In response to the first confirmed cases of COVID-19 in the United States, President Trump restricted most travel from China on February 2, 2020, while exempting U.S. citizens and their families. The next month, he expanded it to include Iranians and most Europeans. By the end of March, the U.S. State Department largely ceased issuing visas to anyone.1 These bans have remained in effect even as the United States has become the leading source for COVID-19 infections worldwide.2

President Trump and his advisors repeatedly touted these actions as demonstrating that they would keep the country safe from the virus.3 However, a review of the research about pandemics prior to COVID-19 with disease profiles similar to influenza shows that the government should have known how ineffective travel restrictions would be at protecting the country.

The research available in January 2020 shows that extreme travel restrictions imposed before a pandemic reached the United States could have theoretically delayed its introduction by a few days or weeks, but that travel restrictions like those the Trump administration did impose could not stop the spread of a disease like COVID-19, even in theory. Rather than pinning all of those hopes on a Chinese travel restriction, the U.S. government should have spent those early moments preparing a vigorous domestic response to the virus—which it failed to do.4

Introduction

Cato’s Alex Nowrasteh has reviewed the research regarding travel restrictions and COVID-19, concluding that “travel restrictions likely delay the spread of COVID-19 in modeled simulations by up to a few weeks.”5 The COVID-19 studies so far have focused mainly on China’s exit restrictions and restrictions imposed by other countries like Japan and Australia.6 Importantly, those papers present theoretical models and do not empirically test their predictions against real-world results.

The following post reviews a dozen pre-COVID-19 studies that analyze the effects of international travel restrictions—or international and domestic travel restrictions—on influenza pandemics globally or in the United States.7 Given the similarities to COVID-19, the influenza research would have been the most relevant research available to the government in January 2020.

A broader view of the pre-COVID-19 evidence supports the World Health Organization’s (WHO) position that travel restrictions are only effective before a pandemic has spread and only serve to delay the spread by a few days or weeks.8 Multiple meta-analyses and reviews by the WHO, U.S. Homeland Security Council in 2006, and United Kingdom (UK) Department of Health in 2018 also back this conclusion.9

The consensus of researchers is that stopping all travel is impossible, unenforceable, and politically unrealistic—an assumption that held in the COVID-19 case—and that if it were even possible to selectively stop it (e.g. excluding all nationals of an affected country), doing so will only delay transmission until the disease spreads beyond the initial selection10 (such as to visitors to the country or to other countries).With any travel, the risk of infiltration grows proportionately to the spread abroad such that even a “tenfold reduction in numbers of visitors delays arrival of infection for approximately as long as it takes global prevalence to increase tenfold to compensate.”11

These dozen studies are unanimous that anything less than an immediate total travel restriction has only a “modest,” “little,” “low,” “negligible,” or no effect at all on the spread of a pandemic.12 If a ban delays the introduction of the virus into the high flu season, it can actually worsen a pandemic.13 Even the most extreme restrictions on travel delay the introduction or progression of flu pandemics by at most a few weeks. For context, the U.S. Homeland Security Council in 2006 expected that pandemic influenza would reach the United States “within 1 to 2 months,” which happened in the case of COVID-19.14

The research also indicates that there is no benefit to international travel restrictions once an outbreak has already become an epidemic inside the destination country.15

Research on Pandemic Spread

  1. A World Health Organization expert consultation (1959) reviews the evidence relating to the effects of travel restrictions from the 1957 flu pandemic.16 It finds that the “most striking example of the possibilities of quarantine was Israel where the restriction or complete absence of international travel between that country and neighboring States (due, not to deliberate quarantine restrictions, but to the political situation) seems to have resulted in a delay of about two months in the onset of the epidemic.” It finds “less convincing” evidence that showed that South Africa’s shipping restrictions may have caused “some delay” in the pandemic there, but in all other countries, “no effect was detected.” They conclude, “If such measures are to be effective, they must be very severe…. a high price to pay for a few additional weeks’ freedom from the disease. Such action could only be justified on technical grounds if the extra time permitted the application of effective specific prophylactic measures.”
  2. In Plos Med, Brownstein, et al. (2006) use the natural experiment of the flight ban and subsequent 27 percent decline in international and domestic air traffic following the September 11, 2001 terrorist attack to evaluate the effect of air travel on the flu season. They find that “the decrease in air travel was associated with a delayed and prolonged influenza season,” with a the epidemic peak 13 days later than average.17 They conclude, “Though our results suggest a possible benefit of airline travel restrictions, without early detection and immediate action, such measures may be ineffective at stemming the spread or mitigating the impact of an oncoming pandemic.”
  3. Bajardi, et al. (2011) model the 2009 H1N1 pandemic’s growth from Mexico in PLoS Med and find that the “40% drop in travel flows [to and from Mexico] observed in reality only led to an average delay in the arrival of the infection in other countries (i.e. the first imported case) of less than 3 days.… too small a magnitude to impact the international spread.”18 They use this model to test the effect of a 90 percent reduction and find that “the resulting delay would be on the order of 2 weeks.” They conclude, “It is unlikely that given the ever-increasing mobility of people travel restrictions could be used effectively in a future pandemic event.”
  4. Hollingsworth, et al. (2006) use a simplified global model in Nature Medicine to estimate that for an influenza pandemic creating 100 cases per day, “travel reductions of the order of 80%, for example, only increase the interval between exports [from the affected countries] by days.” Increasing the restrictions to greater than 99 percent raise the interval “to the order of weeks” even with additional exit screening that excluded all symptomatic passengers.19 They summarize, “Once case numbers exceed a few hundred in source areas, only rapidly implemented and almost total restriction of international travel can prevent the export of cases and the triggering of new epidemics in unaffected areas.” They conclude, “Our analysis also indicates that restrictions on travel will be of limited benefit in slowing global spread of a pandemic influenza outbreak that is not contained at its source. There may be a role for travel restrictions applied to the source country while containment efforts are underway.”
  5. Cooper, et al. (2006) model a flu pandemic in PLoS Med and find that “even large and widely enforced travel restrictions would usually delay epidemic peaks by only a few days; to have a major impact, restrictions would have to be almost total and almost instantaneous.”20 They estimate that even a 99.9 percent reduction in air traffic would not protect most cities from a pandemic. They conclude, “Interventions to reduce local transmission of influenza are likely to be more effective at reducing the rate of global spread and less vulnerable to implementation delays than air travel restrictions.”
  6. In Proceedings of the National Academy of Sciences, Germann, et al. (2006) model the U.S. response to an influenza outbreak that has already seeded from abroad.21 They find “a total reduction in long-distance travel, to 10% of normal frequency” would “slow the virus spread by only a few days to weeks (depending on R0), without reducing the eventual size of the outbreak.” Indeed, their simulations indicate that long-distance travel restrictions do not reduce the total number of infections.22 Table 1 below shows their results for travel restrictions compared to other interventions. They conclude, “travel restrictions alone do not appear to be an effective control strategy, due to the implausibly early and drastic measures required to significantly reduce the large number of local outbreaks that are likely to emerge around the country.”
  1. Epstein, et al. (2007) model a flu pandemic in PLoS ONE and find, “When travel restrictions are imposed, the [First Passage Times to the U.S.] increases by two to three weeks when the outbreak originates in Hong Kong (from 18 days to 31 days) or Sydney (from 7 days to 27 days).”23 This paper’s important contribution is testing the effect of seasonality on the efficacy of a travel restriction. They write, “if travel restrictions are not combined with other measures, local epidemic severity may increase, because restriction-induced delays can push local outbreaks into high epidemic season.” They conclude, “From a public health perspective, it becomes clear that the main purpose of travel restrictions is to delay dissemination of the disease until targeted medical and other interventions can be developed and deployed.”
  2. Ferguson, et al. (2006) model an influenza virus spread in Nature and show that “a 90%, 99% or 99.9% reduction in imported infections might delay the peak of the US pandemic by 1.5, 3, or 6 weeks, respectively.”24 They note for context that SARS reduced Hong Kong travel by 80 percent. They estimate that a tenfold reduction in travel translates to a 12.5‑day delay in epidemic progression. They conclude, “For the high transmissibility scenario only policies using >99% border controls can delay spread by enough to enable pandemic vaccine to reduce attack rates substantially (vaccine is assumed to be produced from month 4 of the global epidemic).”
  3. Colizza, et al. (2007) model avian flu pandemic across 220 different countries using airline data in PLoS Medicine.25 For a pandemic starting in Hanoi in October, they find that halving the number of air travelers from the infected area does “not appear to be efficient neither in reducing the epidemic impact, nor in delaying the prevalence peak,” nor does it “contain the pandemic at the source.” They add that the moment that the epidemic has spread from its source, “the travel limitations are no longer able to considerably slow down the pandemic evolution.” They conclude, “We showed that travel restrictions, which are both economically disruptive and difficult to implement, achieve very modest results, slowing down by only a few days or weeks the overall evolution of the pandemic.”
  4. In Vaccine, Flahault, et al. (2006) model a pandemic influenza outbreak starting in Hong Kong on 52 cities, which would reach all 52 cities in 5 months without interventions.26 A 50 percent reduction in air traffic “bought an average delay to reach epidemic status of 9 days compared with the baseline scenario.” A 60 percent restriction in combination with isolating 10 percent of cases would add 19 days to the baseline transmission estimate but the epidemic peak would be higher.27 Figure 1 shows the effects at varying degrees of travel restrictions. Only a 100 percent travel ban has a significant effect on the global attack rate, and even a 90 percent closure delayed the epidemic by less than a month.
  1. Tomba and Wallinga (2008) model infectious disease transmission in Mathematical Biosciences.28 They use a “simple formula” to demonstrate that with an R0 of about 2 and a generation time of 2.6 days, average delays caused by reductions of importation by 90%, 99% and 99.9% are … 11.5, 23.0, and 35.0 days, respectively.” They state that they focus on the reductions in first importation because “later importations will of course occur, but these will have a minor importance.” They describe this formula as a “simple explanation for the low impact of border control as a countermeasure to the spread of an infectious disease.”
  2. Keirnéis, et al. (2008) model pandemic influenza progression on 52 major world cities in PLoS ONE, varying their estimates based on the profile of the virus (e.g. “a fast, massive pandemic and a slow building, long-lasting one”).29 They find a slight correlation between very early travel restrictions and the number of cities that the pandemic reaches but state, “Regardless of the [virus] profile, restricting air travel (either expressed by the proportion and the date of introduction of transport limitation) had no impact on the global burden of the pandemic.” Compared to other interventions, they find that “travel restriction was the only measure without any measurable effect” on pandemic outcomes. They conclude, “Our results also suggest that travel restrictions would have a limited impact on the spatial and temporal diffusion of an influenza pandemic.… Such restrictions have significant logistical, ethical and economic implications and their impact on an influenza pandemic is currently debated.”

Conclusion

The pre-COVID-19 research is unanimous that governments cannot expect to rely on travel restrictions to prevent the spread of pandemics similar to influenza. Travel restrictions do not prevent the spread of disease and may only delay it for a few days or weeks if implemented prior to the international transmission of the disease. The Trump administration’s travel restrictions waited until after the virus had already entered the United States, and they exempted many travelers from China, not to mention the rest of the world.30

The research shows that the Trump administration should have known that it needed to focus on domestic measures to stop the disease, not international ones, yet it did not.31 Cato’s Alex Nowrasteh argues that the “travel ban created a false sense of security that delayed other more effective methods of containing COVID-19, such as instituting social distancing sooner.”32 In 2014, the CDC has also cited the concern that travel bans could incentivize foreign governments not to report or underreport cases.33 While its motivations remain unclear, China did in fact downplay the extent of the crisis in both the case of SARS and COVID-19—decisions that remain far more damaging than any other.34

The research also shows that the international travel restrictions have virtually no effect today once the outbreak has become an epidemic here. There is no basis in the research for continuing a blanket ban on travel between countries.