Sweden vs. New Zealand, with Taiwan in the lead

Is the Covid pandemic and our collective responses to it best framed as health vs. the economy? That is, are countries forced to make a horrible trade-off between:

• minimising Covid infection rates at high economic cost; and
• accepting higher Covid infection rates, this being the cost of maintaining economic activity?

This framing underlies many of the popular narratives comparing the performance of individual countries. For example, many have contrasted New Zealand’s “elimination” goal to Covid with Sweden’s “mitigation” approach. These narratives typically characterise New Zealanders as valuing health over the economy, and Swedes as valuing their economy over health.¹

We are well into the pandemic now, so we can test this framing against the data.² I plotted countries with their expected economic performance in 2020 (vertical axis, best performers at top) against their performance in keeping Covid case numbers low (horizontal axis, best performers at left).³ I’ve highlighted New Zealand and Sweden.

What can we see in this graph?

• Covid is economically costly. No country is expected to do better in 2020 than it did in 2019.
• The worst position to be is the bottom right corner – many Covid cases along with a high economic cost.
• No country is in the top right corner – many Covid cases and a healthy economy.
• The best position to be is in the top left corner – few Covid cases and a healthy economy. Taiwan, South Korea and China, in particular, have done very well.
• Many countries fall towards the bottom left – few Covid cases but a significant hit to the economy.

What does the graph tell us about health vs. the economy? Well, if it really was a straight trade-off, then we would expect countries to perform well on either health or the economy, but not on both. So, they would cluster along the blue dashed line below. Clearly, they do not.

When the data doesn’t fit, it’s time to start looking for a new theory.

One alternative framing is get health right and the economy will follow. This is intuitively appealing. The policies required to control Covid outbreaks are clearly economically costly, so it follows that success in controlling outbreaks should reduce those costs. If that framing explained what was going on, then countries would cluster along the green dashed line in the diagram above. But the data isn’t a good fit for that framing either.

Tim Harford, a very savvy British economist, recently admonished commentators:

Forget the clash of grand ideas, of Sweden versus New Zealand. Just stop bungling the basics. It’s not much of a slogan. But it might just be a solution.⁴

Harford’s stop bungling framing might explain those countries in the bottom right corner. Divided responsibilities in federal systems, chaotic messaging and policy instability are likely contributors to poor outcomes in, for example, the United States, the United Kingdom, Spain and Brazil. Getting the basics rights, as Harford says, would improve the situation in such countries.

It seems less obvious why Harford picked out Sweden and New Zealand for the “clash of grand ideas”, when Taiwan is doing so much better, as we can see by zooming in.

Being on the left edge is clearly better than being away from that edge. New Zealand’s better health outcomes relative to Sweden may well outweigh the additional GDP losses incurred here.

But real success – and with it the lowest long-term health costs – requires choosing the least-costly policies that control Covid. The comparison with Australia suggests that New Zealand’s lockdowns may have been more costly than necessary. And, in turn, the comparison with Taiwan and South Korea show that well-resourced and well-run testing and tracing, combined with effective community surveillance, a tightly managed border, and well-managed isolation of contacts of confirmed cases appear to be far more cost effective than the “blunt” tools of regional and national lockdowns.

My take-outs

• Cross-country data does not support a health vs. the economy framing of responses to Covid. Nor does the data support a get health right and the economy will follow.
• Achieving the lowest health costs over both the short and long term requires choosing policies that work in protecting citizens from Covid while minimising economic costs.
• New Zealand should seek to improve its policies, finding those that work at lower economic cost. It is likely that we can learn much from Taiwan, South Korea and other high-performing countries.

Header photo:  The Fantasy section of the Te Anau bookstore: Dave Heatley

Notes
1. The economy is not separate from health. Making the country poorer will have adverse health consequences in the medium-to-long term. A lower current and future GDP restricts the income of a country, with corresponding implications for what its citizens (individually and collectively) can afford to spend on maintaining and enhancing their health. See the discussion in my research note A cost benefit analysis of 5 extra days at COVID-19 alert level 4 for more on trade-offs and complementarities between health outcomes and costs to the economy.
2. The data is imperfect, alas. But it’s way better than anything that might have informed us back in April. Things may, and likely will, change as the pandemic further evolves. But that shouldn’t deter us from digging deeply and asking questions.
3. This graph was inspired by a graph in the Financial Times ($paywalled), published on 18 October.
The vertical axis is based on the International Monetary Fund’s projections of each country’s real GDP growth in 2020, relative to actual growth 2019. The IMF published these projections in October 2020. (The original Financial Times graph used actual GDP growth for January to June 2020. I decided to use the recent IMF data, as these should at least partly reflect the expected effects of the “second wave” now affecting many countries.)
Also note that I’m not saying that the direct and indirect effects of Covid are completely responsible for the differences in each country’s GDP between 2019 and 2020. But, for most countries, they are likely to be dominant.
The horizontal axis is the total confirmed Covid cases per million people as at 6 December 2020, sourced from our world in data. These counts are likely understated for countries with low-quality health systems. (The original Financial Times graph plotted Covid deaths to 14 October 2020. Deaths is a lagging indicator, so I prefer to use cases in this instance. Countries offering higher quality medical treatment, whose outbreaks occurred later in the pandemic, have a younger population, or did a better job of keeping the virus out of aged care facilities might be expected to fare better on the deaths metric, relative to the cases metric.)
For clarity, I have omitted labels for countries that are very small or very poor. Also, the datapoints for a few countries fall outside the limits of the graph. These countries are both small and poor, or their GDP is likely affected by events other than Covid (e.g. Iran, whose economy is affected by international sanctions).
4. Lockdown sceptics v zero-Covid: who’s got it right? Tim Harford, 9 October 2020. Financial Times ($paywalled).