Skip to main content
Blog Post

When will the COVID-19 pandemic end in the U.S.?

Insurance Consulting and Technology|Reinsurance
COVID 19 Coronavirus

By Dave Ingram | December 11, 2020

Even with availability of a vaccine, individual behavior will remain a major factor in community transmission and when we can return to normal.

Since July, we have seen the lows and the highs of the COVID-19 pandemic. In mid-September, active reported COVID-19 infections in the U.S. fell to below 160 per 100,000. But by mid-November, we passed 500 per 100,000 and have kept climbing. COVID-19 fatigue, school and business re-openings, cold weather, and lines for voting have all been blamed for the run-up.

At the same time, we received news that not one, but two (and maybe three) vaccines were seeing over 90% efficacy rates in stage 3 testing.

With that in mind, we’ve updated the COVID-19 projections that we did four long months ago. Besides having four additional months of data, we have also learned that COVID-19 is more cyclical than linear: Times with good experience are followed by times with poor experience and vice versa.

As we look into the new COVID-19 projections, we learn that even with a widely available, highly effective vaccine, community behaviors to slow the spread of the pandemic – such as mask wearing, maintaining social distance and avoiding indoor gatherings – will remain of utmost importance for the next six months.

New infection rate

In our previous article exploring different scenarios for reaching herd immunity, we defined the new infection rate (NIR) as today’s new infections divided by the new infections from the prior 14 days. The following shows the average NIRs for the past seven months.

New infection rates of COVID-19 from April to October 2020 ranged from a high of 8.6% in April to a low of 6.4% in August.
New infection rates – monthly average

Percentage of COVID-19 NIR from April 2020 to October 2020

The average for the entire seven-month period from April through October was 7.6%. If we imagine that the NIR will remain cyclical, then a future NIR near the average of 7.6% seems like a reasonable estimate. Looking at the range of monthly experience, it also seems reasonable to believe that the future average NIR will fall between 6.5% and 8.5%.

The NIR is driven by the following:

  1. The natural transmissibility of the disease
  2. The social customs of a population that may result in local variations in actual transmissibility
  3. Community efforts to reduce the transmission of the disease
  4. Immunity to the disease in the population from people who were infected and recovered
  5. People who are immune because they received an effective vaccine

These factors (except possibly the first) are all changing over time. In our model of the pandemic, we set an assumption for the future base NIR and the model layers in future changes due to the immunity factors (4) and (5) that will gradually reduce the observed NIR.


Two, possibly three vaccines have been in the news touting stage 3 test results that show efficacy of over 90%. At least one of these may be available by January. The U.S. government’s Operation Warp Speed promises to deliver vaccines to distribution points in each state as quickly as possible. For the following projections, we looked at possible speed of delivery of effective vaccines.

Vaccine delivery speed and immunization

Four possible speeds of delivery of effective vaccination over the next 12 months
Speed of delivery Percentage of population vaccinated
and immune over next 12 months
Low 12%
Medium 27%
High 50%
Warp speed 78%

With the news of two or three vaccines with 90% or higher efficacy, we will drop the “Low” speed of delivery assumption.

So with three assumptions for new infection rate and three assumptions for speed of delivery of vaccine, we have nine scenarios for the future of the pandemic. These projections do not reflect the potential impact of an overwhelmed health care system.

Nine projections for the future of the pandemic

Possible scenarios through 2021
Cumulative New
Transmission Vaccine Cases -
Dec 31
6.50% Medium 12.1 M 1.6 M 0.3 M 0
6.50% High 12.1 M 1.5 M 0.2 M 0
6.50% Warp speed 12.1 M 1.4 M 0.1 M 0
Total deaths 300 K
7.50% Medium 14.8 M 10.2 M 6.0 M 1.1 M
7.50% High 14.8 M 9.7 M 4.3 M 0.3 M
7.50% Warp speed 14.8 M 8.6 M 1.5 M 0
Total deaths 400 K –
500 K
8.50% Medium 20.2 M 46.0 M 22.8 M 1.3 M
8.50% High 20.2 M 43.4 M 17.4 M 0.4 M
8.50% Warp speed 20.2 M 38.0 M 7.1 M 0
Total deaths 900 – 1.2M

In all nine scenarios, the highest incidence of new cases occurs in the second quarter of next year. And in all nine scenarios, the pandemic is over or has subsided to a very low level by Q4 2021.

But the most important thing to note from the nine scenarios is the extreme importance of community behavior, which drives the differences in transmission rate.

Actual infections have already exceeded the level projected for the end of 2020, which shows that using the 6.5% NIR projection aligned with the NIR seen in May and August is unrealistically optimistic. A figure of 7.5% NIR, similar to the average over the past seven months, indicates we could see 10 to 20 million new infections above the roughly 14 million currently reported, and 130,000 to 230,000 additional deaths beyond the 270,000 lives already lost.

But with community behavior that follows the recent experience in October and early November (near 8.5%), we see up to 40 million additional infections and 600,000 to 900,000 more deaths.

On the other hand, the speed of the vaccine’s delivery has the effect of reducing infections by 7 million at the 7.5% NIR and 25 million at the 8.5% NIR.

From these projections, it appears that community behavior is almost twice as important as the speed of delivery of the vaccine.

Drivers of transmission

Four transmission routes
Four transmission routes

Community behavior remains highly important

When an infected person sheds active virus in can result in

  1. effective transmission with new infection,
  2. Ineffective transmission due to other causes,
  3. Ineffective transmission due to mitigating behaviors or
  4. Effective transmission but to a person who is immune due to being infected and recovered or vaccinated.

During the early stages of a pandemic, when an infected person sheds active virus, there is either an effective or ineffective transmission. At that point, people in the community may have little to no awareness of the existence of the virus. Once the virus becomes known, people will typically start to adopt their own mitigating behaviors and, eventually, public health officials and others in public positions will start to promote specific mitigations that will, over time, change as the most effective mitigations become known. Such mitigations will lead to ineffective transmissions.

As the disease progresses through the population, some of the otherwise effective transmissions will not result in new infections because the person on the receiving end is immune. At first this will be due to some people having recovered from prior infections and developing immunity naturally. At the current time in the U.S., even with the reporting of daily records of infections, this is still not likely to be a significant factor. Prior reported infections in the U.S. were less than 3% of the population at the end of October. Finally, when a vaccine is available, it will mean that more people will be immune.

To look at this graphically, we start with a hypothetical situation where there are no future mitigating behaviors and no future vaccine, and the NIR runs very high at 12%. Under this unrealistically worst-case scenario, there would only be the impact of immunity from prior infections. That pandemic would follow a pattern like the graphic below.

If there is no immunization, and the NIR is at 12%, then infection and recovery become the leading reason for blocked transmission and reaches a plateau in March.
Blocked transmission of infection without immunization, at 12% NIR

If there is no immunization, then infection and recovery become the leading reason for blocked transmission and reaches a plateau in March.

We can compare that with one of the projections in the middle of our set of nine at 7.5% NIR.

With a medium vaccination rate and an NIR of 7.5%, mitigating behaviors like masks become the primary means of blocking transmission of infection until October when immunization and behaviors are nearly equal.
Blocked transmission of infection with medium rate of immunization, at 7.5% NIR

With a medium vaccination rate, mitigating behaviors like masks become the primary means of blocking transmission of infection until October, when immunization and behaviors are nearly equal.

And for our final comparison, we look at the 8.5% NIR and warp speed vaccination.

With a warp speed vaccination rate and the NIR at 8.5%, new infections will plateau in May, but mitigating behaviors can end in July.
Blocked transmission of infection with warp speed immunization, at 8.5% NIR

With a warp speed vaccination rate, new infections will plateau in May, but mitigating behaviors can end in July.

You can see from this illustration that even with the warp speed rate of vaccinations, behavior is the leading factor for the next six months until the number of vaccinated people starts to become a major factor in blocking transmissions.

Comparing projections

In our July projections, we showed the impact of 6%, 8% and 10% NIRs on the course of the pandemic. With the actual NIR since then shown above, you can see that the 10% NIR was not at all close to the emerging course of the disease. But the actual course of the disease did fall between the 6% NIR projection and the 8% NIR projections from July.

Actual cases of COVID-19 were around 9 million in Oct 2020, trending closer to 6% NIR that was projected in July 2020.
Actual versus projected total COVID-19 cases

Actual cases of COVID-19 trended closer to the 6% NIR projected in July 2020

We called this approach to modeling the pandemic a bootstrap approach. In this approach, the daily new infections are projected; the NIR is the observed rate of the daily new infections that have occurred in the past compared to the active infections. This will lead to a realistic projection if community behavior in the future is comparable to community behavior in the past. In addition, this model can reflect changes to the future observed NIR that result from future growth of the part of the population that is immune. Below is an example of that effect for the 8.5% NIR, warp speed immunizations scenario.

Warp speed immunizations scenario

Possible new infection rates through October 2021
new infection rate
new infection rate
2020 October 8.5% 7.95%
2020 November 8.5% 8.46%
2020 December 8.5% 8.30%
2021 January 8.5% 7.92%
2020 February 8.5% 7.29%
2020 March 8.5% 6.45%
2020 April 8.5% 5.52%
2021 May 8.5% 4.64%
2021 June 8.5% 3.81%
2021 July 8.5% 3.01%
2021 August 8.5% 2.19%
2021 September 8.5% 1.38%
2021 October 8.5% 0.58%

Hang in there

The good news about the vaccines is definitely cause for celebration, but under any realistic scenario for actual take-up by the U.S. population, it will be up to six months before the vaccinations have a significant impact on the course of the pandemic. During that time, it is critical that we all continue with the mitigating practices that have been in place around the country. Getting back to normal will only happen after the new cases of COVID-19 have been driven down to a near zero level of infections.


Head of Willis Re ERM Advisory

Contact Us

Related Solutions