Abstract
The Zero-COVID strategy served China well in the initial stages of the pandemic. The emergence of more transmissible but potentially less severe COVID variants and a significant rise in economic cost associated with frequent lockdowns have pointed to the importance of recalibrating the COVID strategy while balancing the still significant health risks. As a result, in late 2022, China started easing COVID restrictions. Experiences from other economies suggest that a prepared and orderly full reopening would help minimize the tradeoff between health consequences and economic costs.
Recalibrating the Covid Strategy1
The Zero-COVID strategy served China well in the initial stages of the pandemic. The emergence of more transmissible but potentially less severe COVID variants and a significant rise in economic cost associated with frequent lockdowns have pointed to the importance of recalibrating the COVID strategy while balancing the still significant health risks. As a result, in late 2022, China started easing COVID restrictions. Experiences from other economies suggest that a prepared and orderly full reopening would help minimize the tradeoff between health consequences and economic costs.
A. Introduction
1. Throughout much of the COVID-19 pandemic, China has relied on a COVID suppression strategy—Zero-COVID strategy (ZCS), for short. The strategy’s goal of suppressing transmission as much as possible has been achieved by extensive reliance on non-pharmaceutical interventions such as rapid and sweeping lockdowns, mass testing, contact tracing, quarantine at government facilities for anyone infected or a close contact of an infected person, and largely closed international and sometimes domestic borders.
2. The ZCS served China well in the initial stages of the pandemic. By quickly containing outbreaks and moving scarce health care resources to where they were most needed, the strategy helped keep Covid cases, hospitalizations, and death rates very low by international comparison despite limited medical resources.2
3. Successfully containing outbreaks also supported the gradual transition to more targeted and flexible containment measures, refining the overall strategy in line with the evolution of the pandemic (Text Figure). After the emergency response to the Wuhan outbreak in early 2020, the zero-COVID policy started taking shape in the spring of 2020. Measures focused on preventing imported cases and containing any rebound of domestic outbreaks by suppressing them at the local level. As the Delta variant started to spread in mid-2021, the COVID policy shifted its focus to more rapid and targeted measures to avoid community spread while recognizing that cases would occur regularly (“dynamic clearing”). In March 2022, amid the Omicron variant and large outbreaks, China’s COVID policy started emphasizing comprehensiveness and strict responses to cut off transmission channels over rapidity.
4. In late-2022, the authorities started easing COVID containment policies notwithstanding ongoing spikes in infections. In November 2022, the COVID policy was significantly relaxed for the first time (“20 measures”) by easing and lifting excessive restrictions imposed previously and those observed in implementation. Further meaningful easing of restrictions has since taken place, including allowing home quarantine, reducing mass testing, eliminating testing requirements to enter public places and transportation, and lifting of requirements for health codes for inter-regional travel in China. The discussion of public health preparations for an eventual full reopening, including targets for elderly vaccinations, has been elevated. At the same time, cases have surpassed earlier waves, suggesting a move away from the zero-COVID strategy towards accepting outbreaks.
Evolution of Zero-COVID Policy
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Evolution of Zero-COVID Policy
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Evolution of Zero-COVID Policy
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
5. Even though the economy bounced back quickly from the 2020 outbreaks, the recovery of private consumption has been lagging—in part as a result of the need for recurring lockdowns and elevated uncertainty. Real GDP rebounded to its pre-pandemic trend by end-2020 but has since lost ground again. While overall cumulative output losses—measured as the percentage share between actual outturns and the January 2020 WEO forecast—amounted to only 1.5 percent of GDP over 2020 and 2021, this performance was predominantly driven by surging exports and investment. Amid significant disruptions throughout 2022, these cumulative losses are estimated at more than 41/2 percent by end-2022. Over 2020 and 2021, private consumption losses amounted to more than 5 percent, and are estimated to have increased to almost 13 percent by end-2022. Supply of services was almost 4 percent below its pre-crisis trend in 2020-21 and at more than 8 percent below its trend by Q3-2022. Consumption was impacted not only by income losses, but also by increased uncertainty under the ZCS. These developments have reversed progress towards a more consumption-based, higher-quality growth model, with the private consumption-to-GDP share falling to around 35 percent from its pre-pandemic share of around 37.
6. Throughout 2022, the emergence of more transmissible variants put the zero-COVID strategy to test and outbreaks have significantly hurt domestic economic activity. Amid these more transmissible variants, China experienced the largest and most persistent outbreaks since the original Wuhan outbreak in the first half of 2022. Under the ZCS, these outbreaks led to widespread and very strict containment measures that severely disrupted domestic economic activity, created significant challenges for the thus-far relatively resilient supply chains, and weakened employment and private consumption. Ever since, China has not returned to extended periods of close to zero confirmed cases—as was the case before the arrival of the Omicron variant—leaving the economy facing frequent, though mostly local, outbreaks that had to be addressed with oftentimes economically painful containment measures under the ZCS (Text Figure), even as the strategy continued to be successful in slowing contagion. Since October 2022, new outbreaks have become larger and more widespread than ever before and have surpassed total daily case levels from earlier in the year—even before testing was rolled back and fewer cases recorded—estimated to significantly hurt economic activity in the near-term even as restrictions have been rolled back.
Cases and Government Response Stringency
(Index from 0–100, 100 = most strict, number of cases)
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Sources: Haver Analytics; CEIC Data Company Limited; Oxford University Stringency Index; and IMF staff calculations.Note: The stringency index shows the aggregated GDP-weighted provincial stringency indices as published by Oxford University.Cases and Government Response Stringency
(Index from 0–100, 100 = most strict, number of cases)
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Sources: Haver Analytics; CEIC Data Company Limited; Oxford University Stringency Index; and IMF staff calculations.Note: The stringency index shows the aggregated GDP-weighted provincial stringency indices as published by Oxford University.Cases and Government Response Stringency
(Index from 0–100, 100 = most strict, number of cases)
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Sources: Haver Analytics; CEIC Data Company Limited; Oxford University Stringency Index; and IMF staff calculations.Note: The stringency index shows the aggregated GDP-weighted provincial stringency indices as published by Oxford University.7. Going forward, a full lifting of COVID containment measures should balance economic costs against still significant healthcare risks. Vaccines remain effective against severe outcomes and should help mitigate health-related risks. The full lifting of the remaining restrictions should thus be preceded by urgent measures to contain the health impact. Specifically, increased vaccination coverage and better treatment options could mitigate some of the difficult tradeoffs between health concerns and economic growth under a durable recovery.
B. Experiences from Other Countries
8. Several other economies have in the meantime lifted their zero-COVID regimes, offering broad lessons for China. Examples of countries removing ZCS-style COVID policies include Australia, New Zealand, Singapore, and Vietnam (see Text Figure). While each has charted a distinct exit path, most did so in the face of rising economic costs and more contagious, less-severe new variants. Some commonalities include the following:
Trigger: Increasing economic costs under less lethal but more transmissible variants featured among the main reasons to abandon the strategy.
Communication: In most cases, the eventual lifting of the ZCS was communicated well in advance, sometimes linked to vaccination goals, providing time for society to internalize the upcoming change.
Vaccinations: For all former ZCS economies, the removal of some or all non-pharmaceutical restrictions has been preceded or accompanied by ongoing pharmaceutical interventions, in particular very high vaccination rates, including of boosters, especially for the elderly, as well as availability of treatment options and enhanced health care capacity.
Shifting goals: Lifting zero-COVID implied shifting the main goal from trying to keep cases close to zero to accepting outbreaks, though the acceptable level of cases might depend on factors such as health care capacity and the degree of protection afforded by the existing level of vaccination.
From strict lockdowns to more targeted measures: Lifting zero-COVID has not implied an immediate end to all restrictions. Rather, countries ended hard lockdowns before sequentially lifting other restrictions. In some cases, voluntary social distancing by the population have complemented official rules.
Flexibility: Measures have been focused on supporting a reduction in the intensity of outbreaks instead of full prevention. In some cases, restrictions have been reimposed and eventually re-loosened as outbreaks worsened and improved. In several economies, some of these restrictions have been implemented differently across regions, depending on local conditions.
Transition measures: Transition measures have included continued safe distancing, mask wearing, quarantines for infected or close contacts (moving from mandatory government facilities to home quarantine), testing (including rapid tests), and group size or capacity restrictions. Contact tracing has been eased and eventually abandoned over time in several economies.
International reopening: Accepting a certain level of domestic cases provided an important prerequisite for opening international borders, as the number of imported cases has often been small compared to local transmission.
Health consequences: Opening up has allowed the virus to spread domestically and large waves of outbreaks have occurred, with high vaccination rates lowering the severity of cases but not preventing transmission, especially as there was very limited natural immunity. Several of the former-ZCS economies reached population infection levels close to levels observed in other countries around the world in a matter of a few months (though, of course, estimates of case numbers can vary considerably depending on the degree of testing). With high vaccination levels, healthcare systems have not been overburdened and death rates have been kept at relatively low levels.
Economic impact: Amid less lethal variants, high vaccination rates (reducing the severity of health impacts), a well-prepared population (avoiding resorting to voluntary restrictions which could have significant economic impacts) and an oftentimes gradual versus sudden opening up, short-term economic costs appear to be mostly contained, and consumption recovered despite outbreaks.
C. The Economic Cost of the ZCS
9. Overdependence on non-pharmaceutical measures has an economic cost. Since the onset of the COVID-19 pandemic in early 2020, China quickly introduced strict non-pharmaceutical interventions (NPIs) such as large-scale lockdowns, contact tracing and mass testing. The measures formed an essential part of the policy package at the early stages of the pandemic when there were limited medical options. As the COVID virus mutated into more contagious, less-severe variants, the existing mitigation policies have been adjusted, primarily aiming at more rapid and targeted responses. However, the impact of non-pharmaceutical measures on economic recovery had been amplified by mobility restrictions taking place at a higher frequency and broader regions, tilting the balance of risk toward higher economic losses as the virus became more contagious and less lethal.
10. A number of stylized facts characterize the economic impact of the ZCS. We apply a panel regression on the congestion data from 95 major cities and a range of economic indicators for matching prefectures from January 2020 to July 2022, estimating elasticities for different levels of containment stringencies, virus variants, mobility, and economic activity.
The ZCS approach evolved as the virus mutated. Table 3 (see Appendix I) suggests that despite the strong average response by the local authorities to rises in cases and fatalities, tightness of the policy measures increased, on average, for the more infectious Delta variant and even more for the Omicron wave.
Voluntary mobility increased during the Omicron waves in the first half of 2022 in contrast to earlier waves. Table 3 also suggests that average voluntary mobility decreased in response to rising cases and deaths. However, voluntary responses reacted negatively to the Delta variant, which was more infectious and lethal than the original Alpha variant. In contrast, voluntary activity increased amid the Omicron variant, which was even more infectious than the Delta, but less lethal than the original Alpha variant.3 The shift in behavior likely reflects a combination of factors: an evolution of public confidence on the protection provided by the vaccines, a change in public perception of the virulence of the Omicron variant compared to earlier mutations, and potentially fatigue of repeated and prolonged lockdown measures.
The reduction in mobility has had a significant impact on economic cost. Table 1 (Appendix I) suggests that, on average, a one percent drop in mobility led to a 2/3 percentage point decline in GDP over the past years since 2020. Due to the continuous adjustment to the NPIs since the initial wave in Hubei in 2020, the impact on GDP during the Delta wave in the second half of 2021 was on average about 1.3 percentage point smaller. Consumption was most heavily affected. With every percentage drop in mobility, retail sales were 1.6 percentage points lower compared to the pre-COVID trend. The impact on consumption was amplified by the subsequent Delta and Omicron waves, with further declines of 3.8 and 7.1 percentage points, respectively, due to the tightening of NPI measures, in line with Stage 3 and 4 of the Zero-COVID policy (Text Figure). Household disposable income was affected to a lesser degree compared to GDP and retail sales, with every one percent drop in mobility leading to 0.4 percent drop in disposable income per capita. The relatively high elasticity between retail sales and voluntary mobility may reflect the combined impact of a reduction in income and a rise in savings in light of elevated uncertainty. However, the Delta and Omicron waves posed additional 3.0 and 4.5 percent contraction in disposable income respectively.
Uncertainty could have an additional dampening economic effect. While not directly captured in our estimates, uncertainty such as the one deriving from fear of lockdowns could provide an additional negative channel impacting economic indicators, especially consumption.
Experience from Other Countries
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Note: For ‘lifting of ZCS’, we assume the timing of the most significant lifting of domestic restrictions even if some residual measures remained in place at that time, based on news reports. The underlying timline is the following. The announcement was made in August 2021 in Australia, October 2021 in New Zealand, June 2021 in Singapore, and October 2021 in Vietnam. For the eventual lifting of restrictions, we assume November 2021 in Australia, September 2022 in New Zealand, April 2022 in Singapore, and May 2022 in Vietnam.Experience from Other Countries
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Note: For ‘lifting of ZCS’, we assume the timing of the most significant lifting of domestic restrictions even if some residual measures remained in place at that time, based on news reports. The underlying timline is the following. The announcement was made in August 2021 in Australia, October 2021 in New Zealand, June 2021 in Singapore, and October 2021 in Vietnam. For the eventual lifting of restrictions, we assume November 2021 in Australia, September 2022 in New Zealand, April 2022 in Singapore, and May 2022 in Vietnam.Experience from Other Countries
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Note: For ‘lifting of ZCS’, we assume the timing of the most significant lifting of domestic restrictions even if some residual measures remained in place at that time, based on news reports. The underlying timline is the following. The announcement was made in August 2021 in Australia, October 2021 in New Zealand, June 2021 in Singapore, and October 2021 in Vietnam. For the eventual lifting of restrictions, we assume November 2021 in Australia, September 2022 in New Zealand, April 2022 in Singapore, and May 2022 in Vietnam.D. The Health Cost of a Sudden Lifting of All Containment Measures
11. An immediate withdrawal of all non-pharmaceutical interventions could have severe health consequences, given the relatively lower level of vaccination for the most vulnerable, a possible lack of antiviral treatments, and still limited medical capacities. The Omicron variant is more infectious and immune evasive than the previous variants, which could pose serious challenges to China’s health care system and existing capacity of hospital beds, especially in intensive care. Cai and others (2022) studied possible outcomes should NPIs be absent in the pandemic prevention and control framework. By maintaining the pace of vaccination as of March 1, 2022, a hypothetical 20 imported Omicron cases in the baseline could lead to a wave of 112 million symptomatic cases, 2.7 million ICU admissions, and 1.6 million deaths over a 6-month period, with three quarters of the deaths among unvaccinated individuals aged 60 and above. This compares to approximately 88 million people aged above 60 who have not yet been boosted and 41 million people aged above 60 who are not fully vaccinated as of July 23, 2022 (or around 23 million people above 80 who have not yet been boosted and 13 million people aged above 80 who are not fully vaccinated as of November 11, 2022). The study also finds that in the absence of NPIs:
Medical capacity would be quickly overwhelmed. Despite enough hospital beds nationwide for respiratory illness, the demand for ICU beds would exceed the availability by 15.6 times during the peak of the wave.
Increasing vaccination and booster rates could lower deaths substantially. Closing the vaccination gap for the elderly (i.e., all eligible individuals aged 60 and above would be vaccinated) would lead to a 33.8, 54.1 and 60.8 percent decrease in hospitalization, ICU admissions and deaths, respectively, compared to the baseline.
Effective antiviral treatment could also help lower the medical burden. Using China’s approved antiviral therapies, at an 80 percent effective rate, could lead to a 36.5, 39.9 and 40 percent decrease in hospitalization, ICU admission and deaths, respectively. In the best-case scenario, if all cases were treated by highly effective medicines, the decrease could be improved to 81.2, 88.8, and 88.9 percent, respectively.
In the absence of high vaccination coverage among the vulnerable and antiviral treatments, only strict NPIs could help lower the medical burden meaningfully. NPI adoption would have to be able to reduce the reproduction number to values no larger than 2 in order to have a substantial decrease in health consequences. The strict NPIs are found only to be able to delay epidemics beyond the projected 6-month horizon, leaving the final impact uncertain.
12. A comprehensive set of measures is needed. The scenarios discussed by Cai and others (2022) suggest that no single measure can reduce the number of deaths to a level close to influenza related deaths in China. A combined effort would be needed to keep medical demand within capacity, for which increased vaccination among the elderly and wide availability of antiviral treatment hold the key.
E. Preparing for the Gradual Lifting of Remaining Restrictions
13. A prepared and orderly strategy to fully lift remaining restrictions could minimize the tradeoff between health outcomes and the economic cost. Any such approach would involve a further move away from the use of the most stringent non-pharmaceutical restrictions, such lockdowns of those not infected and stringent mobility restrictions on the vaccinated population, and pandemic prevention and control mainly relying on the protection from effective vaccines, treatments, and the existing healthcare system, supported by a phased withdrawal of social distancing measures.
14. Filling the vaccination gap, especially for the most vulnerable, is an essential precondition for a full easing of restrictions. With lockdown measures fully lifted, the number of infections would be expected to rise but high vaccination levels would imply a high share of asymptomatic or mild cases which could be isolated at home for recovery without hospitalization. Severe cases and deaths would be expected to remain at low levels if vaccines fully protect the most vulnerable groups and sufficient treatment options are available. The rise in hospitalization and ICU demand could thus remain within the capacity of the medical system.
15. China’s vaccination campaign is well advanced, but important gaps remain for booster shots and vaccination of the elderly. 90 percent of the population have received at least two shots and 63 percent have been boosted as of November 2022, but less than 6 percent of the population received a shot in the last six months. Waning immunity in the general population thus remains a concern (see Text Figure). In addition, vaccination rates among the elderly are still relatively low. Among the especially vulnerable over-80-year-old group, just over 65 percent have received two shots and around 40 percent have been boosted (see Text Figure).
Vaccine Protection has been Waning
(Vaccinations by month, in percent of population)
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Sources: Our World in Data; BMC Infectious Diseases; The Lancet; and IMF staff estimates.Notes: Intertemportal vaccine efficacy (VE) in percent based on studies for US-approved vaccines. Due to irregular reporting intervals, some numbers are linearly interpolated. Numbers cannot account for people who have been both fully vaccinated and boosted in a given time period, so present an upper bound.Vaccine Protection has been Waning
(Vaccinations by month, in percent of population)
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Sources: Our World in Data; BMC Infectious Diseases; The Lancet; and IMF staff estimates.Notes: Intertemportal vaccine efficacy (VE) in percent based on studies for US-approved vaccines. Due to irregular reporting intervals, some numbers are linearly interpolated. Numbers cannot account for people who have been both fully vaccinated and boosted in a given time period, so present an upper bound.Vaccine Protection has been Waning
(Vaccinations by month, in percent of population)
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
Sources: Our World in Data; BMC Infectious Diseases; The Lancet; and IMF staff estimates.Notes: Intertemportal vaccine efficacy (VE) in percent based on studies for US-approved vaccines. Due to irregular reporting intervals, some numbers are linearly interpolated. Numbers cannot account for people who have been both fully vaccinated and boosted in a given time period, so present an upper bound.16. The booster vaccination campaign remains vital considering recent academic research suggesting domestic vaccines’ effectiveness against severe illness. Despite lower efficacy of vaccines against infection, recent research—based on the large Omicron outbreak in Hong Kong SAR in the first quarter of 2022 (McMenamin and others, 2022)—suggests that three doses of Sinovac’s domestic vaccine could be effective at preventing severe illness. While the study showed considerably higher efficacy for BioNTech’s mRNA vaccine under a 1-or 2-dose regime, both vaccines were found to offer very high levels of protection against severe outcomes following a booster shot (above 95 percent across all age groups). Based on the vaccination pace observed across China’s vaccination campaign so far, boosting the entire population could be achieved within few months if the pace went back to where it was during the initial vaccination campaign in 2021 (see Text Figure), assuming no supply constraints and no vaccine hesitancy. Data gaps remain on the impact of new variants and the waning of effectiveness, but ensuring protection is maintained would likely be contingent on repeated booster campaigns.
17. Clear forward-looking communication would facilitate the transition and reduce uncertainty. Communication could elaborate on the sequence and conditions for further withdrawal of measures toward a full normalization of activity. A well communicated strategy could help anchor business and household expectations for the upcoming changes and reduce uncertainty in their decision making.
China’s Vaccination Campaign
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
China’s Vaccination Campaign
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
China’s Vaccination Campaign
Citation: IMF Staff Country Reports 2023, 081; 10.5089/9798400233517.002.A001
F. Conclusions
18. Going forward, China should further adjust its COVID strategy in a way that would both mitigate the impact on economic activity and continue to protect lives. Doing so will require: (i) a well communicated strategy, (ii) ramping up the roll out of booster shots that are effective against new COVID variants and targeting the under-vaccinated elderly, (iii) making antiviral therapies—either domestically developed or sourced from abroad in line with the recent agreements to jointly produce antiviral drugs with foreign companies—widely available and (iv) scaling up health-care capacity as (v) containment measures continue to be gradually adjusted by making them more flexible and less restrictive for economic activity, while containing domestic transmission within healthcare capacity.
Appendix I. Data and Empirical Analyses
Data
19. The city level mobility data is taken from the Gaode 100-city congestion index. It measures mobility as a ratio between the time needed for a road trip and the time needed for the same trip without congestion. As a benchmark, average monthly mobility from 2017 to 2019 is treated as pre-COVID levels. Data since January 2020 is measured in percent deviation from the pre-COVID benchmark levels.
20. The city level economic data are nominal economic variables, sourced from the respective municipal statistic bureaus. The variables are nominal GDP, retail sales, and disposable income. For each city, the Compounded Annual Growth Rate between 2016 and 2019 is treated as the pre-COVID trend growth. The counterfactual benchmark for each variable is calculated by extrapolating using the pre-COVID trend growth. The actual turnouts since January/Q1 2020 are transformed into percent deviation from the counterfactual pre-COVID benchmark levels.
21. The stringency index is sourced from Oxford COVID-19 Government Response Index for each Chinese province. The index collects government’s non-pharmaceutical interventions in response to the COVID outbreaks and translates into a numerical index between 0 and 100 for which a higher value indicates stricter policy.
22. The confirmed cases and death numbers are from the National Health Commission of China. The dominant variant of the COVID virus is assumed to be the Delta variant for the period between Q3/July 2021 and Q4 2021/February 2022; and Omicron since Q1/March 2022.
Empirical Strategy
23. For the economic impact, we apply the following panel regression:
yj,t = βj + α0mobilityj,t + α1 variantt + α2seasonal factort + ɛj,t
24. In the equation above, economic variables (yi,t) measured in percent deviation from pre-COVID trend for city (j) are regressed on the decline in mobility, adjusted for the major virus variants and seasonal factors.
25. To better analyze the sources of changes in mobility, mobility is decomposed into policy stringency and voluntary mobility using the following equation, where the voluntary part is proxied as the residual.
mobilityj,t = βj + α0stringencyj,t + ɛj,t
26. The decomposed mobility measures are used as explanatory variables for the economic activities.
yi,t = βj + α0mobilityj,t + α1voluntary mobilityt + α2variantt + α3seasonal factort + ɛj,t
The regression results are shown below, with 95 percent confidence intervals shown in the brackets underneath each parameter.
Empirical Results
China: Activity
China: Activity
| Dependent variable: | |||
|---|---|---|---|
| GDP | Retail sales | Disposable income | |
| (1) | (2) | (3) | |
| Mobility | 0.66*** | 1.16*** | 0.37*** |
| (0.53, 0.80) | (0.96, 1.35) | (0.13, 0.61) | |
| Delta variant | 1.31* | -3.84*** | -2.99** |
| (-0.22, 2.84) | (-6.04, -1.65) | (-5.79, -0.20) | |
| Omicron variant | 0.48 | -7.24*** | -4.46*** |
| (-1.05, 2.01) | (-9.31, -5.18) | (-7.21, -1.71) | |
| Individual fixed effects | Yes | Yes | Yes |
| Seasonality adjusted | Yes | Yes | Yes |
| Observations | 896 | 815 | 566 |
| R2 | 0.58 | 0.38 | 0.47 |
China: Activity
| Dependent variable: | |||
|---|---|---|---|
| GDP | Retail sales | Disposable income | |
| (1) | (2) | (3) | |
| Mobility | 0.66*** | 1.16*** | 0.37*** |
| (0.53, 0.80) | (0.96, 1.35) | (0.13, 0.61) | |
| Delta variant | 1.31* | -3.84*** | -2.99** |
| (-0.22, 2.84) | (-6.04, -1.65) | (-5.79, -0.20) | |
| Omicron variant | 0.48 | -7.24*** | -4.46*** |
| (-1.05, 2.01) | (-9.31, -5.18) | (-7.21, -1.71) | |
| Individual fixed effects | Yes | Yes | Yes |
| Seasonality adjusted | Yes | Yes | Yes |
| Observations | 896 | 815 | 566 |
| R2 | 0.58 | 0.38 | 0.47 |
China: Activity
China: Activity
| Dependent variable: | |||
|---|---|---|---|
| GDP | Retail sales | Disposable income | |
| (1) | (2) | (3) | |
| Stringency | -0.23*** | -0.28*** | -0.31*** |
| (-0.31, -0.16) | (-0.39, -0.17) | (-0.45, -0.18) | |
| Voluntary Mobility | 0.63*** | 1.14*** | 0.30** |
| (0.49, 0.77) | (0.94, 1.34) | (0.06, 0.54) | |
| Delta variant | 4.84*** | 1.69 | 0.06 |
| (3.15, 6.53) | (-0.76, 4.15) | (-2.99, 3.10) | |
| Omicron variant | 3.52*** | -3.18*** | -1.24 |
| (1.91, 5.13) | (-5.36, -1.00) | (-4.04, 1.55) | |
| Individual fixed effects | Yes | Yes | Yes |
| Seasonality adjusted | Yes | Yes | Yes |
| Observations | 896 | 815 | 566 |
| R2 | 0.58 | 0.38 | 0.48 |
China: Activity
| Dependent variable: | |||
|---|---|---|---|
| GDP | Retail sales | Disposable income | |
| (1) | (2) | (3) | |
| Stringency | -0.23*** | -0.28*** | -0.31*** |
| (-0.31, -0.16) | (-0.39, -0.17) | (-0.45, -0.18) | |
| Voluntary Mobility | 0.63*** | 1.14*** | 0.30** |
| (0.49, 0.77) | (0.94, 1.34) | (0.06, 0.54) | |
| Delta variant | 4.84*** | 1.69 | 0.06 |
| (3.15, 6.53) | (-0.76, 4.15) | (-2.99, 3.10) | |
| Omicron variant | 3.52*** | -3.18*** | -1.24 |
| (1.91, 5.13) | (-5.36, -1.00) | (-4.04, 1.55) | |
| Individual fixed effects | Yes | Yes | Yes |
| Seasonality adjusted | Yes | Yes | Yes |
| Observations | 896 | 815 | 566 |
| R2 | 0.58 | 0.38 | 0.48 |
China: Policy Response and Voluntary Mobility
China: Policy Response and Voluntary Mobility
| Dependent variable: | ||
|---|---|---|
| Stringency | Voluntary mobility | |
| (1) | (2) | |
| (1) | (2) | |
| confirmed cases (log diff) | 0.02*** | -0.002*** |
| (0.02, 0.03) | (-0.004, -0.001) | |
| deaths (log diff) | 0.05*** | -0.02*** |
| (0.03, 0.06) | (-0.03, -0.02) | |
| Delta | 3.46*** | -1.13*** |
| (2.51, 4.41) | (-1.60, -0.66) | |
| Omicron | 4.68*** | 1.48*** |
| (3.50, 5.86) | (0.92, 2.04) | |
| Individual fixed effects | Yes | Yes |
| Seasonality adjusted | Yes | Yes |
| Observations | 2,779 | 2,587 |
| R2 | 0.36 | 0.21 |
China: Policy Response and Voluntary Mobility
| Dependent variable: | ||
|---|---|---|
| Stringency | Voluntary mobility | |
| (1) | (2) | |
| (1) | (2) | |
| confirmed cases (log diff) | 0.02*** | -0.002*** |
| (0.02, 0.03) | (-0.004, -0.001) | |
| deaths (log diff) | 0.05*** | -0.02*** |
| (0.03, 0.06) | (-0.03, -0.02) | |
| Delta | 3.46*** | -1.13*** |
| (2.51, 4.41) | (-1.60, -0.66) | |
| Omicron | 4.68*** | 1.48*** |
| (3.50, 5.86) | (0.92, 2.04) | |
| Individual fixed effects | Yes | Yes |
| Seasonality adjusted | Yes | Yes |
| Observations | 2,779 | 2,587 |
| R2 | 0.36 | 0.21 |
References
Cai, J., and others, 2022, “Modeling transmission of SARS-CoV-2 Omicron in China”, Nature Medicine 28, 1468–1475 (2022). https://doi.org/10.1038/s41591-022-01855-7
McMenamin, M. E., and others, 2022, “Vaccine effectiveness of two and three doses of BNT162b2 and CoronaVac against Covid-19 in Hong Kong”, The Lancet, Infectious Diseases, July (2022). https://doi.org/10.1016/S1473-3099(22)00345-0
Lewnard, J.A., and others, 2022, “Clinical outcomes associated with SARS-CoV-2 Omicron (B.1.1.529) variant and BA.1/BA.1.1 or BA.2 subvariant infection in Southern California”, Nature Medicine. https://doi.org/10.1038/s41591-022-01887-z
Liu, Y., Rocklöv, J., 2021, “The reproductive number of the Delta variant of SARS-CoV-2 is far higher compared to the ancestral SARS-CoV-2 virus”, Journal of Travel Medicine, Volume 28, Issue 7, October 2021, taab124 (2021). https://doi.org/10.1093/jtm/taab124
Liu, Y., Rocklöv, J., 2022, “The effective reproductive number of the Omicron variant of SARS-CoV-2 is several times relative to Delta”, Journal of Travel Medicine, Volume 29, Issue 3, April 2022, taac037 (2022), https://doi.org/10.1093/jtm/taac037
Prepared by Anne Oeking and Fan Zhang.
For example, according to Phua et al. (2020): “Critical Care Bed Capacity in Asian Countries and Regions.” Critical Care Medicine January 2020, China has 3.6 critical care beds per 100,000 population compared to 7.1 in Hong Kong, 7.3 in Japan, 10.6 in Korea, 11.4 in Singapore, or 12 intensive care beds per 100,000 population across OECD countries according to the OECD.
Lewnard and others (2022) show the relatively lower severity of the Omicron variant compared to Delta, beyond what can be explained by previous infections and vaccinations.
