## CHAPTER 2 THE gREAT LOCKDOWN: DIssECTINg THE ECONOMIC EFFECTs

## Source details

**Canonical URL:** [CHAPTER 2 THE gREAT LOCKDOWN: DIssECTINg THE ECONOMIC EFFECTs](https://www.imf.org/-/media/files/publications/weo/2020/october/english/ch2.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/weo/2020/october/english/ch2.pdf.md)
- [Structured JSON version](/-/media/files/publications/weo/2020/october/english/ch2.pdf.json)

---

### Introduction and overview
- Authors: Francesca Caselli, Francesco Grigoli (co-lead), Weicheng Lian, and Damiano Sandri (co-lead), with support from Jungjin Lee and Xiaohui Sun.
- Purpose:
  - Examine whether the economic contraction during the COVID-19 pandemic was driven primarily by government-imposed lockdowns or by voluntary reductions in social interactions due to fear of infection.
  - Inform expectations for the strength and timing of the recovery: if lockdowns were the main driver, activity could rebound quickly after lifting restrictions; if voluntary social distancing dominated, activity may remain subdued until health risks abate.

### Cross-country associations
- Sample coverage and timing:
  - Up to 52 advanced, emerging market, and developing economies.
  - GDP, consumption, and investment data refer to the first half of 2020 (2020:H1).
  - For monthly indicators, first three months after a country reaches 100 confirmed COVID-19 cases.
- Key empirical finding:
  - Countries that implemented more stringent lockdowns experienced larger declines in GDP relative to pre–COVID-19 forecasts, even after controlling for the severity of the local epidemic.
- Lockdown stringency index:
  - Averages subindicators: 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 (Oxford Coronavirus Government Response Tracker).
- Caveat:
  - Cross-country correlations are suggestive but subject to omitted variable and endogeneity concerns; time-series, high-frequency approaches are used to strengthen identification.

### High-frequency evidence: mobility and job postings
- Data and proxies:
  - Google Community Mobility Reports (national and some subnational coverage).
  - Indeed anonymized daily job postings (22 countries, disaggregated by job categories).
- Mobility regression setup:
  - Local projections with country fixed effects and time dummies; controls include the number of COVID-19 cases and lags of the mobility indicator.
- Sample size:
  - Mobility regressions use national-level data for 128 countries.
- Main results:
  - Lockdowns have a considerable and statistically significant negative effect on mobility.
  - Voluntary social distancing in response to rising COVID-19 infections also has strong negative effects on mobility.
  - Overall, lockdowns and voluntary social distancing played a near comparable role in driving the economic recession.
  - The contribution of voluntary distancing in reducing mobility was stronger in advanced economies (AEs).
- Dynamics and asymmetry:
  - Lockdown tightening can reduce mobility by 25 percent within a week.
  - Mobility begins to resume gradually as the lockdown tightening shock dissipates.
  - A doubling of daily COVID-19 cases leads to a contraction in mobility by about 2 percent.
  - Easing lockdowns tends to increase mobility, but the positive effect is weaker than the effect of tightening lockdowns; this asymmetry implies lifting lockdowns is unlikely to rapidly restore economic activity if health risks remain high.
  - The impact of lockdowns on mobility is smaller when infections are relatively high.

### Lockdowns and job postings (sectoral patterns)
- Both lockdown tightening and increases in COVID-19 cases lead to a statistically significant negative effect on job postings.
- Timing and sectoral differences:
  - Contact-intensive jobs (hospitality, personal care, food) declined before stay-at-home orders—likely due to voluntary social distancing.
  - Manufacturing job postings declined closer to the adoption of stay-at-home orders—reflecting lockdown impacts.
  - Job postings in contact-intensive sectors declined more than in manufacturing.
  - Removal of stay-at-home orders coincided with only a marginal increase in job postings, even in manufacturing.
- Sample composition note:
  - Contribution of voluntary social distancing is relatively higher in the sample, which includes mostly advanced economies.

### Distributional and demographic impacts
- Data source and protections:
  - Vodafone anonymized and aggregated mobility indicators (provincial level, including at least 50 customers; data sharing protocol included ethical assessment and privacy safeguards).
- Gender and childcare:
  - Stay-at-home orders for people aged 25 to 44 coincided with a drop of about 20 percent in the number of people who leave their homes on a given day.
  - The effect on women was stronger by about 2 percent (a modest but statistically significant difference).
  - The higher reduction in women’s mobility may reflect childcare responsibilities because stay-at-home orders coincided with school closures.
  - Regions with school closures before national lockdowns showed the gender gap widening at the time of school closures.
- Age-group effects:
  - Adoption of stay-at-home orders led to substantial mobility reductions across all age categories, with considerably stronger effects for younger cohorts.
  - Drops were particularly large for ages 18–24 and 25–44.
  - The impact was substantially weaker for ages 65+, whose mobility was already lower before stay-at-home orders.
- Policy implication:
  - Targeted interventions are needed to protect employment prospects of women and younger cohorts (for example, parental leave, strengthening unemployment benefits) to prevent widening income and intergenerational inequality.

### Lockdowns and infection dynamics
- Timing and effectiveness:
  - Lockdowns can reduce confirmed COVID-19 cases substantially, with effects tending to materialize after a few weeks of delay (accounting for incubation period and testing times).
  - A stringent lockdown leads to a reduction in cumulated infections of about 40 percent after 30 days.
  - Lockdowns are more effective if introduced early in the epidemic and must be sufficiently stringent to reduce infections significantly.
- Nonlinearities and marginal effects:
  - The introduction of additional lockdown measures has a weaker marginal impact on mobility once other measures are already in place.
  - Conversely, lockdowns become progressively more effective in reducing COVID-19 cases when they become sufficiently stringent; mild lockdowns appear ineffective in curbing infections.
  - Definitions used:
    - Low and high stringency refer to the 25th and 75th percentile of lockdown stringency.
    - A lockdown tightening corresponds to an increase in the index by 100 units.
- Policy interpretation:
  - To achieve a given reduction in infections, policymakers may prefer stringent lockdowns over a shorter period rather than protracted mild lockdowns, because tighter lockdowns appear to entail only modest additional economic costs while leading to a considerably stronger decline in infections.
  - Recommendations should be reexamined as non-lockdown interventions (contact tracing, face masks) expand; if these succeed, mild lockdowns could suffice for localized flare-ups.

### Policy recommendations and implications
- Be cautious about removing policy support too quickly given persistent voluntary social distancing and weaker mobility responses to easing restrictions.
- Continue to protect the most vulnerable through social safety net spending.
- Support economic activity consistent with persistent social distancing by:
  - Reducing contact intensity and making workplaces safer (for example, promoting contactless payments).
  - Facilitating gradual reallocation of resources toward less-contact-intensive sectors.
  - Enhancing work-from-home capacity (for example, improving internet connectivity and supporting information technology investment).
- Pursue and scale up alternative public-health tools to contain infections with potentially lower economic costs:
  - Expand testing and contact tracing.
  - Promote face-mask use.
  - Encourage work from home.
  - Deploy targeted measures (for example, protecting vulnerable people and restricting large indoor gatherings).

### Evidence on information technology as a mitigant (Box 2.2 and US case)
- Main message:
  - Information technology adoption may, in the aggregate, significantly shield labor markets against the effects of the coronavirus pandemic.
- Specific empirical findings:
  - The increase in the probability of being unemployed associated with a large drop in mobility (one standard deviation, equal to 10 percentage points) is 25 percent larger in metropolitan statistical areas with low levels of information technology adoption than in those with high levels (5 percentage points versus 4 percentage points).
  - Information technology cushions the unemployment impact of mobility across male and female and across white and nonwhite workers, but not uniformly across all education groups.
- Distributional implication:
  - IT adoption may contribute to widening inequality between individuals with high and low levels of educational attainment.
- Policy-relevant implications:
  - Strengthen IT adoption to increase economic resilience to mobility-reducing shocks.
  - Design policies to ensure IT-mediated resilience does not exacerbate educational inequality (targeted training and access programs, inclusive digital infrastructure).

### Caveats and areas for future research
- Identification concerns cannot be fully dismissed, including regarding the measurement of voluntary social distancing.
- The analysis relies on short-term indicators (mobility, job postings) that provide imperfect measures of economic activity; findings need reexamination as more conventional economic indicators become available.
- The chapter does not quantify important side effects (for example, on educational attainment and mental health issues), which are crucial areas for future research.
- A key research priority is comparing effectiveness of more-targeted instruments (for example, restrictions on dense indoor gatherings, isolating vulnerable people) with blunt lockdowns.

*Chapter 2, “Introduction,” World Economic Outlook, October 2020 — IMF staff analysis*

### Introduction

### Introduction

### Overview and purpose
- Authors: Francesca Caselli, Francesco Grigoli (co-lead), Weicheng Lian, and Damiano Sandri (co-lead), with support from Jungjin Lee and Xiaohui Sun.
- The chapter examines whether the economic contraction during the COVID-19 pandemic was driven primarily by government-imposed lockdowns or by voluntary reductions in social interactions due to fear of infection.
- The analysis informs expectations for the strength and timing of the recovery: if lockdowns were the main driver, activity could rebound quickly after lifting restrictions; if voluntary social distancing dominated, activity may remain subdued until health risks abate.

### Cross-country associations
- Sample coverage and timing:
  - Cross-country analysis across a broad sample of up to 52 advanced, emerging market, and developing economies.
  - GDP, consumption, and investment data refer to the first half of 2020 (2020:H1).
  - For monthly indicators, the analysis considers the first three months after a country reaches 100 confirmed COVID-19 cases.
- Key empirical finding:
  - Countries that implemented more stringent lockdowns experienced larger declines in GDP relative to pre–COVID-19 forecasts, even after controlling for the severity of the local epidemic.
- Data and index construction:
  - Lockdown stringency index averages subindicators: 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 (Oxford Coronavirus Government Response Tracker).
- Caveats:
  - Cross-country correlations are suggestive but subject to omitted variable and endogeneity concerns (policy choices are not random; time-invariant country characteristics may confound results). Time-series, high-frequency approaches are used next to strengthen identification.

### High-frequency evidence: mobility and job postings
- High-frequency proxies used:
  - Google Community Mobility Reports (attendance rates at various locations relative to precrisis levels; national and, for some countries, subnational coverage).
  - Indeed anonymized daily job postings (provided for 22 countries, disaggregated by job categories).
- Mobility regression setup:
  - Local projections with country fixed effects and time dummies; controls include the number of COVID-19 cases and lags of the mobility indicator to mitigate endogeneity.
- Sample sizes:
  - Mobility regressions use national-level data for 128 countries.
- Main results:
  - Lockdowns have a considerable and statistically significant negative effect on mobility.
  - Voluntary social distancing in response to rising COVID-19 infections also has strong negative effects on mobility.
  - Overall, the analysis suggests lockdowns and voluntary social distancing played a near comparable role in driving the economic recession.
  - The contribution of voluntary distancing in reducing mobility was stronger in advanced economies (AEs), where people can work from home more easily and sustain temporary unemployment because of personal savings and government benefits.
- Dynamics and asymmetry:
  - Easing lockdowns tends to increase mobility, but the positive effect is weaker than the effect of tightening lockdowns.
  - This asymmetry implies that lifting lockdowns is unlikely to rapidly restore economic activity if health risks remain high.

### Distributional and demographic impacts (Vodafone and other mobility data)
- Data source and protection:
  - Vodafone anonymized and aggregated mobility indicators (provincial level, including at least 50 customers; data sharing protocol included ethical assessment and privacy safeguards).
- Key heterogenous effects:
  - Lockdowns tend to have a larger effect on women’s mobility than on men’s, especially at the time of school closures—suggesting disproportionate child-care burdens for women that may jeopardize employment opportunities.
  - Lockdowns tend to have a stronger impact on the mobility of younger cohorts, who are more economically vulnerable because they generally rely on labor income and have less stable jobs.
- Policy implication:
  - Targeted policy intervention is needed to protect the employment prospects of women and younger cohorts and to prevent a widening of income inequality.

### Lockdowns and infection dynamics
- Timing and effectiveness:
  - Lockdowns can reduce confirmed COVID-19 cases substantially, with effects tending to materialize after a few weeks of delay (accounting for incubation period and testing times).
  - Lockdowns are more effective at curbing infections if introduced early in a country’s epidemic.
  - Lockdowns must be sufficiently stringent to reduce infections significantly.
- Trade-offs and complementarities:
  - Effective lockdowns that contain the epidemic may reduce voluntary social distancing and thus pave the way for a faster economic recovery—short-term costs could be offset by stronger medium-term growth.
  - Policymakers should pursue alternative containment measures that may involve lower short-term economic costs than lockdowns, including expanding testing and contact tracing, promoting face-mask use, encouraging work from home, and deploying targeted measures (for example, protecting vulnerable people and restricting large indoor gatherings).

### Policy recommendations and implications
- Be cautious about removing policy support too quickly given persistent voluntary social distancing and weaker mobility responses to easing restrictions.
- Consider measures to support economic activity consistent with social distancing, including:
  - Measures to reduce contact intensity and make workplaces safer (for example, promoting contactless payments).
  - Facilitating gradual reallocation of resources toward less-contact-intensive sectors.
  - Enhancing work-from-home capacity (for example, improving internet connectivity and supporting information technology investment).
- Pursue and scale up alternative public-health tools to contain infections with potentially lower economic costs (testing, contact tracing, face masks), and, as understanding of transmission improves, favor more targeted measures instead of blunt lockdowns.

_Chapter 2, “Introduction,” World Economic Outlook, October 2020 — IMF staff analysis_

### CHAPTER 2 THE gREAT LOCKDOWN: DIssECTINg THE ECONOMIC EFFECTs

### CHAPTER 2 THE gREAT LOCKDOWN: DIssECTINg THE ECONOMIC EFFECTs

### Impact of Lockdowns and Mobility
- Lockdown tightening can reduce mobility by 25 percent within a week.
- Mobility begins to resume gradually as the lockdown tightening shock dissipates.
- A doubling of daily COVID-19 cases leads to a contraction in mobility by about 2 percent.
- Lockdowns have a strong negative impact on mobility; results are robust to controlling for COVID-19 deaths, subindicators of mobility, testing, contact tracing, public information campaigns, and cross-country heterogeneity.

### Voluntary Social Distancing versus Government Mandates
- Voluntary social distancing in response to rising infections plays a large role in reducing mobility and can be roughly as important as lockdowns in emerging markets.
- The contribution of voluntary social distancing is smaller in low-income countries and larger in advanced economies, reflecting differences in the ability to work from home and rely on savings or social security benefits.
- Because voluntary social distancing is substantial, lifting lockdowns may produce only a partial rebound in economic activity if health risks persist.
- The impact of lockdowns on mobility is smaller when infections are relatively high.
- Easing lockdowns tends to have a positive effect on mobility, but the magnitude is weaker compared with the impact from a lockdown tightening; this difference is statistically significant.

### Policy Implications and Recommendations
- Policymakers should be wary of removing policy support too hastily to avoid precipitating a further downturn and should continue to protect the most vulnerable through social safety net spending.
- Support economic activity consistent with persistent social distancing by:
  - Reducing contact intensity and making workplaces safer (for example, promoting contactless payments).
  - Facilitating reallocation of resources toward less-contact-intensive sectors.
  - Enhancing working from home by improving internet access and supporting firm investment in information technology.
- Box 2.2 indicates that supporting firm IT investment can protect employment during the pandemic.

### Lockdowns and Job Postings
- Both lockdown tightening and increases in COVID-19 cases lead to a statistically significant negative effect on job postings.
- Lockdowns and voluntary social distancing played an important role in driving the reduction in job postings during the first three months of each country’s epidemic; the contribution of voluntary social distancing is relatively higher in the sample, which includes mostly advanced economies.
- Sectoral patterns:
  - Contact-intensive jobs (hospitality, personal care, food) declined before stay-at-home orders, likely due to voluntary social distancing.
  - Manufacturing job postings declined closer to the adoption of stay-at-home orders, reflecting lockdown impacts.
  - Job postings in contact-intensive sectors declined more than in manufacturing, reflecting a larger drop in aggregate demand from voluntary social distancing.
  - Removal of stay-at-home orders coincided with only a marginal increase in job postings, even in manufacturing.

### Unequal Effects across Gender and Age
- Stay-at-home orders for people aged 25 to 44 coincided with a drop of about 20 percent in the number of people who leave their homes on a given day.
- The effect on women was stronger by about 2 percent (a modest but statistically significant difference).
  - The higher reduction in women’s mobility may reflect childcare responsibilities because stay-at-home orders coincided with school closures.
  - Regions with school closures before national lockdowns showed the gender gap widening at the time of school closures.
- Age-group effects:
  - Adoption of stay-at-home orders led to substantial mobility reductions across all age categories, with considerably stronger effects for younger cohorts.
  - Drops were particularly large for ages 18–24 and 25–44.
  - The impact was substantially weaker for ages 65+, whose mobility was already lower before stay-at-home orders.
- These patterns suggest lockdowns disproportionately affect younger workers and women, calling for targeted policy interventions (for example, parental leave) to avoid long-lasting effects on employment opportunities.

### Lockdowns and COVID-19 Infections
- Lockdowns tend to have a negative impact on infections:
  - A stringent lockdown leads to a reduction in cumulated infections of about 40 percent after 30 days.
- The effects of lockdowns on confirmed COVID-19 cases tend to materialize after at least two weeks, consistent with the COVID-19 incubation period and testing delays.
- Early implementation matters:
  - Countries that imposed lockdowns faster after their first case experienced better epidemiological outcomes.
  - Differences are more pronounced when countries are categorized by the number of COVID-19 cases at the time of lockdowns.

*International Monetary Fund | October 2020 — CHAPTER 2 THE gREAT LOCKDOWN: DIssECTINg THE ECONOMIC EFFECTs*

### 1. Response of Infections to a Full Lockdown

### Response of Infections to a Full Lockdown

### Lockdown timing and infections
- Countries that adopted lockdowns when COVID-19 cases were still low witnessed considerably fewer infections during the first three months of the epidemic compared with countries that introduced lockdowns when cases were already high.
- Lockdowns are powerful instruments to reduce infections, especially when they are introduced early in a country’s epidemic and when they are sufficiently stringent.

### Trade-offs and dynamic economic effects
- The common narrative that lockdowns involve a trade-off between saving lives and protecting livelihoods should be reconsidered given evidence that rising infections also have severe detrimental effects on economic activity.
- By bringing infections under control, lockdowns may pave the way to a faster economic recovery as people feel more comfortable resuming normal activities; therefore, the short-term economic costs of lockdowns could be compensated through higher future economic activity, possibly even leading to positive net effects on the economy.
- These medium-term gains remain an important area for future research as more data become available.

### Individual lockdown measures and nonlinear effects
- Lockdown stringency index combines multiple measures (travel restrictions, school and workplace closures, stay-at-home orders, etc.). These measures are highly correlated and often introduced in rapid succession, meaning empirical estimates capture marginal impacts conditional on measures already in place.
- The analysis uses quadratic terms of the lockdown index to examine whether further tightening continues to have similar effects.

Key empirical patterns:
- The introduction of additional lockdown measures has a weaker marginal impact on mobility once other measures are already in place—that is, when the lockdown stringency index is already relatively high. This implies lockdowns have marginally weaker negative economic effects as they become more stringent.
- Conversely, lockdowns become progressively more effective in reducing COVID-19 cases when they become sufficiently stringent; mild lockdowns appear ineffective in curbing infections.
- Interpretation: Preventing only a few personal contacts (for example, closing schools alone) is often not enough to reduce community spread significantly; additional measures such as workplace closures or stay-at-home orders are needed to bring the virus under control.

Empirical notes and definitions used in the analysis:
- Low and high stringency in panels 2 and 3 refer to the 25th and 75th percentile of lockdown stringency.
- The shaded areas in panels 2 and 3 correspond to 90 percent confidence intervals computed with standard errors clustered at the country level.
- A lockdown tightening corresponds to an increase in the index by 100 units.

Policy implication derived from nonlinearities:
- To achieve a given reduction in infections, policymakers may want to opt for stringent lockdowns over a shorter period rather than protracted mild lockdowns, because tighter lockdowns appear to entail only modest additional economic costs while leading to a considerably stronger decline in infections.
- These recommendations should be reexamined as the pandemic progresses and as non-lockdown interventions (contact tracing, face masks) expand; if these succeed, mild lockdowns could be sufficient for localized flare-ups.

### Voluntary social distancing and economic persistence
- Voluntary social distancing in response to rising infections has severe detrimental effects on the economy and contributes importantly to the recession.
- The contribution of voluntary social distancing in reducing mobility is particularly high in advanced economies, where people can more easily stay at home due to teleworking arrangements, higher personal savings, and more generous social security benefits.
- Lifting lockdowns tends to have a more modest impact on mobility compared with the impact of a lockdown tightening; lifting lockdowns prematurely when infections remain high can lead to weaker mobility rebounds because people’s decisions are driven by fear of contracting the virus.
- As long as significant health risks persist, economic activity is likely to remain subdued; policymakers should refrain from withdrawing policy support too quickly and preserve spending on social safety nets.

Recommended policy measures consistent with persistent social distancing:
- Encourage work from home.
- Facilitate reallocation of resources toward less-contact-intensive sectors.
- Promote adoption of new technologies to limit contact intensity within given sectors.
- Expand testing and contact tracing, promote the use of face masks, and encourage working from home as alternative ways to contain infections with potentially lower economic costs.

### Distributional impacts and targeted policies
- Lockdowns tend to severely affect economically vulnerable segments of the population:
  - Mobility data for some European countries show lockdown measures—especially school closures—generate a larger drop in women’s mobility, likely reflecting women’s disproportionate role in childcare and jeopardizing their employment opportunities.
  - Lockdowns generate a sharper reduction in mobility of younger cohorts; younger workers rely on labor income and often have temporary job contracts at greater risk of termination.
- Targeted policy interventions are needed, such as strengthening unemployment benefits for vulnerable categories and supporting paid leave for parents, to prevent widening gender and intergenerational inequality.

### Evidence on information technology as a mitigant (US case)
- Firms’ adoption of information technology can dampen the economic impact of lockdowns and voluntary social distancing by facilitating teleworking, promoting online sales, or organizing contactless delivery.
- Empirical findings:
  - The increase in the probability of being unemployed associated with a large drop in mobility (one standard deviation, equal to 10 percentage points) is 25 percent larger in metropolitan statistical areas with low levels of information technology adoption than in those with high levels (5 percentage points versus 4 percentage points).
  - Information technology cushions the unemployment impact of mobility across male and female and across white and nonwhite workers, but not uniformly across all education groups (see the original analysis for subgroup coefficients).

### Caveats and areas for future research
- Identification concerns cannot be fully dismissed, including regarding the measurement of voluntary social distancing.
- The analysis relies on short-term indicators (mobility, job postings) that provide imperfect measures of economic activity; findings need reexamination as more conventional economic indicators become available.
- The chapter focuses on economic consequences of lockdowns and does not quantify important side effects (for example, on educational attainment and mental health issues), which are crucial areas for future research.
- A crucial area of research is comparing the effectiveness of more-targeted instruments (for example, restrictions on dense indoor gatherings, isolating vulnerable people) with blunt lockdowns.

*Source: IMF staff calculations and chapter text from the provided PDF content.*

### 3. Dampening Effects of IT on

### 3. Dampening Effects of IT on Unemployment, by Worker Type

### Unemployment and Lockdowns in the United States
- The chapter analyzes the relationship between lockdown measures and unemployment in the United States, highlighting heterogeneity across worker types.
- Figures referenced: "Unemployment and Lockdowns in the United States" and "Unemployment and Mobility in the United States" illustrate links between unemployment rate increase and mobility drop (figure axes labeled "Unemployment rate increase" and "Mobility drop").

### The Role of Information Technology Adoption during the COVID-19 Pandemic (Box 2.2)
- Information technology (IT) adoption is examined as a mitigating factor for labor market impacts during the coronavirus pandemic.
- Key finding:
  - "Information technology adoption may, in the aggregate, significantly shield labor markets against the effects of the coronavirus pandemic."
- Distributional implication:
  - "It may also contribute to widening inequality between individuals with high and low levels of educational attainment."
  - IT adoption particularly mitigates impacts for individuals who have a low level of education, but overall dynamics can widen education-based inequality.

### Empirical and Literature Context
- The box situates its conclusions within a broad literature on COVID-19, lockdowns, mobility, labor markets, and technology adoption (References list numerous working papers and studies addressing: lockdown effectiveness, labor market outcomes, social distancing impacts, the ability to work from home, and IT shields).
- Representative topics in the referenced literature include:
  - Multi-risk SIR models with targeted lockdowns.
  - Socio-economic network heterogeneity and pandemic policy responses.
  - Economic and health impacts of social distancing policies.
  - The future of working from home; how many jobs can be done at home.
  - IT Shields: Technology Adoption and Economic Resilience during the COVID-19 Pandemic.

### Policy-relevant implications (derived from the box's findings)
- Strengthen IT adoption to increase economic resilience to mobility-reducing shocks.
- Design policies to ensure IT-mediated resilience does not exacerbate educational inequality:
  - Target training and access programs for workers with low levels of educational attainment.
  - Support inclusive digital infrastructure and labor-market policies that facilitate equitable access to remote-work opportunities.

*Source: IMF — World Economic Outlook: A Long and Difficult Ascent, Chapter 2 (Box 2.2), October 2020.*

---


_Source: https://www.imf.org/-/media/files/publications/weo/2020/october/english/ch2.pdf_