Reopen the economy when the time is right based on data

Setting the context

Some countries have begun to reverse the upward trend of COVID-19 cases, while others remain on an upward trajectory despite their best efforts. However, countries in both categories have started reopening their economies. With markets, industries and educational institutions closed for weeks or months, governments are faced with deciding how long they are willing to sustain the negative social and economic impact of restricted movement. It is too early to tell how reopening the economy will impact the long-term health recovery trajectory. What we attempt in this section is to briefly highlight some of the emerging trends we are seeing and to identify the role that some of these sector decisions might play in the health recovery of countries as they respond to the pandemic. These insights can inform a broader narrative that can inform effective practice moving forward.

This section therefore focuses on two key areas: 

  • Effective use of data and relevant information to drive decision-making 
  • Effective practice in balancing the recovery in health with reopening of the education sector (as an illustration of resuming a critical area of public life)

Key findings from the analysis

How have countries used data effectively to inform their decisions on easing of restrictions, and what can this tell us about trends in recovery? 

Key Finding 1: Countries that have begun to reopen sectors on the basis of improvements in COVID-19 data seem to be able to continue progress on their health recovery.  

It is too early to make firmly evidence-based assumptions on what is or is not working when it comes to balancing economic reopening measures with maintaining health outcomes. Many factors feed into the GCI recovery index, and the impact will vary over time as different stages of easing measures come into play. But as the situation evolves rapidly, clear trends in best practice are beginning to emerge. These include:

  • Use of big data to track real-time rise in infections and respond rapidly.[58] Often this coincides with countries that have compact demographics and can deploy mobile communication technology for rapid contact tracing, and targeted responses to minimise resurgences. Whilst this raises an important debate around use of data and privacy concerns, where advanced mobile apps with geolocation data are combined with human contact tracers, such as Singapore, Taiwan and South Korea, cases have continued to decrease rapidly even after key sectors such as education and public transport are reopened. Smart technology has played an outsized (and sometimes controversial) role here. As China began reopening its public transport system, QR codes were placed on Beijing’s subway trains and on buses in Shanghai to encourage passengers to register contact information to improve track and trace efforts. 
  • Adapting the model of reopening to disease progression data. Many countries are planning a graduated model of reopening, easing restrictions in stages based on data that informs decision makers on the progression of the disease, the capacity of the health system to respond and the preparedness of the general public to adhere to measures.[59] Many countries are adopting traffic light systems to communicate their deliberate and phased roadmaps, from Australia’s ‘levels’, to New Zealand’s ‘alert system’ – both countries in the top 10 rankings of the GCI recovery index.
  • Phasing reopening by sector. Effective countries have prioritised the sectors to be released from lockdown based on analysis of four key considerations: (1) sectors less suitable for remote operations; (2) sectors that contribute larger percentage of gross domestic product (GDP) to the national economy; (3) value add per worker (releasing fewer workers, but releasing those with more productive contributions); and (4) business viability and the impact that state support can have on stimulating the economy.[60] It is important to consider combining this approach with real-time tracking of virus progression so that corrections can be made where the number of cases spikes. An MIT Digital Economy initiative, for example, developed a framework that ranks the viability of reopening specific industries based on the trade-off between risk of infection and importance to the economy. Work like this could provide a useful baseline to help countries think through the impact of opening specific businesses to the long-term health recovery.[61]

Phasing reopening by sector: Harvard analysis on two case studies: Switzerland and Spain

  • Switzerland – By constructing key measures of analysis (e.g. an indicator of ‘value loss in lockdown’ which combined ease of switching to remote operations and relative weight of contribution of the sector to the economy) and comparing this with an evaluation of the risk of contamination (e.g. density of the workplace amongst other factors) – the team concluded that employees in the health, construction and manufacturing sectors should be targeted for release in the first wave of easing measures. They could then identify which sectors to target in a second stage of release, and finally, once there were zero new cases – a third stage – a strategy could be developed that would progressively reopen businesses without risking overwhelming the public health system.
  • Spain – Similar analysis for Spain would see manufacturing, construction and health services released in stage one of reopening, with retail and real estate restarting operations in stage two. All other sectors (mainly tourism, financial and administrative services) could reopen in the final stage of reopening.
  • There is no one answer here. However, if governments use data continuously to adapt their phased reopening strategies, it could give them the space needed to continue learning about the changing dynamics of the pandemic to inform their ongoing strategy.

  • Maintaining snap-back mechanisms. On initially easing restrictions, Germany witnessed an increase in the COVID-19 reproduction number (going above one again in early May from a drop to 0.65 before reopening). Yet the government agreed to monitor the data and snap back into targeted lockdown measures if any county exceeds 50 new coronavirus cases per 100,000 residents.[62] This type of coherent national framework, combined with cooperative local implementation to improve adherence, will require strategies to rebuild public trust and manage uncertainty. Yet early indications are that these reopening strategies can pay off, with Germany maintaining good health recovery outcomes on the GCI.
  • Using data to manage and monitor recovery. This is increasingly important as restrictions are eased, with health dashboards helping governments manage the impact of reopening on resurgence in cases, business activity and the trade-off between the two. In the US, for example, a new ‘Management Metrics for Cities’ tool released by Bloomberg Philanthropies, Johns Hopkins University and Delivery Associates, provides one set of expert-vetted benchmarks to help local leaders use the best data to manage through the recovery period.[63]

Case Study: New Zealand

The full reboot model

  • The government in New Zealand waited until new COVID-19 cases effectively reached zero before restarting full economic and social activity. This required a set of conditions very specific to New Zealand as a remote group of islands, including tight border controls, high levels of test, track and trace, and the ability to get high levels of adherence to an early, strongly enforced and lengthy lockdown.[64]
  • Whilst New Zealand has since recorded new confirmed cases of COVID-19, it is clear that keeping its borders closed appears to be key to New Zealand’s broadly successful strategy to protect its elimination of community spread, with ‘travel bubbles’ being designed to stimulate economic progress in a measured way. This could first include travel between Australia and New Zealand but could be extended to COVID-free South Pacific island nations in July.[65] This approach might be unique to its context, essentially providing laboratory conditions under closed-border restrictions.
  • Contact tracing is another foundational element of the strategy, learning from Singapore’s model of a mobile-based app and continuing its globally high levels of testing.
  • The data roadmap released by the New Zealand government highlights some of the unique characteristics that enable New Zealand to be data-driven in its policymaking. Amongst these, the fact that 90 per cent of residents have access to the internet, the growing data economy which is projected to contribute over NZD 4.5 billion this year, and its dedication to an open data environment (ranking fifth on the Global Data Barometer) all enable data to play a unique role in government decision-making and the pandemic has been no different.[66]

Case Study: China

The graduated reopening approach

  • ‘First in, first out’ is not just a handy catchphrase for China’s experience of COVID-19, but highlights the importance of learning lessons from its experience, both in terms of what to adopt and what to adapt or change.
  • China was one of the first countries to adopt a graduated approach to lifting restrictions and reopening the economy. The government developed a national framework to guide decisions, but decentralised decisions on implementation to the local level. The framework graded businesses according to their transmission risk, and then phased low-to-medium risk businesses such as construction and manufacturing first, with high-transmission risk industries such as hospitality and education reopening more than a month later. This localised approach has clear benefits in terms of targeting efficiencies. However, it also poses logistical challenges for supply chains of the many businesses that operate across provinces with different rules.
  • Equally controversial is the government-endorsed health code system,[67] which gauges a person’s infection risk and colour codes them accordingly. This information is centralised on a data platform using QR codes to crowd-source information, which raises important questions on privacy and civil liberties while simultaneously enabling more rapid targeted isolation of outbreaks.

Key Finding 2: Some countries have experienced a decline in recovery after reopening, raising questions on the impact of reopening too much, too fast.

On the flip side, early evidence suggests that countries that haven’t invested in data-informed strategies have seen a negative impact on their health recovery. Doing this well isn’t easy. There are many statistical challenges with the data being used to inform these strategies, including limited access to the tools and data needed under lockdown conditions, inconsistent approaches to recording data (leading to complications in the assessment of impact and implications), and lower frequency in data reporting, meaning policymakers are making decisions based on infrequent and/or out-of-date information.[68]

One of the key challenges seen is limited track and trace programmes. Many countries have been unable to identify real transmission rates at a local level so far, reducing their ability to target response measures effectively. When a cluster of new cases was identified in Seoul, South Korea on 10 May, questions were raised on whether the government had reopened too much too soon. Yet with the capacity to test, trace and isolate individual cases in real-time, the government is arguably getting health recovery back on track without having to reinstitute lockdown for entire sections of the economy.[69] This is not the case everywhere:

Case Study: Israel

Surge in cases after measures deployed to reopen the economy

  • Israel had to ‘put the brakes on’ the reopening of its economy after reports of steep increases in infection rates at the start of June. When schools, shopping malls and restaurants reopened at the end of May, the country saw five times as many new infections each day (100 daily new infections up from 20 the week before reopening).[70]
  • Initially, Israel had been seen as a relative success case, with the National Security Council concluding that the infection rate had only increased by 0.1 per cent within 10 days of reopening schools and the retail economy. Yet as the planned measures to reopen aspects of the economy are put on hold, questions are being raised over the evidence that informed decisions over which aspects to reopen (e.g. shopping malls before public rail) and whether too much was reopened too soon.
  • Since then, Israel has had to close many areas of its economy again in an attempt to control further outbreaks. Schools, for example, that were reopened on 3 May had to close a month later after over 2,000 students, teachers and staff tested positive for COVID-19.[71]

The calculation is not a simple one, especially for populous, developing economies. The debate is ongoing on the cost of social distancing measures to contain the virus given the significant impact on economies with large informal sectors. What is becoming clear is the need for global partnerships, data exchange and advanced analytics to help governments make effective and evidence-informed decisions that work for them. Sometimes the best health interests will conflict with the best interests of economies, requiring political leaders to exercise judgment in difficult circumstances.

Case Study: Pakistan

The political challenge in targeting the right response

  • Prime Minister Imran Khan of Pakistan made the decision to lift lockdown measures despite data showing 20,000 new COVID-19 cases in the previous three weeks. That number more than doubled in the three weeks after measures were eased.[72] Whilst testing is increasing, adding complication to any assumptions being drawn on number of cases alone, the data suggests that community transmission is probably still widespread, leading local health experts to urge the government to reinstate some lockdown measures.
  • The graph below plots the ratio of total deaths to total cases through 27 April. Once broken down by demographics (for example age and gender of the deceased), critical policy implications become clear in terms of segments of the population that could remain on lockdown in order to stimulate the economy through safer demographics. Yet decisions in the country appear to not have prioritised the data, with mosques remaining open despite the clear risk to the most vulnerable age groups.

A close up of a map

Description automatically generated

Figure 47: Total deaths/total cases

A screenshot of a cell phone

Description automatically generated

Figure 48: Pakistan’s deceased demographic background

Source: Reuters

  • Reopening came with recommendations to follow social distancing rules, with face masks being made compulsory in public. Many still fear this won’t be enough to stop the continued rise in cases as shopping malls and markets have reopened through a Supreme Court ruling. Education is the only major sector that remained closed as of the end of May. Even with the government’s emphasis on citizen responsibility, official data suggests the infection rate has started increasing again, doubling every nine to eleven days since 1 April.[73]
  • Pakistan has a poor GCI recovery rating of just 2 as of 17 May, placing it 136th out of 184 countries analysed.

It remains too early to be definitive about the best model for reopening, with the graduated approach appearing to minimise the risk of increased viral reproduction rates, but arguably compounding the impact to the economy without a full sharp reboot.[74] However, with real-time data available, countries such as New Zealand, Austria, China and Switzerland have been able to develop economic roadmaps that are targeted and phased, but also flexible enough to adapt to the rapidly changing dynamics of the pandemic.

Key Finding 3: Key trends are emerging in countries that have reopened parts of their education sector and maintained a recovery in health metrics.  

As of June 2020, most countries had not yet reopened schools, and many still plan to end their academic year online. Schools in Israel, Norway, Japan, Denmark, China and Taiwan were among the first to reopen, with teachers, parents and students alike being forced to adjust to a new reality. By mid-May, an estimated 22 countries in Europe alone had reopened schools for certain grades, with education ministers insisting this has not led to a significant increase in infections amongst children, parents or school staff.[75] The graphs below speak to this trend, with minimal impact to progress on health recovery rates (based on the GCI). The lessons we can learn from these countries, however, can only provide initial indications. It is too early to extrapolate from these individual experiences to broader effective practice beyond what the current data can tell us.

Figure 49: Reopening of economies Source: GCI

These early trends highlight some key considerations for governments reopening their education sector, through a narrow lens of the role these decisions have played in the health recovery. 

  • Social distancing both inside and outside of the classroom is a key tactic in Denmark’s reopening strategy. The first country in Europe to restart classes on 15 April, much of the learning in Denmark’s schools has moved to outdoor public parks and spaces. Class ‘pods’ (groups of students) have gradually increased from groups of four to groups of 28 students in line with weekly updates to government guidelines in line with the data.[76]
  • Testing and screening measures have played a key role in China’s partial school reopening. Despite provincial variation in return dates and regulations, the Ministry of Education is requiring all students to have their temperature checked at school entrances, displaying a green sticker to align with the national smartphone health code programme.[77] Similar to Denmark, the latest data on the recovery index suggests a continued reduction in COVID-19 cases despite the partial reopening of the education sector in China.
  • New Zealand opted for staged returns to on-site schooling to allow time to monitor the data and evaluate the effects of the increased movement of people and community transmission. The government laid out a detailed set of guidelines for each of the four levels of ‘alert’ in response to COVID-19, which saw all educational facilities closed at level four, through to current level two status in which all education facilities, including schools, early learning centres and tertiary institutions are open with distance learning options for those unable to attend school. In Europe, 17 of the 22 countries that had begun reopening by mid-May were only allowing kindergarten, primary and final year secondary students back to school, to prioritise return-to-work policies and final examinations for secondary leavers.[78] This contextualisation to allow for adapting response strategies to local recovery conditions appears to be key in early strategies to reopen education to stimulate economic recovery whilst maintaining a strong GCI recovery index.

Case Study: Denmark

Success in reopening schools without impacting recovery

  • The Danish government worked closely with school community representatives (e.g. Danish Union of Teachers) to ensure buy-in for guidelines on school reopening across the ecosystem. This led to transparent publication of information on decision-making, informed by the Statens Serum Institut (SSI), the governmental public health and research institute. The SSI’s scientific modelling showed the infection rate was at 0.6 (R number) with just 433 activeCOVID cases in Danish hospitals at the time of reopening on 15 April.

Reopening was phased to allow social distancing between students. Classes were split in two, with learning ‘pods’ initially restricted to four students. With weekly updates to the guidelines from government, this has increased to class ‘pods’ of 28 students, complimented by strong hygiene practices including near-hourly handwashing in every classroom. Social distancing guidelines have been reduced from two metres to one, meaning most schools can now accommodate all pupils in one session. The government has not made guidelines definitive leading to schools interpreting them in different ways, with many still offering blended options of split days of on-site learning with online teaching to avoid overcrowding at schools.

A group of people in a room

Description automatically generatedImagen que contiene exterior, persona, gente, parque

Descripción generada automáticamente

  • Timetables were adapted to enable smaller class sizes and shift learning to outside spaces wherever possible, informed by SSI modelling that showed children had a low susceptibility rate to the coronavirus (including as carriers) and that infection spread was reduced significantly outside vs inside the classroom.
  • Initial fear of the virus led to just 50.7 per cent of pupils returning to primary school within the first three days, but by the third week of reopening, that number had jumped to 90.1 per cent of registered pupils. The R number reportedly rose from 0.7 to 0.9 within two weeks of reopening schools to students up to age 11, but quickly decreased again. Six weeks on from the first reopening of schools, it has remained at 0.7, in line with reductions in infections and deaths from coronavirus countrywide. The SSI reported as few as 112 COVID-19 patients in Danish hospitals at the end of May (a drop from 380 when primary schools reopened) leading epidemiologists to suggest the process has proven to be safe.

Recommendations: Interventions for countries to consider

Recommendation 1: Use data to guide decisions on when to open the different sectors of the economy and, ultimately, the economy as a whole.

  • With the impact of COVID-19 ongoing and a vaccine still months or years away, the transition period will last for some time to come. Strategies therefore need to be flexible and adapt to data-informed trends. 
  • Ensuring critical data is collected and used sector by sector and region by region is clearly important to manage the consequences of reopening the economy. In the short-term, when good real-time data may not yet be available, using the best that can be found makes sense. This might mean using temporary estimations, imputing data from alternative sources, or enhancing the transparency of data sharing to encourage quality controls in real-time.
  • Open Data Watch consolidated a set of resources that can help strategists in effective use of open data in responses and recovery plans from COVID-19.

It is important to bring equity to the forefront here. Even where data use has been effective in informing reopening strategies, most countries still have gaps in their levels of disaggregated data leading to large equity concerns across the world. The pandemic has proven to impact some groups disproportionately according to gender, race and ethnicity. Yet even where this disturbing fact has been recognised, such as the UK where Public Health England recently published a report on minority ethnic groups and the effects of COVID-19, the lack of disaggregated data makes it hard to make significant recommendations on how to address the problem.[79] This and other equity concerns should be at the heart of any data-driven approach.

Recommendation 2: Reopen targeted parts of the public sector (such as education or transport) with adequate support and guidelines in place to facilitate safety.

  • Deploy a contextually relevant range of testing and screening measures from self-administered health screening to thermal digital technologies to identify and isolate individual clusters and avoid reclosing whole sectors.
  • Establish a staged roadmap for reopening, analyse lead indicators and adjust the strategy in real-time as necessary.
  • Devolve some decisions to the local level to enable local contextual factors to be taken into account.


[58] Reopening the Economy, Moving Forward This Week: The Coronavirus Workplace Series, Michael Bernick, Forbes, 20 April 2020

[59] Beyond the Curve: How to Restart in the Wake of COVID-19, Amanda Brimmer, Vincent Chin et al., BCG, 16 April 2020

[60] The Case for Reopening Economies by Sector, Jean-Philippe Bonardi, Arturo Bris et al., Harvard Business Review, 19 May 2020

[61] What should reopen first? Data shows the way, Sara Brown, MIT Management Sloan School, 18 May 2020

[62] Renewed outbreaks in South Korea, Germany and China show continued risk as more countries seek to reopen, James Griffiths, CNN, 11 May 2020

[63] Management Metrics for Cities in the COVID-19 Crisis, Johns Hopkins University and Medicine

[64] Beyond the Curve: How to Restart in the Wake of COVID-19, Amanda Brimmer, Vincent Chin et al., BCG, 16th April 2020

[65] Australia and New Zealand reopen after coronavirus cases plummet, Rebecca Falconer, AXIOS, 12th May 2020

[66] Data strategy and roadmap for New Zealand, data.govt.nz, 2020

[67] China’s Coronavirus Back-to-Work Lessons: Masks and Vigilance, Alexandra Stevenson & Cao Li, The New York Times, 12th May 2020

[68] Keeping Economic Data Flowing During COVID-19, IMFBlog, 26 May 2020

[69] Renewed outbreaks in South Korea, Germany and China show continued risk as more countries seek to reopen, James Griffiths, CNN, 11 May 2020

[70] Israel hits “emergency brake” on reopening as coronavirus cases rise, Kisa Archeta, 8 June 2020

[71] https://time.com/5868098/schools-reopening-coronavirus-denmark-south-korea-israel/

[72] Coronavirus cases surge in Pakistan after PM gambles on reopening the economy, Charlotte Greenfield & Umar Farooq, Reuters, 5 June 2020

[73] Lockdown over, Pakistan’s COVID-19 deaths, infections tick higher, Reuters, 21 May 2020

[74] Beyond the Curve: How to Restart in the Wake of COVID-19, Amanda Brimmer, Vincent Chin et al., BCG, 16 April 2020

[75] Schools reopening has not triggered rise in COVID-19 cases, EU Ministers told, Guardian, 18 May 2020

[76] How Denmark sent children safely back to school, ITV, 20 May 2020

[77] COVID-19: Countries around the world are reopening their schools. This is what it looks like, WEF, 2 May 2020

[78] Schools reopening has not triggered rise in COVID-19 cases, EU Ministers told, Guardian, 18 May 2020

[79] Data in the time of COVID-19, Open Data Watch

Previous page Next page
Contact Us

Find out more about Delivery and Pemandu Associates' work on Global Lessons in Tackling COVID-19. We would love to hear from you, whether you have questions or simply want to learn more about how we could work with your organisation.

Please complete the form below to get in touch, and we will do our best to send you a prompt response.

Wish to connect with us?