Covid-19: The lockdown decision that taught Prime Minister Jacinda Ardern to love caution

Prime Minister Jacinda Ardern arriving for a post-Cabinet press conference. Photo / Mark Mitchell

"Out of an abundance of caution …"

For two years, that is how Government ministers have begun their sentences about new or prolonged Covid regulations. In many ways, caution has defined New Zealand's response to Covid-19.

As Anzac Day approaches, and with it the anniversary of one of the Government's signature cautionary decisions, it's worth re-examining.

By April 20, 2020, New Zealand had spent close to four weeks in the world's most stringent lockdown, according to the University of Oxford's Stringency Index.

That day, Prime Minister Jacinda Ardern took to the Beehive podium and told the country it would spend an additional five days in harsh lockdown. Level 4 would lift, not on April 22 as expected, four weeks after the restrictions were imposed, but rather at the close of Monday, April 27, following the long weekend.

Shaun Hendy, the University of Auckland physics professor whose team had provided modelling for the Government's decision, told Radio NZ that morning that his preference was for an extension of Level 4 by a full two weeks, but that lifting on the 27th split the difference.

Just hours later, then National leader Simon Bridges criticised the decision.

The extension, he said, showed a lack of preparation by the Government to have the country ready to ease restrictions. He said he feared the harm from staying in lockdown would be greater than from easing more quickly.

It was a fateful position to take. Bridges attracted widespread criticism: in the media, from commentators and from the public online.

Many within his own party made their annoyance known, and his outspoken response appeared to play an important role in his ouster, following sinking poll results, just a month later. The subsequent 18 months were defined by disarray in National's ranks.

But time, and new information, casts Bridges in a more flattering light.

Critically, it is now clear that the Prime Minister herself was in possession of advice that broadly mirrored Bridges' position.

A second modeller, Rodney Jones of Wigram Capital, was also providing advice directly to the Prime Minister's Office. His summary on April 19 was this: "The Hendy simulation does not provide a basis for delaying an exit from Level 4 until April 27. The key conclusion from [Hendy's] simulation is that it is critical that Level 3 is effective."

In his correspondence with the Prime Minister's Office (at times responding explicitly to "the PM's questions"), Jones wrote: "The Hendy model systematically underestimates the impact of interventions. In the case of Level 4, on 8 April their draft model was projecting around 2330 cases by 18 April, a number that was revised down to 1780 in the final paper on 9 April. As we know, the final number was 1409."

Jones' point was that the effectiveness of interventions was poorly understood, and that "models and simulations" were only useful to a point. The graphs of Google-tracked population-level activity that he supplied showed a 70 per cent drop in movement in New Zealand at Level 4, a much deeper drop than in other countries which also pursued elimination at the time. In Australia, activity was dented by less than 50 per cent, and the drop was more shallow still in Taiwan and Hong Kong.

"Decisions around exit have to be carefully balanced and take into account the substantial economic and social costs," Jones argued. "The gain from delaying until April 27 is marginal at best, and certainly does not justify the additional cost of delay."

What that cost amounted to was subject to remarkably limited debate.

In a subsequent cost-benefit analysis (using only data available at the time the decision was taken) the Productivity Commission calculated a net cost of $750 million for the Level 4 extension.

By Anzac Day 2020, New Zealand had spent close to four weeks in the world's most stringent lockdown, according to the University of Oxford's Stringency Index. Photo / NZME

Perhaps unsurprisingly, that work met with harsh criticism from Hendy and his group of modellers at the University of Auckland-based research centre Te Pūnaha Matatini (TPM). The Government ultimately spent and contracted to spend more than $6m for Covid-19 modelling from TPM, eclipsing spending on all other such modelling work.

Hendy maintained that a fundamental error in the commission's work had failed to properly take account of the dynamics of virus spread at very low levels of disease, and he wrote to the commission and requested it issue a press release retracting the paper's findings.

The commission, however, stood by its work and the paper remains published on its website.

Hendy and colleagues ultimately undertook their own cost-benefit work, which considered a second situation: the use of Level 3 to eliminate Covid during the August, 2020, Auckland outbreak.

The TPM work found that using Level 4 would have achieved elimination at either a comparable cost to Level 3 or at a lower cost. However, the work has never been peer reviewed and it did not consider the social costs of greater restriction.

The upshot is that the benefits of the Level 4 lockdown and its Anzac Day extension remain in question, but the political fallout does not.

The Prime Minister learned that an abundance of caution cost her nothing, whatever the monetary price.

And National found that an apparent absence of caution, despite sound reasoning, can still be clobbered by emotion.

The Treasury, whose job it is to conduct cost-benefit analyses, has not undertaken a single such piece of analysis in relation to Covid-19 decision-making.

Indeed, it is clear that cost-benefit analysis is not a tool that Treasury's political master (and Ardern's right-hand man), Finance Minister Grant Robertson is keen to deploy.

The Government subsequently leaned heavily and successfully (given its landslide election victory later in 2020) on the line that: "a strong public health response is the best economic response".

The statement was magnetically simple, and concise enough to fit handily into Twitter debates and the condensed news snippets that are often repeated on radio and television broadcasts.

It was also simplistic: there was a compelling argument to be made that Level 3 was a strong health response, and Level 4 was overkill.

It's notable that after the Government's initial use of Level 4 in March and April 2020, it managed to eliminate Covid-19 with Level 3 lockdowns thereafter; it wasn't until the arrival of the more transmissive Delta strain of the virus in August, 2021 that it resorted to Level 4 stringency again.

At that point it failed to achieve elimination, and caution, though still potent, has been losing its power ever since.