• Moderate lockdown measures that promote social restrictions have a high likelihood of reducing the physical death toll in the short-term:

It could reasonably be expected to reduce the death toll from COVID- 19 by about 12,000 if implemented for ~9 months. In addition, lower social interaction would also mean less traffic and hence fewer traffic deaths, saving about 550 lives. Finally, with lower pressure on the health care system, it is estimated that 3,800 more people with HIV will not die over the coming 5 years. In total, a moderate lockdown (social restrictions) will likely reduce the death toll by 16,350 over the next five years.

However, moderate social restrictions will also lead to less health outreach and more malnutrition, causing more deaths from malaria (3,700), from TB (4,000) and from child mortality (1,700) along with about 100 extra maternal deaths. There is evidence to suggest that total but unmodeled impacts from air pollution, non-communicable diseases, mental health, and from unemployment would be more likely to increase rather than decrease these extra deaths. In total, it is likely that the complications following a moderate social restriction policy would result in at least an extra 9,550 deaths over the next 5 years.

Total number of deaths avoided by a moderate lockdown (social restrictions) is likely to be around 6,800.

• In this scenario, most of the deaths avoided would be from relatively older individuals, while the additional lives lost would be from relatively younger individuals. Our estimates suggest moderate social restrictions would lead to 397,000 life years gained, but contribute to 423,000 life years lost. This means that moderate social restrictions may increase total number of life years lost. In other words, moderate social restrictions could lead to lower number of COVID-19 related deaths, but the life years lost could outweigh the life years gained due tomore youthful populations losing their lives over the medium-term.
  
• The social value of the net number of deaths avoided is $228 million ($61-394 million).
  
• Closing schools for 9 months means that each child will receive 9 months less education. This will make each child less productive in their adult years. In total, it is estimated that the social cost of closing schools for Malawi will be around $5.2 billion - the present value of income loss for 6 million children over the next 50 years. Given that the benefits from school closure cannot result in a higher number of deaths avoided than what would come from moderate social restrictions (6,800 deaths), the maximal benefit from this policy would be $228 million. This means that on current knowledge, a policy of school closures will leave Malawi much worse off with a net social cost of around $5 billion.
  

Moderate movement and livelihood restrictions would generate a higher economic cost, which would lead to a loss of $6.7 billion - the estimated present value of GDP loss over the next 30 years. Compared to maximal benefits of $228 million, it means that on current knowledge, a policy of moderate movement and livelihood restrictions will leave Malawi much worse off with an economic cost of around $6.5 billion.

• Valuing all the costs and all the benefits in economic terms, using Value of Statistical Life to convert deaths and avoided deaths, it means that on current knowledge the costs vastly outweigh the benefits from moderate lockdown that promotes social restrictions.

The COVID-19 pandemic presents policy-makers with difficult trade-offs. Based on this analysis, this report suggests three headline policies that balance out the need to contain COVID-19 with other concerns:

• Do not increase social restrictions more than what is already in place, to avoid escalating both health and economic costs that have lasting effects on the economy and livelihoods. The analysis has shown that impacts on the economy from moderate COVID mitigation strategy are likely to be very large and significantly higher than potential benefits. However, continuing with a series of preventative low-cost social restrictions - including ensuring physical distancing and non-contact greetings, cocooning of the elderly and vulnerable, restricting large gatherings and promoting hand washing - appear effective. Bottom line, to the greatest extent possible, avoid mass livelihoods disruption.
• Mitigate education loss and open schools when possible: The analysis indicates that school closures have a large cost in terms of lost future productivity that vastly outweighs even the most optimistic benefits. In the presence of school closures, authorities might consider non-face-to face education modes for all school-going children to avert the opportunity cost of reduced future productivity of students especially for the girl child in a country where early marriages affect 42% of girls. However, this is likely to be a challenge for Malawi where most families cannot manage to support their children to access such modes of education especially in rural areas. Therefore, opening schools, even in phases that consider exam-writing classes amidst strict COVID-19 containment measures, should be a top priority at the earliest opportunity.
• Keep key community health services funded and operating: The analysis suggests that there is a risk of significant loss of life associated with the disruption to health services. To ensure that long-term health and economic welfare is not heavily compromised, at a minimum crucial areas should not see their funding cut when considering diverting resources towards fighting the COVID-19. These include child nutrition and survival programmes, maternal health, family planning, and support towards major infectious diseases like TB, HIV/AIDS and malaria.

Sources