What do economic models say about Covid-19?

Dr Bryce Wilkinson
The National Business Review
22 March, 2020

No one doubts the economic costs of lost jobs and income from Covid-19 will be substantial in the short run. But what can we know about the likely magnitude of those costs?

The virus is clearly highly contagious. Guesses about the proportion of a population likely to be infected if nothing is done range between 20% and 60%. In contrast, the common winter influenza normally affects about 10% of the uninoculated population.

Similarly, there is little clarity about how many of those infected will die. If it is 2%, that’s about 20 times higher than the flu. By comparison, the death rate for those infected by SARS in 2002 was about 10-11% but its spread was contained by standard public health measures.

Even economic modelling of the assumed epidemiological effects are complex.

For instance, contagious viruses tend to cause labour productivity losses from workers taking sick leave or time out to care for family members. Fear of contracting a virus causes people to take defensive actions to self-isolate. Tourism drops away. Attendance in public places, such as restaurants, pubs, cafés and cinemas, also dips, with many leisure activities falling away completely.

Firms lose business, creating anxiety and stress for owners and employees. Companies experience big upward swings in demand for services like acute respiratory care and downward global swings in many other products, not just oil. Disruption to international supply chains also reduces production.

Heightened pressure

Highly geared companies face bankruptcy as share prices, wealth and confidence all fall. Banks come under pressure to tide over companies and households experiencing financial stress. Public health systems also struggle, which only serves to heighten anxiety about contagion. The public demands politicians ‘do something,’ so leaders are likely to respond with regulations to reduce the risk of a virus entering the country (and better containment if it does) along with monetary and fiscal stimulus to maintain a degree of confidence.

These sorts of effects will differ from country to country according to the quality of their health systems; the strength of their trade and people-flow linkages to the source country (China); the degree to which its businesses and governments are leveraged; and how governments respond.

Taking all these considerations – and more – into account, Australian National University economics Professor Warwick McKibben and Roshen Fernando have developed a model to assess the economy-wide impacts of Covid-19 for seven scenarios for each of six industry sectors in 20 countries and six regions. They published the results on March 2, 2020.

Three of their scenarios presumed the virus spread from China to the other countries. The mildest of them assumed an infection rate in China of 10% with a 2% mortality rate. The most severe of the three had an infection rate of 30% and a mortality rate of 3%. (The middle scenario assumed 20% and 2.5% respectively).

For China, the falls from otherwise expected output and incomes (GDP) in 2020 ranged from 1.6% to 6.2%. For other countries, the impact on expected GDP was higher or lower depending on each county’s unique mix of the above considerations.

In the model’s three general pandemic simulations, roughly half the 20 countries had a tougher time than China, with Japan the hardest hit, followed by Germany. Australia was among the next most adversely affected group given its tourist and trade links.

Unfortunately, New Zealand was not included in the models but would presumably have shown similar proportionate vulnerability to Australia.

The modelled GDP losses for Australia in 2020 from otherwise expected GDP ranged from 2.1% to 7.9% and deaths ranged from 21,000 to 96,000. (At March 16, only five people have died in Australia.) The projected mortality rate for its total population ranged from .09% to 0.4%. By comparison, the projected mortality rate for China in each scenario was over twice as high, presumably because wealthy countries such as Australia have better health systems.

Globally, the modelling projected from 15 million to 68 million deaths across the three scenarios. To put this into perspective, the WHO says the typical global death rate from annual flu is somewhere between 290,000 to 650,000.

Serious pandemic

Projections are not predictions. Projections merely depict the modelled implications of assumptions. These assumptions may be seriously awry and the modelling full of over-simplifications and errors in assumed values for parameters – for example, people’s hygiene responses. The latest global GDP forecasts from the IMF and the Asian Development Bank are much more optimistic – for now.

New Zealand’s Treasury did conduct some insightful modelling after the SARS scare in a March 2006 policy perspective paper. The Treasury’s simulations assumed a 40% infection rate and a 2% case fatality rate, with 40% absenteeism and an average time off work of three weeks.

Those assumptions postulate a serious pandemic. The simulated lost production and income in the pandemic year were about 5-10% of GDP. Over four years, the cumulative loss jumped to 10-15% of one year’s production. That is a lot. Consider that the central government’s spending on health and education is about 11% of GDP.

The authors added that New Zealand has experienced six incidents in the past 100 years and the ball-park loss in real GDP was 8%, compared to otherwise.

Of course, none of this is grounds for either complacency or panic, both of which would only increase economic costs and cost lives. Healthy people and children appear not to be dangerously affected by Covid-19 but we all have a communal responsibility to do our bit to avoid unwittingly infecting others.



Stay in the loop: Subscribe to updates