Assessment of monthly economic losses in Wuhan under the lockdown against COVID-19

COVID-19 emerged in Wuhan at the end of 2019 and rapidly spread throughout China and globally. To curb transmission, China enacted an unprecedented lockdown of Wuhan, a metropolis of over 10 million with a comprehensive industrial system. This intervention, while aimed at disease control, had substantial social and economic repercussions necessitating scientific quantification. The study aims to assess, for the first month of lockdown (January 23–February 23, 2020), the interplay between epidemic prevention measures and socio-economic development by estimating monthly health losses (physical and mental) and meso-economic losses (direct and indirect industrial losses). This provides reference for managing major infectious disease outbreaks and understanding the broader impacts of city-wide lockdowns.

Prior work on COVID-19 has emphasized (1) understanding transmission dynamics for prevention and control, and (2) assessing economic losses and public health burdens. Studies have evaluated cost-effectiveness of interventions using transmission models, estimated economic impacts of epidemics, and used surveys to capture absenteeism. Lockdowns and closures of factories, services, and schools have led to the use of CGE models to assess economic losses. Input-Output (IO) models have a history of application to disaster impact assessment (e.g., earthquakes, hurricanes) and address sector-specific and inter-industry loss propagation. This study builds on these strands by combining a transmission model and health burden assessment with an IO framework to quantify both health-related and meso-economic losses from Wuhan’s lockdown.

Design: A data-driven modeling study estimating monthly health and meso-economic losses in Wuhan during the first month of lockdown (Jan 23–Feb 23, 2020).

Data sources: Three categories of data were used: (1) epidemiological data on reported COVID-19 cases (Chinese CDC, WHO, Health Commissions of China and Hubei) and parameter values from literature and official handbooks; (2) business and revenue data for directly affected industries (transport: airline, railway, expressway, metro, taxi, buses; logistics and warehousing; postal; accommodation; food and beverage) from official reports and databases (e.g., Variflight, ICAO, Wuhan Metro Group, East Money database, Wuhan Bus Group, Hubei Traffic Investment Group, Ministry of Transport, Wuhan Bureau of Transportation, Wuhan Statistical Yearbook); (3) supporting socio-economic data for valuation (treatment costs, GDP per capita for time valuation, population figures, state compensation standards, national IO table for 149 sectors).

Health loss assessment: A compartmental SIR-type transmission model was constructed and extended to account for healthcare capacity and isolation. Compartments included Susceptible (S), Mild patients (I), Severe patients (I1), Mild patients in mobile cabin hospitals (F), Severe patients in designated hospitals (H), Recovered (R), and Dead (D). Only mild patients were assumed infectious; severe patients were quarantined. Two stages reflecting capacity changes were modeled: Stage 1 (Jan 23–Feb 5) before mobile cabin hospitals opened; Stage 2 (Feb 5–Feb 23) after opening. Bed capacity effects were parameterized via ratios b1 (mobile cabin hospitals affecting mild patient occupancy, transfer, transmission) and b2 (designated hospitals affecting hospitalization and mortality of severe cases). Initial conditions on Jan 23, 2020: S(0)=55,439; I(0)=269; F(0)=1,000; I1(0)=129; H(0)=494; D(0)=23; R(0)=31. Representative parameters (from literature) included contact transmission rate β≈0.813; recovery rates γ=1/17 for mild, γ1=1/20 for severe; mortality rates for severe d≈0.016; and transition probabilities among compartments reflecting hospital capacity effects. The model simulated cumulative cases by category (mild, severe, deaths) over the one-month period.

Health burden calculation: Health endpoints were defined by disease severity (mild, severe, death). Economic loss per patient Li combined direct medical costs (diagnosis, treatment, beds, nursing, medication) and indirect time costs (work loss during treatment and government-mandated 14-day post-discharge quarantine) valued by Wuhan daily GDP per capita (2018 average 135,136 CNY/year with 7% growth). Total health loss L summed across endpoints as Li×Pi, where Pi are simulated cumulative cases. Average treatment costs were derived from China Health Statistics Yearbook and hospital surveys; time durations for treatment and quarantine were based on official guidance and literature. Mental health loss was proxied by (a) monthly revenue of cultural and entertainment services (annual 2018 revenue converted to monthly) lost due to closures, and (b) state compensation standard for violating personal freedom (346.75 CNY/day) applied to approximately 10 million quarantined residents for one month.

Meso-economic loss assessment: Direct losses were computed for sectors directly shut or curtailed: air transport, railways, expressways, metro, taxis, buses, logistics and warehousing, postal services, accommodation, food and beverage. Examples: Air transport direct loss per canceled flight combined aircraft fixed cost (~70,000 CNY/day) and passenger loss (50 CNY per passenger) using average load factors and Spring Festival flight schedules; monthly shutdown losses for rail, road tolls, metro, buses, taxis derived from reported daily/monthly/annual revenues; logistics and warehousing and postal monthly losses estimated from annual revenues adjusting for Spring Festival closures; accommodation and food & beverage monthly losses from 2018 annual revenues.

Indirect losses across all industries were evaluated using a static Leontief IO model with the 2017 national IO table (149 sectors). Assuming stable inter-industry relationships and Wuhan’s structure similar to national averages, the complete consumption coefficient matrix B=(I−A)^{-1}−I was used to translate direct final demand losses ΔY in directly affected sectors into total output losses ΔX=(B+I)ΔY. Sectoral indirect losses were computed by multiplying each directly affected sector’s direct loss by the corresponding complete consumption coefficients, then aggregating across all directly affected sectors to obtain each of the 149 sectors’ indirect losses and total indirect loss.

Policy effect analysis: The study examined changes in the proportions of mild, severe, and fatal cases over three subperiods (Jan 23–Feb 4; Feb 5–Feb 13; Feb 14–Feb 23) and tracked the time-varying effective reproduction number R(t) to infer lockdown impacts on transmission and clinical severity distribution.

• Transmission and cases: The SIR-based simulation estimated 55,616 cumulative confirmed cases in Wuhan during Jan 23–Feb 23, 2020: 41,430 mild, 12,100 severe, and 2,086 deaths.
• Health burden (physical): Total one-month economic loss from patients’ health was approximately 4.4899 billion CNY. Breakdown of total losses: mild 0.990 billion CNY; severe 1.253 billion CNY; deaths 2.247 billion CNY. Costs per patient: mild ~23,886 CNY; severe ~103,588 CNY; death ~1,077,132 CNY.
• Mental health loss: Cultural and entertainment closures implied a proxy monthly loss of about 10.5204 billion CNY; personal freedom restriction compensation proxy contributed about 104.025 billion CNY. Total monthly mental health–related loss estimated at approximately 114.5454 billion CNY.
• Direct meso-economic losses (selected sectors): Total about 21.6094 billion CNY across directly affected industries. Illustrative components: flight cancellations ~1.0892 billion CNY; rail ~0.3843 billion; expressways ~1.8567 billion; public transport (taxis, buses) ~1.090 billion; metro ~1.0047 billion; logistics & warehousing ~6.5661 billion; postal ~1.2523 billion; accommodation ~1.064 billion; food & beverage ~7.302 billion.
• Indirect meso-economic losses (economy-wide): Total approximately 36.3966 billion CNY across 149 sectors via IO linkages. Most affected sectors included manufacture of refined petroleum products and processing of nuclear fuel (~1.813 billion CNY), financial services (~1.670 billion), farming (~1.576 billion), business services (~1.547 billion), road transport (~1.447 billion), extraction of crude petroleum and natural gas (~1.376 billion), electricity and steam supply (~1.196 billion), parts and accessories for motor vehicles (~1.118 billion), retail trade (~1.103 billion), and real estate (~1.089 billion.
• Total monthly economic losses: Approximately 177.0413 billion CNY, comprising health burden (4.4899 billion), mental health loss (114.545 billion), direct sector losses (21.6094 billion), and indirect losses (36.3966 billion).
• Macroeconomic context: Monthly total losses equaled about 11.06% of Wuhan’s 2019 annual GDP and 52.73% of 2019 Q1 GDP. Health and mental health losses accounted for 7.44% of annual GDP and 35.45% of Q1 GDP.
• Policy effectiveness: Lockdown associated with reductions in the shares of critical and fatal cases over time and declines in the effective reproduction number R(t) to below 1 later in the month. The policy is estimated to have prevented over 180,000 infections, saved about 20,000 lives, and avoided nearly 30 billion CNY in medical costs, indicating significant long-term societal benefits.

The study’s objective was to quantify, on a monthly basis, the health and meso-economic losses attributable to Wuhan’s lockdown and to assess the overall policy effect. The combined SIR and health burden framework linked clinical severity distributions and case counts to monetary health losses, while the IO model captured how direct shocks to transport, logistics, postal, accommodation, and food services propagated as indirect losses to upstream and downstream sectors. Results show that although the mental health proxy losses dominate the monthly total, the health burden from morbidity and mortality is substantial, and direct shutdown losses in key services further amplify into sizable economy-wide indirect losses. Nevertheless, the lockdown reduced transmission (R(t) fell below 1), lowered the proportions of severe and fatal cases over time, and likely averted very large medical costs and deaths. In this sense, while the short-term economic impacts are significant, they are counterbalanced by avoided long-term health and economic damages. Sectoral analysis highlights heightened vulnerability in secondary industries (e.g., refined petroleum, manufacturing) and critical service sectors (finance, real estate, business services, retail), implying the need for targeted mitigation and recovery strategies. Overall, findings support the conclusion that decisive containment policies, though costly in the short run, can keep total economic losses at a controllable level when they effectively suppress the epidemic.

This study provides a first-month assessment of Wuhan’s lockdown impacts by integrating an extended SIR-based health burden assessment with an IO-based meso-economic loss evaluation. It estimates total monthly economic losses of approximately 177.04 billion CNY, including 4.49 billion CNY in physical health burden, 114.55 billion CNY in mental health–related loss proxies, 21.61 billion CNY in direct sectoral losses, and 36.40 billion CNY in indirect losses across 149 sectors. Despite these losses, the lockdown likely prevented over 180,000 infections, saved roughly 20,000 lives, and avoided nearly 30 billion CNY in medical costs, indicating net long-term benefits. The sectoral footprint underscores the need for targeted support to highly affected industries and for strategic recovery planning. The approach is generalizable to other cities and contexts. Future research should extend to comprehensive macroeconomic assessments (consumption, exports, investment), incorporate direct losses across all sectors, integrate dynamic CGE analyses, and account for both adverse and beneficial indirect health effects (e.g., delayed care vs reduced accidents) as better data become available.

• Scope of sectoral coverage: Direct economic losses were estimated for selected directly affected sectors (transport, logistics and warehousing, postal, accommodation, food and beverage) but not for all meso-economic sectors.
• Macroeconomic perspective: No full macro-level assessment of impacts on consumption, exports, and investment due to data limitations; IO model is static and does not capture price adjustments or behavioral responses.
• Data limitations and proxies: Health treatment costs and time costs used averaged/representative values; mental health losses were proxied by cultural service revenues and state compensation standards, which may not fully reflect actual mental health impacts.
• Model assumptions: Assumed only mild cases are infectious; used national 2017 IO table as a proxy for Wuhan’s structure and assumed stable inter-industry relationships; two-stage capacity effects simplified complex healthcare dynamics.
• Indirect health effects: Potential opposing indirect health effects of lockdown (e.g., reduced traffic fatalities vs increased mortality from delayed care) were acknowledged but not quantified.
• Parameter uncertainty: Transmission and clinical parameters were drawn from early literature and may be uncertain; results depend on these inputs.