Could the United States have weathered the Covid-19 pandemic with fewer fatalities if only we adopted a different suite of policies based on countries that have shown greater success at controlling the disease?
We often see this claim repeated in the press, specifically invoking the examples of Australia and New Zealand to argue that the United States failed to properly lock down. Both Pacific nations have thus far escaped with relatively few internal Covid outbreaks, as have many other countries around the Pacific Rim.
Japan, South Korea, and Taiwan often join Australia and New Zealand on the list of “model” responses, with the media and public health officials such as Anthony Fauci touting their examples while also scolding the US for insufficiently locking down. As Fauci stated in an interview from late October:
“Australia is one of the countries that has done actually quite well, I believe…New Zealand has done well. Some of the Asian countries have done well … I would like to say the same for the U.S., but the numbers speak for themselves.”
Such claims have been commonplace since relatively early in the pandemic. Time Magazine ran an article back in June crediting Australia, New Zealand, South Korea, and Taiwan for their successful responses. Devi Sridhar, a health anthropologist who has become a prominent pro-lockdown voice in the UK media, frequently invokes New Zealand and the Asian countries with low case counts as examples for Britain to emulate.
The John Snow Memorandum (JSM), a petition drafted by lockdown supporters in the public health profession to counter the Great Barrington Declaration, similarly credits “Japan, Vietnam, and New Zealand, to name a few countries, that have shown that robust public health responses can control transmission.” The United States and Europe would do well to copy and adapt their policies, it follows, which in the eyes of Fauci and many of the JSM signers usually means more aggressive lockdowns.
There’s a fundamental problem however with statements invoking the Pacific “success stories” as a model to copy: there is no Pacific model, but rather a half-dozen countries with completely different policy responses.
The American press seems particularly keen on pushing for the replication of Australia and New Zealand-style responses. Both countries adopted aggressive lockdowns to “eliminate” the virus – although, as we’ve seen numerous times now, these ‘Prince Prospero’ strategies depend on maintaining tight border restrictions, which makes them inherently fragile whenever a case slips through by accident or human error.
South Korea, Japan, and Taiwan have generally avoided heavy-handed lockdowns, although the advocates of lockdown-style restrictions seldom urge replication of their policies even when invoking their successes. By contrast, my colleague Pete Earle suggested the world should copy South Korea’s approach all the way back on March 12 – the same week that Fauci in the US and Neil Ferguson in the UK threw their weight behind a lockdown strategy explicitly modeled after the draconian suppression measures employed by the Chinese government in the Wuhan province.
In practice, the five Pacific “success stories” may have similar results thus far in limiting outbreaks. But they also employed five completely different policy responses to the virus. One would not know this by reading the remarks of Fauci or the many commentators who invoke these countries, only to stress the supposed propriety of locking down again. Although the Oxford Stringency Index remains a deeply imperfect measure of lockdown severity, it nonetheless captures the extreme policy variation in the Pacific “success stories” when compared to the United States:
When a lockdowner accordingly invokes Australia, New Zealand, South Korea, Taiwan, and Japan as examples of successful pandemic responses, the appropriate inference is not the claimed vindication of lockdowns but rather to ask which one of these five countries we should copy and why, seeing as their responses are completely different. They range from frequent heavy lockdowns to almost no lockdowns at all, to say nothing of numerous smaller differences in testing, local restrictions, public health recommendations, school closures, and border quarantine policies.
When evaluating the Pacific “success stories” we must also keep in mind that other factors besides policy likely explain how they have fared during the pandemic. The most obvious is geography. Four of the five are islands, and the fifth – South Korea – sits on a geographically isolated peninsula.
Note that geography alone does not fully explain their low outbreak numbers, but it does appear to help with one characteristic that most of these locales share: all five enacted varying degrees of external border controls on international visitors. Crucially, they also appear to have done so before the virus became widespread within their borders.
To the extent that such strategies worked for the Pacific “success stories,” it is simply not replicable for much of the rest of the world. Countries in Europe and the United States also tried border closures without success, mainly because the virus had already arrived and reached widespread undetected transmission back in January and February. In the case of Australia and New Zealand in particular, their comparative remoteness at the time of the border closures likely had more to do with their subsequent courses in the pandemic than any internal lockdown measure.
A narrative nonetheless persists among journalists and lockdown supporters that credits Australia and New Zealand for showing us the “right” way to control Covid. While we’re past the point of implementing their border strategies, so this narrative goes, we may still copy their lockdowns – and do it “correctly” this time until the virus is fully suppressed and eliminated.
By implication, these health policy experts contend, the ineffectual and costly lockdowns that swept through the United States and Europe last spring only to be revived in the fall wave were not “true lockdowns” – an argument that bears striking similarity to the line used by those who insist that the Soviet Union was not “true socialism.”
They were lockdowns with too many holes in them, too many exceptions, and too many people breaking their terms by venturing out into public. Therefore we must try locking down again and do so even more aggressively, because this time will somehow yield different outcomes than the last two times that the lockdown approach failed to meaningfully alter the course of the pandemic.
Such thinking carries an implicit – indeed axiomatic – assumption that lockdowns necessarily work as intended when done “right,” and the only reason that their previous iterations failed to reverse the course must stem from faulty implementation. So who did it “right,” and who could show the rest of the world a lockdown model to replicate?
New Zealand’s appeal remains high among the lockdowners, although when pressed they will usually concede that it is also the most remote of the Pacific “success stories.” As a nation of less than 5 million with few large cities and a comparatively low international air traffic even before Covid, its position remains highly dependent on its geography-assisted border closure. Indeed, the epidemiology literature has long recognized that border quarantines can be effective at staving off influenza pandemics in small island nations where almost all international travel comes through airports. Most of the world does not meet these conditions though.
In the last few months the comparatively larger Australia has emerged as the new favored model. Although its population is only 25 million and its air traffic is still low compared to London, Paris, or New York, Australia still has several large cities, a comparable level of economic development, and a large dispersed geography (Australia’s 3 million square miles of territory is only slightly smaller than the continental United States).
Australia’s Covid outbreaks have also occurred in regional waves, mostly around Melbourne and Sydney, which the lockdowners now tout as examples of “successful” regional suppression strategies that might be adapted to American or European cities. For example, a recent Washington Post article proclaims that Australia “has almost eliminated the coronavirus – by putting its faith in science.” Numerous other stories of this type appeared in early December as something of a victory lap after Melbourne relaxed its lockdowns following a rapid drop in its cases (although the virus reemerged around Sydney only two weeks later, once again confirming the fragility of this roving lockdown strategy).
Yet one other aspect of Australia’s lockdown strategy has gone almost entirely unnoticed amid the touting of its approach as a model for the rest of the world. Far from revealing a superior and more effective way to conduct a lockdown, Australia’s strict lockdown policies have had no more actual effect upon its population’s willingness to venture into public than those employed in the United States.
This surprising pattern becomes clear when we examine anonymized GPS data from smartphones as released by Google for each country. The data show percentage changes in mobility as measured against a pre-Covid baseline of “normal” mobility from January.
Certain Google mobility indicators are particularly well-suited for estimating the combined effects of (a) voluntary behavioral pattern changes in response to the pandemic and (b) restrictions such as lockdowns. Although it is difficult to completely separate these two factors since they often occur simultaneously, their cumulative effect appears in the rates at which people venture out into public for shopping and similar activities and the amount of time spent sheltering in place at home. Mobility data therefore give us a real-time indicator of what people are actually doing during a lockdown, as distinct from what the lockdown policy purports to do.
A curious pattern emerges when we compare two key mobility metrics for Australia and the United States. First, Google’s “retail and recreation” index measures patterns of people venturing out into public spaces. Between the start of measurement in February and the end of June – a period that encompasses the spring lockdowns in both countries – Australia and the United States exhibit virtually identical mobility patterns.
Retail and recreation dropped off precipitously in both countries at almost the exact same rates and levels. Indeed, the only visible difference is that the United States started to drop about one week before Australia. Both drops were due to a combination of uncertainty-induced voluntary behavior and the imposition of lockdowns in mid-March (43 of 50 American states and nationwide in Australia).
Yet the earlier drop in the United States belies the ubiquitous claim that Australia “got ahead of the virus” through proactive policy, as opposed to combinations of geography, seasonality, and sheer luck.
An even more interesting pattern emerges when we look at the data after July 1st. By mid-summer, the reopening in the United States stalled (especially compared to Europe) and Google’s retail mobility index basically plateaued, where it has remained ever since (excepting a few brief spikes and drops, coinciding with holiday travel).
Australia, by comparison, had a second albeit much more modest drop between mid-July and mid-October (almost exactly coinciding with the regional lockdowns in Melbourne and larger Victoria). After late October however, Australian mobility surged past the United States and has remained so ever since, subject to the above-noted holiday fluctuations.
A directly corresponding pattern may also be seen in Google’s index of time spent in one’s own residence – an approximation of how much the residents of each country are sheltering at home. Once again, the US and Australian patterns display an almost perfect match through the end of June.
Once again, Americans started sheltering at home about a week before Australians adopted the same behavior. The patterns still peak at the same time and level, and still closely track as the countries begin to reopen. From July to October, the pattern of Australians sheltering at home slightly overtakes the United States, corresponding with the Victoria regional lockdowns. From November until the end of the year though, the United States also overtakes Australia on the residential mobility index – a period that also coincides with the reimposition of fall lockdowns in several American states.
These patterns each warrant closer investigation including within specific states in both countries that locked down. At the national level however, the United States and Australia show far more similarities in their mobility patterns than differences.
If Australia locked down the “right” way – as the media narrative of the moment holds – one might reasonably expect to see two very different patterns, with Australia reducing its mobility to the home both earlier than the United States and to a much greater degree. Instead, we see the US shutting down slightly earlier and the patterns closely matching thereafter. If the posited benefits of lockdowns come from causing a reduction in public interactions as their epidemiological advocates claim, then the lockdowns in both countries achieved strikingly similar reductions. Yet one country’s outbreak continued to surge to current levels while the other country’s case totals remained comparatively modest.
Those wishing to use Australia as a “model” for the United States would be wise to investigate other causes besides lockdowns that explain their differences. They may also soon find that some of those causes are not only incapable of being replicated elsewhere – they also defy the same experts’ pretenses of actually controlling what they imagine they can design through a centrally planned suite of restrictive lockdowns.