left); (4) further demand feedback via job losses and income effects. Conclude by noting that COVID-19 is distinct from prior crises in that BOTH sides of the economy deteriorated simultaneously and interactively.
Expert Opinion on the Balance of Supply and Demand
An IGM (Initiative on Global Markets) poll of leading economists — from both the European and US IGM panels — was conducted on the question of whether the demand effects of COVID-19 would dominate the supply effects. The statement polled was: "The economic effects of COVID-19 coming from reduced spending will be larger than those coming from disruptions to supply chains and illness-related workforce reductions." Among the European panel, 47 per cent agreed or strongly agreed, while 41 per cent expressed uncertainty. Among the US panel, 44 per cent agreed, while 52 per cent were uncertain. A very small minority disagreed. This expert consensus — weighted toward agreement but with substantial residual uncertainty — is consistent with the theoretical analysis above, which identifies demand contraction (through precautionary savings, income loss, and uncertainty) as the primary driver of the output decline, even though supply disruptions were the initial trigger.
Expert surveys suggest that demand effects were expected to dominate supply effects over the course of the COVID-19 recession, consistent with the empirical literature on financial crises and uncertainty shocks.
- COVID-19 began as a supply shock (labour supply reduction, supply chain disruption) but rapidly generated simultaneous demand contraction (uncertainty, income loss, precautionary savings, investment postponement).
- Supply and demand deterioration interacted via two feedback loops: (a) demand collapse causing firm insolvencies and supply destruction; (b) job losses reducing household income and further depressing demand.
- The pandemic is best understood as a shock to market frictionlessness rather than a conventional shock to fundamentals.
- The cumulative welfare loss — the "red wedge" — represents surplus destroyed by the spiral of mutually reinforcing contractions.
- Expert opinion, as captured by the IGM poll, broadly favoured the view that demand effects would dominate supply effects in magnitude, though with significant uncertainty.
Bibliography
Acemoglu, D., Carvalho, V.M., Ozdaglar, A. and Tahbaz-Salehi, A. (2012) 'The network origins of aggregate fluctuations', Econometrica, 80(5), pp. 1977–2016.
Ahir, H., Bloom, N. and Furceri, D. (2018) 'World Uncertainty Index', Stanford mimeo. Available at: https://worlduncertaintyindex.com (Accessed: [date]).
Barrett, P. et al. (2021) 'After-effects of the COVID-19 pandemic: Prospects for medium-term economic damage', Chapter 2 in World Economic Outlook, April 2021. Washington, D.C.: International Monetary Fund.
Brien, P. et al. (2022) 'Economic impact of COVID-19 lockdowns', House of Commons Library Research Briefing. London: House of Commons Library.
Carroll, C.D. (1997) 'Buffer-stock saving and the life cycle/permanent income hypothesis', Quarterly Journal of Economics, 112(1), pp. 1–55.
Carvalho, V.M. et al. (2021) 'Supply chain disruptions: Evidence from the Great East Japan Earthquake', Quarterly Journal of Economics, 136(2), pp. 1255–1321.
Diamond, P.A. (1982) 'Aggregate demand management in search equilibrium', Journal of Political Economy, 90(5), pp. 881–894.
Kimball, M.S. (1990) 'Precautionary saving in the small and in the large', Econometrica, 58(1), pp. 53–73.
McKinsey Global Institute (2020) COVID-19: Implications for Business. McKinsey & Company, 9 March 2020.
Office for National Statistics (ONS) (2022) Monthly GDP, UK: February 2020 to July 2022. Newport: ONS.
Pigou, A.C. (1920) The Economics of Welfare. London: Macmillan.