COVID-19’s impact on jobs and education: using data to cushion the blow

In the wake of the pandemic, infection mitigation strategies curbed the virus but had inevitable consequences for school children and job-holders. Almost two years on, there are still no perfect solutions to controlling the virus and COVID-19’s impact on jobs and education. But city-based pandemic data modelling may indicate the best approach.

The pandemic’s inducement of nationwide lockdowns has led to business closures and bankruptcies. In 2020, there was an 8.8% reduction in working hours and over 114 million people were left unemployed. School closures have also disrupted education, resulting in over 100 million children falling below the minimum reading proficiency level. These consequences are particularly stark in developing countries like the Philippines, where inequalities are greater.

As of September 2021, Metro Manila - the Philippines National Capital Region - had undergone four lockdowns of varying stringency, totalling nine months of restrictions on work, school and other public gatherings. These lockdowns have widened disparities in education and employment. The digitalization of schoolwork inevitably favours the privileged as only 55% of students in the National Capital Region have internet access.

What’s more, public schools are having difficulty adapting due to a lack of resources and preparedness. The Philippines also reached an unemployment rate of 10.3%, the highest in 15 years. Given these challenges, what strategies can help reopen schools and offices and halt the widening gap between rich and poor?

Data can inform safe school reopenings and return to work

At the start of the pandemic, predicting COVID-19’s spread and its impact was challenging given how little we knew about the virus, which meant the Philippine government’s decision-making was guided by domestic reports on positive cases and patients’ symptoms. This retrospective approach and subsequent stay-at-home orders limited the virus’ impact but it came at the expense of education and livelihoods of those unable to work from home. What communities need now is a strategic response to the pandemic and disruptions to these sectors.

One way to develop an apt strategy is by harnessing pandemic data and feeding it into a computational model called SEIR, used in past epidemics. Although the model’s accuracy depends on historical data, it can aid decision-makers in proactively evaluating plans before implementation.

The model works by segregating individuals into the following disease states:

By categorizing individuals this way, we can simulate the rate they move across groups according to time. This sequencing makes it possible to visualize infection dynamics and forecast transmission with different interventions at play, for example, the extent of reopening, physical distancing and mask-wearing.

City-level modelling

We used age-dependent contact data, within an agent-based model and piloted it in one city - Pasig in Metro Manila - for which reliable data was available. We then simulated different scenarios:

We then checked whether 100% opening is possible when integrating different levels of mask-wearing and physical distancing:

Based on the modelling, 100% of school and office openings in Pasig City would not substantially increase infections if mask-wearing and physical distancing were maintained. With moderate measures, 100% opening of schools and offices increases infections by 3.9% and 5.2%, that is 3,900 and 5,200 individuals per 100,000 of the population, respectively.

The ‘city bubble' approach

Leveraging pandemic data to foresee how interventions affect transmission will likely bring about a more strategic reopening. However, a national implementation may spur infections causing unfavourable health and economic repercussions.

How then can we translate this into policy?

The COVID-19 bubble concept – a method where more than one household (usually two) can have physical contact during restrictions – has been applied in New Zealand, Australia and the United Kingdom. Building upon this approach, data-guided reopening policies can be implemented city by city. This is how:

1. Start with one city, one bubble

Selecting a single city to roll out a pilot study can determine if the model-based reopening strategies work. The chosen city should have an efficient contact tracing system and not be a COVID-19 hotspot, to minimize the risk of the pilot going awry.

2. Reopen schools first

Our data suggest that reopening schools results in low infections among primary and secondary students regardless of opening strategy, unlike workplaces where 50% and 100% opening might significantly increase the infected working population. Nevertheless, mitigation strategies like vaccinating school personnel and early isolation of suspected cases should be considered.

3. Reopen according to age group

Studies have shown that children have a lower susceptibility to COVID-19 compared with adults. Despite that, similar infection rates in school and work reopening can be observed – some children are likely asymptomatic but can infect their parents. Nonetheless, starting from the least susceptible age group enables careful monitoring of new infections. This is further strengthened by observing additional precautions like limiting school hours, downsizing classes and establishing infection control task forces.

4. Adapt and apply the policy to other cities

In the pilot phase, it is important to catalogue what did and didn’t work, and to improve the model with data on vaccine efficacy, asymptomatic cases and virus variants. These learnings can help scale and replicate these strategies in other cities. The pandemic highlights the importance of collaboration between policymakers, healthcare providers, scientists, educators, and business leaders to ensure systematic and safe reopening strategies. As developing countries have especially limited healthcare resources, the city bubble approach could address the trade-offs between infection spread and harms from prolonged disruptions to school and work.