Analysis by Keith Rankin.

This chart shows total deaths in a number of comparable countries with high or highish life expectancies. The countries with most deaths have older populations.

New Zealand should perhaps be compared most with Ireland, Scotland, Denmark and Finland; all countries with just over five million people. And with Australia.

Australian mortality has been similar to New Zealand’s in recent years, though, as more New Zealand citizens migrate to Australia, in the next few years New Zealand will age faster than Australia.

Ireland is exceptional because of its relationship with United States’ high technology companies and its full membership of the European Union. Thus Ireland has many elite ‘tech’ workers at present. Further, in past years of difficulty – especially 2008 to 2014 – Ireland was able to unload much of its underclass to other countries. While the health and financial circumstances of Ireland’s sixty- and seventy-somethings requires further investigation, Ireland will neither have had as big a baby bust as New Zealand in the 1930s nor as intense a baby boom from 1946 to 1965. So, it is likely that the numbers of deaths in New Zealand will rise faster in coming years than the number of deaths in Ireland.

The high death numbers in Scotland, Denmark and Finland do not reflect lower life expectancies in those countries. Rather, they reflect populations with comparatively fewer younger people compared to older people. These countries’ mortality numbers in 2018-2022 are the best guide we have to what death numbers will be like in New Zealand in coming years, as the baby bust generation passes on and the 1940s’ and 1950s’ baby boomers reach the days in which they dominate death data.

Most importantly, the experience of these three countries suggests that we will see the demand for health care in New Zealand surge from now on – peaking in the 2030s and 2040s – at a time when current projections show that New Zealand’s healthcare workforce will trough. There seems to have been minimal, if any, demographic analysis of the implication in New Zealand of a baby bust generation giving way to baby boom generations. This is despite record numbers of policy analysts and cost analysts contracted by government.

Some Particular Comments about other countries

I have included Qatar and its near-neighbour Oman to show how low death numbers are at present in small Arabian countries with relatively large numbers of working-age residents. I think that this observation also applies to Israel.

We note also the death incidence in the higher life expectancy countries of Latin America – Costa Rica, Colombia and Chile – on account of their relatively low numbers of people in their eighties.

We note that Portugal and Japan have relatively high numbers of elderly people in their populations. Portugal has been a retirement magnet within Europe, with strong links to the United Kingdom. I have generally been puzzled as to why Japan has so many older people, though we should note that the generation which fought in World War II has largely passed on. I guess that, as in England, many Japanese children in the war were transported into the countryside so that they were not in the cities which suffered very intense bombing from the United States. Overall, Japan is one of the most age imbalanced countries; the low birth rates in recent decades contribute most to this.

Germany is a country which suffered particularly from Covid19 and similar diseases in 2022. But its high 2018 death tally suggests demographic causes which still need unravelling. Despite Germany being a major labour inflow country in Europe, it still has a median age about ten years higher than New Zealand’s (47 compared to 37).

On the flipside of Germany’s role as a labour-inflow country within the European Union, we have Finland and the other Baltic states as outflow countries. Hence the high death tallies in Finland and the Baltics relative to their resident populations.

Neither Finland nor Denmark look particularly happy in this chart. I predict that New Zealand’s death tally will soon be like Finland’s, given both countries’ propensities to lose labour to bigger neighbours. The situation of Greece is similar to that of the Baltic counties; too great a loss of their younger people to the employment centres of the European Union.

Re a few countries not in the chart, I can affirm that both England and Netherlands have population-adjusted death tallies very similar to the United States. And Canada’s adjusted numbers are very similar to Norway’s.

Finally, we should note Sweden, which was neutral in World War II. So Sweden does not have the extreme demographics of older people which New Zealand and other war participant countries exhibit. And, Sweden was less impacted by Covid19.

______________

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Analysis by Keith Rankin.

These three countries are very useful comparators because they have broadly similar demographics – especially population age structures – to each other. Further they have comparable living standards. The Republic of Ireland has a population the same size as New Zealand (and a similar climate); Australia has close to five times the population of each of the others.

Of the above charts, Ireland best shows the three main waves of mortality in the Covid19 Pandemic. [Note that I will capitalise the word ‘pandemic’ for a WHO-declared pandemic. Otherwise uncapitalised.] There are very clear covid mortality peaks in Ireland in April 2020, January 2021, and December 2022. Other than these peaks there are clear periods of elevated mortality, the second half of 2021 and most of 2022. 2023 also, from March.

Ireland’s population has been growing more slowly this century than Australia’s and New Zealand’s. Death tallies before the Pandemic years were not noticeably growing from 2015 to 2019; compare Australia below. There was an influenza pandemic from late 2016 to about April 2018; the high numbers of deaths in Ireland in January 2017 and December 2017 reflect this. (I have omitted 2016 and 2018 to avoid chart clutter. For Ireland, influenza pandemic deaths actually peaked in January 2018, and extended into March of that year.)

The most obvious difference in the New Zealand chart is the southern hemisphere seasons. The second most obvious difference in New Zealand is the lack of obvious Covid19 waves.

The scales of the two charts are fully comparable, because of the near-identical populations of each country. But a careful look will show that ‘normal’ – ie baseline – summer deaths in Ireland are lower than in New Zealand. Ireland’s population may have better baseline health than New Zealand’s. Or, New Zealand may have more deaths because it has a higher population of post-war ‘baby-boomers’ than Ireland; a population which is now starting to die in greater numbers.

The New Zealand data are worrying in other ways, however. While 2017 clearly shows the 2016-2018 influenza pandemic, with its July 2017 mortality peak, summer data for 2017 and 2019 don’t show large increases in deaths arising from population growth. The period from March to July 2019, in the absence of known epidemic illness, nevertheless looks like a protracted period of deaths triggered by early seasonal viruses. (Indeed, I recall from my former workplace that there were a lot of ‘bugs’ around for parts of 2019.)

If we regard the April to June periods in 2017 and 2019 as having elevated death tallies, then 2021 looks like a normal year in New Zealand, even allowing for population growth. Yet it wasn’t a normal year. It was the peak year of the Covid19 panic; the year of the most extreme public health mandates, with an effectively shut international border and with face-masking required in many settings. The big question is to ask why 2021 was not more like 2020. In the winter of 2021, New Zealand had no Covid19 to speak of, and no influenza.

Elevated death tallies reappeared in March 2022, continuing through to January 2023. While these were clearly linked to Covid19, there was no mortality peak anything like that which Ireland experienced in December that year. My guess is that the timing of mortality in New Zealand reflected the timing of booster vaccinations against Covid19, whereas Ireland was caught unawares that December.

A good check on these tentative conclusions for New Zealand is Australia. (The scale is five times higher, reflecting that Australia’s population is five times greater.) Australia shows most of the same features as New Zealand in the years before 2020, though in a muted way. Australia shows more consistently than New Zealand the impact of population growth before 2020 being reflected in more deaths each year than the previous year. We see that in the spring months (September to November) Australian deaths are generally lower than New Zealand’s; probably because winter lingers for longer in New Zealand.

Australia shows the same problem in 2021 as New Zealand; normal winter deaths despite highly abnormal circumstances. As in New Zealand, there almost certainly were ‘killer viruses’ in both countries that year. Deaths to some extent will have been people who would otherwise have died in 2020, but avoided viruses then because of the lockdowns and physical distancing. Also, weakened immunity arising from the lack of normal exposure to respiratory viruses in 2020 will have increased the chances of vulnerable people dying in 2021 after contracting such a cold virus. The 2021 mortality peaks were higher in New Zealand than in Australia.

Australia shows a classic Covid19 mortality peak in January 2022, before the ‘Omicron’ variant of Covid19 was discovered in New Zealand. Australia had covid exposure peaks in December 2021, much of that being the Delta variant, pre-Omicron. In the autumn and early winter of 2022, Australian mortality data show a shorter and lower ‘Omicron wave’ than New Zealand data.

All three countries continue to show elevated levels of mortality this year, though this is obscured in New Zealand by the problematic numbers of deaths in the autumns and early winters of 2017 and 2019. New Zealanders in the first four months of this year have had slightly more deaths (adjusting for population) than Australians. This may reflect New Zealand’s relatively more overstretched healthcare system, noting from having myself spent some time in Australia this year that Australians also see their healthcare system as overstretched.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The Pandemic can be assessed a bit like games of football, with deaths being the score. (Or, given that there are many ‘teams’ competing together, a better analogy may be a Marathon race. Nevertheless, I will use the language of the football metaphor.)

The winning country would be that with the fewest number of deaths attributable to Covid19. While this football metaphor is indeed useful, our perceptions of ‘who did best’ are strongly coloured by the pandemic’s first year, when media attention was greatest, when public health measures were most ‘in our faces’, and when the pandemic response was at its most bureaucratic.

As a result, the half-time scores are the scores that most seeped into public consciousness. Then, deaths which were classed as ‘covid-deaths’ were implicitly seen as more tragic, more requiring of daily tallying, than other deaths.

The two Tables below look at this Pandemic ‘World Cup’ through the simple demographic criteria of increases in deaths, all deaths. We may note four ‘ordinary-time’ phases of the pandemic. Together, they add up to a period of four years.

First, the warm-up, from May 2019 to February 2020. The warm-up, going back to 2019, is important to include because countries with unusually low numbers of deaths due to respiratory illnesses in 2019 would typically have higher death tallies in the next respiratory epidemic, whatever virus that might be.

Second is the first half of the Pandemic proper, which I would date as March 2020 to March 2021. Third is the second half, from April 2021 to April 2022, which includes the waves associated with the Greek-alphabet variants (especially Alpha, Delta and Omicron).

Following ‘ordinary-time’, there was ‘extra-time’ which I define as May 2022 to April 2023. We may note that the WHO declared the Pandemic to be over at around the end of April this year. So, we may formally categorise the period from May 2023 as ‘post-pandemic’.

Table 1 below indicates the score at the end of ordinary-time. It shows the percentage increase in deaths for a number of countries for the three years from May 2019 to April 2022, compared to the three years from May 2015 to April 2018. In the right-hand column is a counterfactual which is a best estimate of what the increase in deaths would have been in the pandemic period had there been no pandemic. (The counterfactual is calculated by comparing deaths in the 24 months ending April 2019 with deaths in the 24 months ending April 2017. I have used April years because, in both hemispheres, the period in late April and early May is generally free from epidemic respiratory deaths. This method minimises the impact to this calculation of the severe influenza global epidemic which lasted from late 2016 to early 2018.)

Table 2 indicates the ‘extra-time score’, comparing the year-to-April 2023 with the year-to-April 2019. It uses the same ‘trend’ counterfactual as Table 1. Whereas Table 1 is sorted to place the ordinary-time ‘winners’ at the top, Table 2 is sorted to place the ‘extra-time losers’ at the top. (We note that, for Table 2, some countries are laggards in publishing their mortality data; and also that the most recently published numbers are subject to upwards revision. The countries which are problematic in this regard have their data marked with asterisks, the number of asterisks indicating the degree of estimation required.)

In the pandemic proper, the two countries with easily the least increases in deaths were Norway and Sweden. The others in the ‘Top Eight’ (the ‘quarterfinalists’, to use the football metaphor) were the other Nordic countries, Australia and New Zealand, and Taiwan.

Based on the media coverage in New Zealand and the world news channels that New Zealanders mainly follow, the only surprise in that Top Eight would be Sweden, which pursued a very different policy response, especially in the ‘first-half’ of the Pandemic. In the 2020 New Zealand election campaign, political parties generally agreed that Taiwan was the exemplar for other countries to follow.

We may note that only New Zealand and Taiwan had counterfactuals showing higher projected increases in deaths than what actually happened. Thus, these two may be declared the ‘ordinary-time’ winners. The problem is that the Pandemic World Cup had ‘extra-time’. (It must also be noted, however, when we take the ‘non-death costs’ of the pandemic and its associated health policies, Sweden’s non-death costs were easily the lowest. So, on this basis, it could be argued that Sweden was the true ordinary-time winner, despite having been way behind at ‘half-time’.)

Extra-Time

When we look at Table 2 we see clearly that the East Asian countries performed very poorly. Most of these were deemed to be success stories in ordinary time. Taiwan is very prominent here. So is South Korea which has a conservative estimate in this table for its ‘extra-time’ increase in deaths. Macao is very important here, because it is the best proxy we have for China. Taiwan has had a recent resurgence in deaths in May 2023, and Macao has had a resurgence of Covid19 cases in recent weeks. So, these countries’ pandemic problems are far from over. (There are also signs that New Zealand’s seasonal death tally is picking up early this year.) The Macao situation, combined with other reports that all is not well in China right now, suggest that China may be presently going through a significant third wave of Covid19. This will add to global supply-chain problems.

New Zealand and Australia are in the top (ie worst) half of Table 2. So are two of the Nordic countries, Iceland and Finland, the Nordic countries which imposed more restrictive health mandates than their neighbours. So is Ireland near the worst, more restrictive in its public health mandates than the United Kingdom countries. Norway, top of Table 1, is in the middle of the Table 2 pack. Of the Nordic countries, only Sweden – easily the least restrictive in Europe, especially in the first-half of the Pandemic – performed well.

Quirky Counterfactuals

Creating consistent counterfactuals for each country is difficult because there are quirky demographics at play. First, we note that there are three main reasons why death increases might trend high for a given country. The first is a general increase in the population of a country: more people, more deaths. Second is the aging of a country, represented by increases in the median age of living persons. Third is a deterioration of general health, especially of those middle-age cohorts whose deaths ‘come under the radar’, given that deaths are dominated in most countries by people aged over 75.

It is likely that the high counterfactual for New Zealand is due to a mix of these. We know that New Zealand has some of the same issues of underclass deprivation as the United States, which include obesity, diabetes, and substance abuse. And we know that the United States has a lower life expectancy than other ‘western’ countries; a life expectancy now known to be falling.

The other two main quirks to look out for are birth rates in the troubled second quarter of the twentieth century. The Great Depression and World War Two were the main events that impacted on birth rates. There was also warfare in the 1950s in Korea and Malaysia. Sweden is an interesting case, comparable with Switzerland, neutral in World War Two, so having a lesser demographic impact from the War. Also, Sweden came out of the Great Depression early, meaning it will have had comparatively high birth numbers in the 1930s; Sweden’s peak deaths since 2015 will have been higher than otherwise, on that account.

While New Zealand is possibly the western country with the fastest population growth this century, this is offset by the fact that low birth numbers in the 1930s are translating to lower deaths since 2015. (See my recent ‘Smithometer’ analysis, in Granny Smith.) Aging and population growth are not the whole story of New Zealand’s upper quartile trend of increasing deaths. (Unlike, say Portugal, which is known to attract retirees in Europe as Florida does in the United States.)

New Zealand also has the additional factor of having, in June and July 2022, too many vulnerable people who were denied, for unexplained political reasons, a timely second booster Covid19 vaccination. The July 2002 mortality peak, almost entirely experienced by older European-ethnic New Zealanders – the Granny Smiths – came to a prompt end once these people became eligible for second-boosters. This sharp July peak – and drop-off – appears to have been a New Zealand specific phenomenon.

High counterfactual notwithstanding, New Zealand performed very poorly in extra-time, though not as badly as the East Asian countries which imposed the most ‘sterile’ public health policies on their people.

East Europe

I would like to note two other groups of countries. First, it was Eastern Europe which had the highest reported per-capita Covid19 death tolls. These countries do not look as bad in this analysis, though they (except the Baltic states) still look bad in Table 1, especially in light of their often negative counterfactual death trends. The main demographic problem that these countries have been facing is emigration of working-age adults, especially those Eastern European countries in the European Union. Generally, these countries look much better in Table 2, in extra-time.

Most of these countries successfully imposed severe public health restrictions in the first half of 2020, but abandoned those restrictions – or were unable to easily reimpose them – in the later stages of ‘the game’. The result was that these countries’ populations had substantially compromised immunity going into the winter of 2020/21. Their death peaks were much higher than the death peaks in the west earlier in 2020. The second problem was that, on account of their departed youth, their populations were aging as well as falling. Hence the high Covid19 per capita death tolls that savage winter.

In ‘extra time’, East Europe has ‘performed’ best. This would appear to be in part because so many of their most vulnerable people had already died; respiratory viruses had lost much of their human ‘fuel’. Also, these countries had re-established (the hard way) high levels of natural immunity to respiratory illnesses.

Russia continues to supply mortality data, though it excludes deaths in the Ukraine conflict zone; so its not included in the tables. And Ukraine has certainly stopped supplying data, due to governmental priorities as well as a lack of will to publicise its present demographic plight. Kazakhstan is probably the best proxy for assessing the impact of the Pandemic in Russia.

South America

These countries (plus Mexico) are among the worst performers in both Tables. Typically, they exhibit many of the ‘underclass’ socio-economic problems apparent in the United States, United Kingdom, and New Zealand: inequality, poverty, homelessness, obesity, crime, violence. It is likely that they will see ongoing increases in annual mortality on account of these factors; factors exacerbated by both the Pandemic and its associated mandates.

While Latin American populations are much younger than Eastern European populations – due to both higher births and less emigration – there will also have been a significant growth of numbers of people of peak-dying-age (over 75) contributing to ‘trend’ counterfactuals in some cases as high as New Zealand’s.

Another factor in these American countries is the high proportions of people living in or near the tropics at high altitudes. Under normal circumstances, these are unusually healthy environments, in which seasonal respiratory illnesses do not circulate as much as in temperate climes. But, it makes people living in these zones more vulnerable to pandemic respiratory illnesses when they do happen. It’s an old story that goes back to the time of Spanish colonisation in the sixteenth century.

This last factor is only apparent here in Colombia and Ecuador. Other similarly affected countries – Peru and Bolivia – had very high early death tolls, but (presumably due to political crises) have not released ‘extra-time’ data. Venezuela was even less forthcoming with useful data.

Africa including Qatar

Mortality statistics from Africa are rare. Egypt is now the best, and it certainly suffered. South Africa, which in ‘ordinary-time’ had a similar experience to that of East Europe, used to supply good quality data; but no more as its present economic crisis deepens. Signs are that the African continent was less impacted directly by Covid19 than other regions, though its more fragile economic supply-chains have become victims of the Pandemic’s ‘extra-time’ environment.

Qatar is an interesting case, because of its unusual demographics. Qatar’s resident population is heavily weighted towards the younger working-age population. So, while its death rates per capita have been very low, its percentage increases in deaths have not been low. We do need a good comparative analysis of the health impact of Covid19 on working-age populations, though made difficult by demographic data today still focusing on sex and ethnicity rather than age or occupation or labour force status.

South Asia

After China, the biggest Pandemic uncertainties relate to South Asia, with Inda being the largest country. We may also add the very populous country that is Indonesia. This region is a demographic black hole, which experiences high levels of emigration as well as of death. (We may note here – see Hundreds of Pakistanis dead in Mediterranean migrant boat disaster, CNN 19 June 2023 – that the majority of victims of the overcrowded refugee boat which sank last week off the coast of Greece were from Pakistan.) This region has suffered a huge upheaval since 2020, with the Pandemic a significant contributing factor.

Conclusion

There remains a lack of competent demographic analysis of recent and former pandemics, partly due to poor (sometimes politically-motivated) record-keeping, and partly due to the low status of demography among the social sciences. Analyses like mine here – amateur in the sense of being unpaid, but not in the sense of quality – help to fill the gap.

The most striking conclusion is that the ‘extra-time’ of the Pandemic gives a very different picture of the Pandemic’s human cost. The imposition by governments of sterile environments for long periods is not a recipe for good health outcomes, although it may give good headlines in the early phases of a pandemic when the Press is at its most attentive.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Over Easter I relistened to Jim Mora’s RNZ interview (17 May 2020) of Johan Giesecke, “world leading epidemiologist” and Professor Emeritus at the Karolinska Institute in Stockholm, Sweden. In the period April to June 2020, Sweden gained notoriety for its divergent public health policies with respect to the management of the Covid19 pandemic. People – including me – widely pointed to Swedish authorities then as being more concerned about retaining a pretence of their economic normality rather than caring about people’s lives.

Swedish exceptionalism became a thing, again; this time seemingly for all the wrong reasons. Hitherto ‘progressive’ New Zealanders had regarded Sweden as an exceptional policy exemplar; now it seemed to be an outlier of classical liberalism.

Here are two summary measures of pandemic and post-pandemic mortality; comparing Sweden with Finland, Germany, New Zealand and Japan:

We note that all these countries have rising populations of older people, so some increase in deaths was to be expected in all of them. Sweden had covid vaccination rates comparable with these other four representatives of ‘the rest of the civilised world’, so differences in vaccination uptake cannot explain its mortality difference.

It’s worth relistening to this Giesecke interview, now with the perspective of hindsight. The context is that, in the contest (as it was then framed) of Sweden versus the rest of the civilised world (with the World Health Organisation settling on the counter-Swedish majority view), Sweden has come out a clear winner. The scandal is the failure of ‘the rest of the civilised world’ to acknowledge the statistical reality.

(Note that I use ‘civilised world’ with mock irony. In New Zealand at least, few politicians or high-profile commentators believed that there could be anything New Zealand authorities could learn from the experiences of West Europe, South America, or Africa; instead, the policy elite contemptuously assumed such countries to be ‘basket cases’. See the use of this phrase in 1980s: Days of greed and glamour, NZ Herald, while noting that we are still waiting for a balanced history of the ‘Muldoon Years’ referred to.)

Highlights from the 2020 Interview, and the interview itself can be heard here, and read in synopsis form: Johan Giesecke: Why lockdowns are the wrong approach, Radio New Zealand, 17 May 2020.

Relating to points not covered in RNZ’s synopsis, Giesecke draws a direct comparison with Finland, which was pursuing a public health policy very close to New Zealand’s. His concern – shared by Finland’s state epidemiologist – was that the authorities’ actions were creating a significantly vulnerable population in Finland.

Giesecke, from that May 2020 perspective, mentions that if a good vaccine would come quite soon then New Zealand’s outcome might be better than Sweden’s. The irony is that, while a good vaccine did indeed come quickly, New Zealand’s authorities were slow to embrace the vaccine as the answer; having already decided that New Zealand had eliminated the virus as per the China policy. Then, after New Zealand’s people were vaccinated, the government doubled down on the lockdowns, not at all trusting the vaccine to work as a substitute for lockdowns. (Indeed, New Zealand only abandoned its border-quarantine policy in 2022 because that policy failed on its own terms. Had the border policy been implemented without error, New Zealand presumably would have followed a set of draconian restrictions through 2022, with a timeline similar to that of China. New Zealand’s border mishaps proved to be a blessing in disguise.)

In mid-2020, Johan Giesecke’s main expectation was that the mortality experience of all countries in the OECD (essentially the rich western plus the rich eastern countries) would all be about the same; and that Sweden’s major benefit would be in its substantially lesser disruption to normal life.

Where Giesecke was wrong was that the OECD ‘WHO countries’ (a label for the ‘civilised world other than Sweden’) ended up with substantially higher increases in deaths than did Sweden; he was wrong in a way that favoured Sweden rather more than he had expected.

The pandemic has nevertheless had an adverse impact on Sweden’s mortality. Sweden did experience the West European surge in deaths from respiratory illnesses late last year. Its people no more live in a bubble than do New Zealand’s. Overall though, Sweden got the win-win outcome: fewer deaths, and less social and economic dislocation. (David prevailed over Goliath.)

A very basic summary of the difference between the Swedish and the Goliath approaches is that Sweden focussed on its people whereas the prevalent strategy focused too much on the virus; the world by-and-large pursued a strategy of covid exceptionalism. (One consequence of covid exceptionalism was that a death clinically ascribed to Covid19 became a more noteworthy death than most other deaths.) Sweden focussed on having people with good levels of immunity, whereas in 2021 much of the rest of the world went down the unhelpful path of obsessing over the various mutant variants of the novel coronavirus.

Giesecke clearly had a better understanding the history of human coronaviruses than most other epidemiologists; these are viruses for which specific immunity is short-lived, from which we top-up immunity naturally through living our daily lives in a normal manner, and for which vaccine-conferred immunity would also be short-lived.

Sweden understood the science better; indeed, the interview tells us that there was a substantial scientific contest of interpretations of the evidence in Sweden, a good sign that actual science was taking place. Not only did a number of Sweden’s scientists prove to be among the better predictors of the future, people such as Johan Giesecke were also much more prepared to offer humility to their own people and to the world if they had got it wrong.

I am still waiting for our authorities – including the scientists – to do a proper retrospective comparison of New Zealand (and other countries, as in my table above) and Sweden. I am still waiting for a little gracious humility from our authorities in Aotearoa New Zealand. Humility is an important characteristic of civilised behaviour.

In the middle of the interview, Jim Mora noted: “Our readers are quite fascinated with Sweden; I think the world is”. When and why did that fascination stop? Or is it just that Goliath’s information mediators stopped being fascinated when the ‘contest’ moved in David’s favour?

————-

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Analysis by Keith Rankin.

A pandemic can end in three ways. Either the death rates attributed to the pandemic disease cease, or at least drop back to pre-pandemic levels. Or normality is re-established, with the pandemic disease still present, but displacing other causes of death. Or a ‘new normal’ is established, with higher ongoing rates of death normalised.

So, to some extent, a pandemic’s duration is a state-of-mind; meaning that the post-pandemic period is when that ’emergency’ mindset has departed. To a large extent, that happens when the most burdensome public health restrictions become untenable; in New Zealand’s case, that was when the substantive closure of the international border finished. Deaths, covid or otherwise, may still be a problem, but they cease to be newsworthy!

For most of the world, the post-pandemic period started around February 2022. East Asia was the principal exception. Table 1 below shows mortality in the first year of the new normal.

Table 1: Back to Normal?
total deaths:	2018/19*	2022/23**	increase
Macau	2199	3596	63.53%
Hong Kong	47056	62056	31.88%
Singapore	21323	26829	25.82%
Taiwan	170483	212665	24.74%
Thailand	484272	603662	24.65%
Iceland	2180	2702	23.94%
South Korea	291529	357298	22.56%
Chile	107408	130970	21.94%
Colombia	236488	285944	20.91%
Ireland	30051	35650	18.63%
Brazil	1325677	1569617	18.40%
Japan	1360950	1607011	18.08%
Northern Ireland	14998	17504	16.71%
Australia	161466	188155	16.53%
Scotland	55633	64807	16.49%
Malaysia	171015	199069	16.40%
Finland	53458	62112	16.19%
New Zealand	33310	38682	16.13%
Netherlands	148356	171826	15.82%
Germany	925309	1069924	15.63%
England & Wales	515610	592677	14.95%
Norway	39819	45650	14.64%
Austria	80544	92325	14.63%
Canada	279510	319140	14.18%
Uruguay	34655	39512	14.02%
United States	2812658	3193088	13.53%
Greece	122940	139406	13.39%
Mexico	726738	819268	12.73%
Portugal	111815	124757	11.57%
Spain	415025	458846	10.56%
Switzerland	66396	73311	10.41%
Italy	641280	705564	10.02%
Israel	45488	49996	9.91%
Denmark	53578	58826	9.80%
Peru	157650	172825	9.63%
France	591364	647762	9.54%
Czechia	110671	120448	8.83%
Slovenia	20603	22307	8.27%
Belgium	107810	116284	7.86%
Slovakia	54017	58196	7.74%
Poland	405241	435401	7.44%
Sweden	88633	94436	6.55%
South Africa	527630	561992	6.51%
Egypt	570015	605500	6.23%
Croatia	52144	55093	5.66%
Albania	21717	22900	5.45%
Hungary	131229	135090	2.94%
Kazakhstan	131089	134709	2.76%
Bulgaria	109175	112080	2.66%
Serbia	101699	104390	2.65%
Romania	263338	270222	2.61%
Moldova	37314	36554	-2.04%
*	year ended April 2019
**	latest available 12-month period

Table 1 shows a number of countries’ most recent annual death tallies compared with the year ended April 2019, the best baseline period available. May 2018 to April 2019 was chosen because it represents the first full year after the silent influenza pandemic of November 2016 to April 2018. While not a media event, that largely invisible 2017 pandemic was a substantial mortality event, at least in the ‘global north’. A pandemic does not require an authentication from WHO to be an actual pandemic. A pandemic is simply a globally widespread experience of a disruptive contagious disease.

Broadly, Table 1 shows the countries which followed ‘elimination strategies’ near the top for post-pandemic mortality. Not only did countries in the east of the eastern hemisphere (including Aotearoa New Zealand) pursue the most stringent anti-covid policies (and practiced them for the longest time periods), some  prematurely claimed to have eliminated (though not eradicated) the disease. For some in East Asia, the 2003 experience of SARS was uppermost in health officials’ minds.

Table 1 also shows that some of the countries worst-hit by the pandemic (especially those in the Southeast European ‘Balkans’) have returned to death tallies comparable with base-year numbers. If South Africa and Egypt are a suitable guide, that return to health normality applies to Africa as well.

The only West European countries with post-pandemic deaths under nine percent more than pre-pandemic deaths are Sweden and Belgium, both countries with high covid death tallies in the first wave of the pandemic, but well below European mortality averages in the second year of the pandemic. Sweden’s 6.55% increase actually overstates its situation by about two percentage points, because, more than in most other countries, deaths there were particularly and unusually low from May 2018 until the start of the pandemic. Also, Australia’s 16.53% is an overstatement, probably by at least two percentage points; this is because tardy Australia’s most recent annual deaths’ data includes the months of December 2021 and January 2022, both high mortality months compared to the following December and January.

New Zealand’s most recent death data uses December 2022 and January 2023, not December 2021 and January 2022. In contrast to Australia, New Zealand’s Table 1 increased mortality experience is understated by a percentage point, because March and April 2019 (included in the pre-pandemic baseline year) had somewhat higher deaths than those same months in 2018. If we had used a baseline year from March 2018 to February 2019, then New Zealand would have had a mortality increase of 17.40%, not 16.13%.

Re East Asia, the numbers for Macau and Hong Kong give a hint of the recent reality in China. For that region we should note also that the South Korea increase in Table 1 (22.56%) is a substantial understatement of reality, because South Korea has not reported ‘total deaths’ after July 2022, and we know that Korea has had many covid cases since then.

We also note that post-pandemic death tallies are high for Japan, Ireland, Scotland, Iceland, and Finland. These are all countries which, for their regions, were known for their more restrictive public health policies. Finland was widely acclaimed for being the most restrictive of the Nordic countries during the pandemic years. (We also note that Finland had many more deaths than Norway both pre-pandemic and post-pandemic, despite having about the same population as Norway; it suggests that many more young Finns are working abroad than young Norwegians. Likewise, we see that New Zealand has more deaths than Ireland, despite both countries having essentially the same population.)

Germany, which has had a particularly worrying recent run of deaths, in Table 1 is not out of step with its western neighbours; although we should note that southern Western Europe has generally had a post-pandemic more normal than northern Western Europe (Sweden excepted).

The critical question, looking to 2023 and 2024, is whether, for the countries towards the top of Table 1, the pandemic has triggered a new normal with persistently higher mortality than in the 2010s’ decade. Or have these countries simply experienced a delayed pandemic mortality experience which will soon subside? If the latter, then we should expect a substantial mortality drop in East Asia and West Europe in the year to April 2024.

Demography and the challenges of predicting the pandemic’s influence on 2020s’ mortality

Demography is a complex subject. Pandemic death rates per capita were high in Eastern Europe because those countries have lost many of their young people to emigration. Increases in death tallies, however, were never so high in those countries. The demography of Europe is particularly complex because many of their older people were born either side of, or during, World War Two; a war with substantial demographic consequences which have not yet fully played out.

The Scandinavian countries in particular had diverse experiences in that war. Sweden was neutral, Norway and Denmark were occupied, while Finland was successively friend and foe to the allied powers. So the change in the number of older people may differ in Sweden compared to the others. Nevertheless, Sweden still compares well with the other neutral countries: Switzerland, Ireland, Spain and Portugal. (Though noting that Spain had its own especially large demographic trauma in 1936 to 1939.)

Another problem in unravelling the demographics of Europe is the substantial international migration between present and former European Union countries, and immigration from former (or present) empire countries. So many people these days die in different countries from which they were born. We know little about the different pandemic and post-pandemic death experiences of immigrants compared to people born in the country of their death.

In most countries deaths in 2022/23 would have been higher compared to 2018/19. The main determinant of death rates is the numbers of people in the oldest age cohorts. About half of all deaths in most countries are of people in their eighties. So the biggest increases, for reasons other than the pandemic, would be due to the rate of increase or decrease of a country’s population of octogenarians. Some countries will have significantly fewer octogenarians after the pandemic, because the pandemic itself took so many.

The second most important reason for changing death tallies is the underlying health of the people. Pandemics take more people in countries which already have substantial populations – especially populations in the 65 to 74 age group – with compromised pre-pandemic health or compromised general immunity. In pandemic years, the main reason for more death is worse underlying health. In other years changes in health status may either accentuate or offset changes in the numbers of people over eighty. While there are health-compromised people of all ages, compromised health – high morbidity or low general immunity – is more likely to have prematurely fatal consequences for people aged 65 to 74.

To summarise the two previous paragraphs, I would argue that the two main predictors of a country’s normal death tally are the numbers of octogenarians in the population, and the numbers of people in the population aged 65 to 74 with compromised health or general immunity. (In addition, some developing countries still have unacceptably high levels of infant mortality.)

The two key aspects of the health status of living populations are morbidity and immunity. The countries which fare best in a novel virus pandemic (or from wave pandemics of pathogens which induce only-short-lived specific immunity) are those with low morbidity and high general immunity. With respect to the present post-pandemic period, the covid coronavirus increased both the morbidity and the immunity of populations. Where these two increases balance out, then a new normal appears which looks substantially like the old normal.

Before the twentieth century, people living rurally were more likely to experience longevity. That changed in the twentieth century, when people living in metropolises gained super-high-immunity levels from living in close proximity to each other (improving immunity); and urbanised populations experienced reduced morbidity as a result of access to a wider range of foods, from more timely access to healthcare services, less exposure to conditions such as malaria, and safer supplies of drinking water.

Big cities still reduce life outcomes for people immigrating from rural areas; for people not yet adapted to city levels of exposure to pathogens, and often having to settle for inferior housing and employment experiences. When governments tamper with the finely-tuned immunity equilibria in our big cities, the potential for deadly unintended consequences has always been there. Such tampering may include the required overuse of facemasks, and the creation of fear around the use of public transport.

The post-pandemic experience of East Asia is not a particularly good advertisement for disruptive public health practices. Sweden was conspicuous by taking the opposite policy tack from that taken in East Asia, minimising disruptions from normal social interaction. Sweden’s different approach was not a result of its greater wisdom or greater laisse-faire liberality; rather it was a result of a mistaken assumption that, by mid-March 2020, many more people had already been infected by the new coronavirus (making it too late for restrictive policies) than actually had been infected.

The interregnum between the two recent respiratory pandemics

Finally, it is worthwhile to suggest reasons why deaths in much of the developed world were especially low from May 2018 to February 2020; a phenomenon particularly marked in Sweden. This was most likely because of the 2017 influenza pandemic – the invisible pandemic (invisible even to demographers, then more attentive to issues other than heightened seasonal mortality). This world disease event left populations more immune, and (because that pandemic took so many) it meant that the post-influenza 2018 population was more healthy and had more immunity than the pre-pandemic 2016 population.

It is normal for post-pandemic death rates to be low for a couple of years. Indeed, it was true around 1919 and 1920, after the great influenza pandemic of 1918. Will it prove to be so this time, from 2023 to say 2025? We should be watching aggregate mortality – in our own countries and other countries – with as great interest as we watch the inflation, unemployment and economic growth data.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Analysis by Keith Rankin.

The ‘Young Elderly’ are in essence the post-war baby-boomers. An average young elderly person in these charts was born around 1950 to 1952.

The charts look at ‘quarterly excess deaths’, so do not show week-by-week fluctuations in deaths. For example, data for the very end of 2022 covers the whole of the last three months of 2022.

As in my previous recent charts (see my Spiralling Deaths in Germany, Evening Report, 14 March 2023; and Examples of Germany and Denmark, Evening Report, 12 March 2023), I have emphasised Germany, because late-pandemic mortality has been so bad there. And because Germany’s differences with the rest of Europe create a very useful point for epidemiological analysis.

In the first chart (above), of the countries shown only Germany and New Zealand had excess deaths in this age group below ten percent in the first six months of Covid19. The United Kingdom was easily worst then.

The United States had a really bad pandemic, for two years from April 2020 to April 2022. But, subsequently, for nearly two years since April 2021 Germany has been for the most part easily the most deathly of these countries, for the young elderly, with only the USA contesting Germany for this dubious honour. For some of 2022, New Zealand was in second place out of these seven countries.

We note that Belgium, Netherlands and France all had high death rates early in the pandemic, but have subsequently had much lower death rates than Germany for this age group.

The two obvious avenues for investigation are diet (the French surely have a more healthy diet?) and differences in the policy responses to the Covid19 pandemic. My understanding is that, while all countries had similar public health restrictions during the peak weeks of Covid19, Germany was much the slowest of these countries to remove mandated public health measures. Germany’s abundance of caution may have backfired big-time. Yet the only reason given here (on Deutcshe Welle) is: “that diseases other than Covid19 are bouncing back because fewer people are wearing masks amid a general relaxation of pandemic rules in comparison with the past two years”. There is no hint of comparative analysis in this particular DW media report.

The second chart compares Germany with its four nearest Eastern European countries, and with the two Mediterranean countries which were the first to experience substantially elevated pandemic death rates.

For the young elderly ‘boomers’, Spain and Italy show quite the opposite pattern to Germany; they started with high death rates and then moved to generally lower rates. Both had summer mortality peaks in 2022, to some extent due to the summer heat waves but mainly due to the rebounding of tourism with Covid19 still present. Covid19 flourishes in bars and restaurants, and in airport terminals.

The eastern countries had more deaths overall in the middle seasons of the pandemic, despite (or maybe because of) the success of the measures taken in the early months of the pandemic. Also, for these countries, with lower life expectancies than their western neighbours, the young elderly are on average closer to their eventual deathdays. It is important to note that these eastern countries had the fewest pandemic-related deaths after March 2022. Presumably their people most at risk of dying had already died, and the remainder had higher natural immunity to respiratory illnesses than did the older citizenry of Germany. (I am not aware that Polish, Bohemian or Hungarian cuisine is particularly noted for its health benefits, in contrast to the much-touted Mediterranean diets; so a better diet is probably not the reason.)

There is one other country with a similar pandemic death-profile to Germany; its southern geographic and cultural neighbour, Austria. The final chart here shows the smaller countries of western and central Europe, plus New Zealand. (Australia and Sweden do not provide age-group data.)

First we note that Austria and its neighbours Slovenia and Switzerland start out closely synchronised. Switzerland drops off Austria’s high young-elderly mortality path from March 2021, and Slovenia drops off a year later (though has a high summer peak in line with its Italian neighbour). The Scandinavian countries had generally low death rates for the young-elderly age group. (They did however see rising deaths from mid-2021 for the older elderly.)

Any valid epidemiological analysis for Germany’s 2022 tragedy needs to take into account the similar experience of Austria, as well as the generally different experiences of the other European and neo-European countries.

The United States still has a worse overall pandemic record than Germany, for the young elderly. The worry for Germany, though, is that the reasons for its really bad 2022 have not necessarily been resolved; 2023 may be just as bad. Time will tell; so long as an asteroid strike or a nuclear war don’t displace infectious diseases as drivers of excess mortality in Europe.

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

https://www.dw.com/en/winter-illnesses-burden-germanys-intensive-care-units/a-64135331

https://www.dw.com/en/top-german-virologist-says-covid-19-pandemic-is-over/a-64214994

Germany’s Deadliest Weeks since World War Two? – Spiralling Deaths in Germany

The four weeks ending 8 January 2023 have seen easily the most deaths in Germany of any four‑week period since 2015. The worst week was the week ending Christmas Day, with 28,481 deaths. While it’s hard to compare with pre-1990 years, due to the East Germany question, it may well be that this week last December had a greater percentage of excess deaths than any other week since the last world war.

Baseline weekly deaths for 2022 would have been just over 17,000; a baseline of 68,700 for four weeks, as shown in Table 1.  Therefore, winter illnesses have raised peak deaths at the end of 2022 to 65 percent above what they would have been in a normal non-winter week.

While the December 2022 statistic is remarkable, it seems that most Germans themselves are not aware of this. Many people suffering individual tragedies will typically not be aware if their ‘micro’ tragedy is part of a much bigger ‘macro’ tragedy. This DW story (23 Jan 2023) The impossible task of calculating global pandemic deaths, only looks at 2020 and 2021, and gives no commentary on the Germany chart included. The best I can find on DW discussing the health situation in Germany last December is: Winter illnesses burden Germany’s intensive care units, 17 Dec 2022.

What is remarkable is that this latest winter toll comes very soon after three other periods of high peak mortality in 2022, listed in Table 1 as ‘autumn wave’, ‘summer wave’, and ‘omicron wave’. So, from the Grim Reaper’s point of view, the ‘low-hanging-fruit’ should already have passed.

These recent mortality waves compare unfavourably with the three ‘classic’ Covid19 death waves, each of which had weekly peaks in 2020. By ‘classic’ I mean the original ‘Wuhan’ coronavirus strain, before ‘variants’ and ‘vaccinations’ became a thing in 2021.

It is also noteworthy how high some of the pre-covid death peaks were. The influenza ‘pandemic’ of late 2016 to early 2018 was particularly pronounced. (I use single-quote-marks, because this actual pandemic was never granted pandemic-status by the World Health Organisation.) Germany’s two peaks for this influenza pandemic were in February 2017 and March 2018. We also note a particularly bad season of epidemic influenza in early 2015.

Refer to my Death Spikes and Covid Dissonance? Examples of Germany and Denmark for charts recently published, comparing Germany’s excess deaths with those of its neighbour, Denmark. The December 2022 mortality peak is reproduced to some extent in most (but not some eastern) European Union countries, and in the United Kingdom and United States. However, Germany’s year-of-death in 2022 is probably the most dramatic. (One other country which appears to have an equally problematic mortality, maybe worse, in 2022 is South Korea. I wait in hope for the eventual publication of South Korea’s complete dataset.)

Chart 1 below shows ‘excess deaths’ – as distinct from total deaths – for Germany, by age group.

Chart 1

Germany’s demographics are unusual (but maybe not unusually unusual) on account of World War Two. The oldest Germans – shown in red – were all born before that war. The German post-war baby-boomers are shown in green. Germany shows disturbingly high rates of pandemic death for its baby-boomers, from 2021. (It should be noted that Covid19 deaths tended to peak from November to January, whereas epidemic influenza death tended to peak in February or March. Thus the big reductions in excess deaths each February and March are mainly due to high death-norms set by pre-covid influenzas.

Countries regarded as having pursued the best anti-Covid19 public health policies in 2020 have not had a good 2022. Germany is one of those countries. South Korea, Japan, Taiwan, Thailand, Singapore and China are others. So are Australia and New Zealand. Once having acknowledged the 2022 death statistics for what they are, terrible, the question is whether the problems of 2022 in these countries will extend into 2023. While my hunch is that new vaccinations could make a difference, in 2023 at least, I am concerned that societies have already passed a demographic turning point and that life expectancies are already declining from their peaks, and may continue to decline for decades.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The chart above is quite alarming, and not only because of Germany’s high death toll. It’s the difference between reported Covid19 deaths and excess deaths, in the context where four waves of excess deaths in Germany in 2022 are clearly ‘epidemic’ in nature. And we see the same death waves in Germany’s northern neighbour, Denmark.

Both countries are assiduous in releasing their demographic data on deaths; most countries are not. And both countries had a reputation for having among the best sets of public health data re the Covid19 pandemic. Yet while public health data may be released with much media fanfare, demographic data usually is not. While certain public health data might be classed as ‘black verse’ poetry, and hence of interest to mainstream media, the more prosaic data sourced from ‘births, deaths and marriages’ tends to be overlooked.

In both countries, easily the worst month for epidemic deaths this decade has been December 2022. Yes, 2022, not 2020. In both countries, the people would seem to be largely unaware of this collective death experience. (Awareness would largely be confined to people’s own families.)

The widely endorsed supposition is that the pandemic is over; a concept that may mean different things to different people. To an epidemiologist, an ended pandemic may be followed by a new era with a new normal of permanently higher death rates. To the lay public, or to the pollyanna-ish finance industry which never stops telling us that we are all going to live longer, an ended pandemic means a return to something like pre-covid normality.

I watch DW (Deutsche Welle) world news semi-regularly, and I have heard nothing about the December death spike in Germany. The few stories on DW that I have found in a search today are these:

Winter illnesses burden Germany’s intensive care units, 17 Dec 2022.

Top German virologist says COVID-19 pandemic is over, 26 Dec 2022.

Germany’s new COVID debate [about removing Covid19 restrictions], 29 Dec 2022.

Are Germany’s hospitals in critical condition?, 5 Jan 2023.

WHO discusses end of COVID-19 emergency status [mainly a reference to China], 27 Jan 2023

So, as in New Zealand, there’s acknowledgement of overstretched hospitals and seasonal illnesses. There’s little suggestion that the main culprit is another covid wave, and we certainly have no indication that Covid19 has recently become more virulent.

We should also note that excess deaths in Europe have looked quite dramatic, since 2020, each December. This is because ‘normal’ influenza outbreak deaths – in the 2010s at least – tended to peak in February, not December. The big falloff in excess deaths from January 2023 is to some extent due to death numbers needing to be higher to be classed as ‘excess’.

Both countries started to downplay their reporting of Covid19 deaths from mid-2021; with the exception of Denmark during the first ‘Omicron wave’ of early 2022. We can see that, in February and March 2022, many people in Denmark died of or with Covid19, while significantly fewer people died of other causes. (This substitution of death-causes was much less true for Germany.)

But the situation in Denmark reversed in December 2022, with many more excess deaths than reported Covid19 deaths. Indeed, in Denmark there were five death waves in 2022, and all (except perhaps the relatively small June death wave) correlating with reported Covid deaths (but see the Denmark chart below); expect that the peaks in reported deaths have been increasingly lagging the actual death peaks. The Denmark data in this chart definitely points to all of these death peaks as being associated with recurring waves of Covid19.

We certainly see the same lags in the German data, although the reported death waves are less pronounced. Generally the ‘peaky’ nature of the data suggests that these populations have been facing repeated outbreaks of respiratory viruses; outbreaks for which their immunity levels have been decreasingly able to cope. While waning immunity appears to be the main problem – a problem long known in relation to human coronaviruses – and that waning immunity here likely relates in particular to vaccination immunity, it is also possible that each wave of infection creates new morbidities in the most vulnerable people. So, a combination of waning immunity – covid immunity and general immunity – combined with new comorbidities is leading to progressively higher death tolls. We also note that December in particular is a month characterised by much social ‘mixing and mingling’; conditions ripe for coronaviruses transitioning from epidemic to endemic.

These two charts, for each country separately, also include ‘case’ information. Again, while underreporting is also an increasing feature of case data, this is data that underpins the timing of Covid19 outbreaks (as distinct from non-covid respiratory viruses). Nevertheless, case data may also lag actual deaths, as people are slow to test for covid, and report, until they are fully aware that a new outbreak is taking place.

When we look at Germany, we can clearly see the correlation of case data (in blue) with excess deaths (in red). When there were new covid variants, as in March 2021, August 2021, and January 2022, we can see the uptick in reported cases leading the uptick in deaths. Otherwise, we tend to see the uptick in deaths coming first, with case reporting lagging ever-further behind. December 2022 was particularly striking, with an eventual upsurge in cases suggesting that a significant number of people who died in the death spike were indeed infected with the coronavirus. In Denmark the pattern of excess deaths correlating with reported cases of Covid19 is, if anything, even stronger.

If we think of Covid19 as a three-year pandemic, we can see that in both countries the second half of the pandemic was worse than the first half, regardless of the official cause of death attributed to each casualty.

A death is a death is a death. To dying persons, and their families, it matters little if a death is directly or indirectly caused by Covid19. An indirect death may be due to a loss of general immunity, or to any other factor linked either to the biology of the virus, to fear-induced behaviour changes arising from the attention given to the virus when it was a big story, or to the government mandates ‘to keep us safe’ but which may have (to some extent) substituted one risk for another.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The most important question about the Covid19 pandemic of the early 2020s is ‘how many people died?’. (The second-most important question relates to the impact of the pandemic on people’s ‘quality of life’.) The data here, available since last week, is the starting point for an answer to the first question. This data is as close as can be got to ‘pure facts’, ‘body counts’ in which no expert interpretation plays a role. (This contrasts with ’cause-of-death’ data which requires a doctor’s opinion.) This is raw data. Raw data is true.

There is not really such thing as a ‘global pandemic’, because a pandemic is, by definition, a global event. In a pandemic, individual countries may be understood as ‘administrative regions’. National differences of mortality during a pandemic will be a mix of fortune, prior circumstances, and quality of administration. Re ‘quality of administration’, ‘body counts’ – while most important – do not represent the whole story. We note here my second-most important question, above.

The data above will never be a global total, no matter how long we wait for laggard countries to report. Some countries simply don’t register deaths; these countries are mainly in South Asia and Africa. Some other countries do not share their death tallies with the rest of the world.

The data above is irrefutable, in that it is a simple count of deaths, covering two periods each of four years (209 weeks for those countries which report on a weekly basis). This contrasts with ‘official’ Covid19 death tallies which depend, in each administrative jurisdiction, on some interpretation of what counts as a Covid19 death. ‘Total deaths’ data does not distinguish direct from indirect pandemic deaths.

For most countries, regardless of covid, there would have been an increase in deaths in the most recent ‘quadrennium’ (four-year period) vis-à-vis its predecessor. The major single cause of such covid-unrelated increased deaths is changes in the numbers of ‘elderly’ people, with the precise age of ‘elderly’ being higher in some countries (say Denmark) compared to others (say Lithuania). A country with a high proportion of elderly people need not have a higher percentage increase in deaths from one period to another; however, in these times, most countries are experiencing faster annual increases in their elderly populations than in their younger people.

One complication here is that World War Two ended in 1945, meaning that in 2020, a person born in 1945 turned 75 in 2020. While we are very sure that most countries had higher birth rates after 1945 than before, we are less sure about which countries had the biggest post-war ‘baby booms’.

One question that may be asked is ‘why include 2019 with the other pandemic years?’, given that the pandemic started in 2020. There are two reasons. First, as we have eight years of data conveniently tabulated by ourwordindata.org, the simplest procedure is to compare one quadrennium against the other. The second reason is that death rates in one year may ‘inversely’ impact on the following year’s data. Countries which have above-average levels of epidemic influenza in the year-or-so before a pandemic are likely to have reduced deaths in the first year of that pandemic, because many of the people most vulnerable to infectious diseases have already died. Likewise, re the present pandemic, a benign influenza year in 2019 (such as in Sweden) would of itself postpone deaths until 2020.

Table 1 is not a ‘league table’ of administrative competence, jurisdiction by jurisdiction. Nevertheless, the data shows broad categories of national experiences, and interesting variations (and non-variations) between countries regarded as like. It is a factual unnuanced measure of the different experiences of the Covid19 pandemic in different countries.

Some data highlights:

Scandinavia

As with many social indicators, Scandinavian countries had the lowest amounts of ‘increased death’ arising from the covid pandemic. Within that Nordic group, Sweden is a clear ‘winner’. This is particularly interesting because Sweden gained much publicity in 2020 for its contrary approach to public health administration during the pandemic. Sweden’s state epidemiologist, Anders Tegnell, famously said that Covid19 was a “marathon, not a sprint”. The marathon is now over, and Sweden has at least taken ‘line honours’.

However we should note that Sweden’s second-worst month (for excess deaths) for the whole pandemic was December 2022. (Its worst month was April 2020.) This significant though largely unnoticed fact is also true for other Western European countries. For some the 2022-2023 festive season was the worst three weeks for the entire pandemic. So, we may be looking at Covid19 as an ‘ultra-marathon’ rather than a marathon; if so, we still have years to wait before we can conclusively evaluate the demographic consequences of this pandemic.

The countries which ‘did best’ in the pandemic were those able to confine most of their covid-diagnosed deaths to people who, had they not died of Covid19, very likely would have died from other causes during the pandemic quadrennium.

Western Europe

Countries in Western Europe outside of Scandinavia had increased deaths mainly in the six‑percent to ten‑percent range, with Belgium and Netherlands both just outside of that range (though on either side of it). Interestingly, in the first wave of Covid19, Belgium had many more recorded covid deaths (per capita) than Netherlands. But it was Netherlands which ended up with an ‘above 10 percent’ increase. Netherlands had a bad pandemic.

The United Kingdom came very much in the middle of the Western European ‘pack’.

Two other ‘western’ countries to note are Canada and Israel. Both have increased deaths higher than the European Union and United Kingdom countries.

Australasia

Australia and New Zealand have increased deaths very similar to Western European levels.  ‘Officially’, both have reported fewer covid deaths per capita than do these European countries. This may be due in part to unusually large increases in the elderly populations of Australia and New Zealand; if so, many of these recent additional deaths will be neither directly nor indirectly due to Covid19.

Eastern Europe and East Asia

Both these groups of countries have, for the most part, increased deaths in the ten‑ to twenty‑percent range. This, for East Asia at least, may be a big surprise to the many people who believed that East Asia set the exemplar for best public health policy during the pandemic.

In East Asia, South Korea is a country of particular concern. South Korea has not released weekly death tallies since July 2022; it used to be a reliable reporter of such data. Subsequent Covid19 case data from South Korea suggests that it has experienced two recent waves of Covid19.

Another country for which the Table 1 data may be understated is Hong Kong. December 2022 was known to be China’s worst month, and this showed in the alarming excess death toll for Macao (Hong Kong’s close neighbour) for that month. So the recent Hong Kong data may be substantially revised, or we may see a much bigger toll for Hong Kong in January 2023. (We should note that, in the United Kingdom, there are signs that many people who die in the end of any December have their deaths counted in the following January. Different administrative practices can may weekly data hard to compare across countries.)

For Eastern Europe, I have generally restricted this table to countries in the European Union, though I have included Serbia, showing that its experience is comparable to its European Union neighbours. Eastern Europe did particularly badly in the ‘official’ Covid19 death tallies, in large part due to their high proportions of elderly people. Eastern Europe is a major source of economic migrants. (And, with lower life expectancies than in Western Europe, the threshold age that defines ‘elderly’ in these countries is lower. We may note, as a matter of interest, that the typical life expectancy in Eastern Europe is comparable to New Zealand’s ‘Pasifika’ population.)

An interesting group of Eastern European outliers are the Baltic countries: Lithuania, Latvia, and Estonia. While these recorded high numbers of Covid19 deaths relative to their total populations, the percentage increase in deaths is not so large. This is due to their high but unchanging prevalence of older people. Indeed, their populations probably got slightly younger in 2020 and 2021, as previous high levels of youth emigration will have been stemmed by Covid19 public health controls within the European Union.

South America and the United States

The typical increase in deaths for South American countries is between twenty and forty percent, with Uruguay, Chile and Brazil looking best for those countries with available data. (Argentina is extremely slow at releasing its total death tallies.) Uruguay is easily best.

The high Covid19 mortality of the United States is very apparent in this simple tally of deaths. Indeed USA probably compares better with South America than it does with its European allies. The demography of the United States is like that of New Zealand in some respects, but like South America and Mexico in other respects. Western European (and Australasian) populations have life expectancies above 80. The USA and most South American countries do not. While Covid19 was a disaster for the United States, it may not be that the different public health responses within USA made much difference. It may be that certain known comorbidities – such as diabetes, drug dependency, mental unwellness – are more present in American than in European populations.

Further Interpretation

I have here confined my interpretation of the data to the points which would be best understood by a professional statistician. Further interpretation takes us into the realm of scientific speculation. The science – the testing of plausible explanatory hypotheses with adequate datasets – needs to be done.

The first question begged by the data presented here is why Sweden in particular (and Europe in general) have come out of the pandemic rather well (so far! the ultra-marathon is far from over). The second question is why East Asia has come out so poorly, despite early indications to the contrary.

Sweden coming out of the pandemic marathon so well, and East Asia so problematically, is the inconvenient counter-narrative which happens to be the truth – the poorly understood truth – of the matter.

————-

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Quantity of Life

With mortality data now available for many countries to the end of 2022 (and, for a few countries, for a month into 2023), we are now able to properly assess the demographic cost of the Covid19 pandemic. There is a proviso, in that for a number of countries, December 2022 experienced substantial covid-related mortality, suggesting that the demographic consequences will last into 2023 and probably 2024. The demographic cost includes mortality indirectly caused by both the disease itself and unintended consequences of the public health measures taken to contain the covid coronavirus.

It is important to note that there is an additional impact – the socio-economic impact – arising from the pandemic; a disruptive impact on the living, adverse on balance, that cannot be measured by death tolls, though which may be reflected if life expectancies beyond 2023 fall below previous projections. This is a ‘quality of life’ rather than a ‘quantity of life’ consequence of the public health events of 2020 and 2020.

We also need to note that, with eight years of data shown, we can compare the death tally for the second four years’ [2019‑22] directly with the first four years [2015‑18] (both representing ‘quadrennial’ periods of 209 weeks). It is appropriate that 2019 is included with the subsequent pandemic years, because events in 2019 may contribute to 2020 death numbers; in particular, if 2019 had lower than usual seasonal illnesses, then many people who would normally have died in the winter of 2019 would have died in of 2020 or 2021 instead.) We note that, for the pre-covid years in the 2010s’ decade, the winter death peaks (typically late winter in the northern hemisphere) mainly relate to influenza, with a more general seasonal mortality effect triggered by ‘common colds’ and the like.

(And one piece of technical information. The weekly death tolls charted have been ‘smoothed’ to reduce the effects of random variation. Thus, the peak numbers for Australia and New Zealand – in July 2022 – each represent a weighted average of the deaths for that actual week and the deaths for the weeks immediately before and after. It means that the actual numbers of deaths in the peak weeks were higher than shown.)

Australia and New Zealand

I start with Australia, which in the past I have been unable to chart very well because the Australian authorities are unnecessarily slow to release mortality data, and when they do it’s typically up to ten weeks of data at a time. We note, with reference to the New Zealand chart, that Australia’s population is slightly more than five times more than New Zealand’s, and has a similar age structure.

A baseline mortality toll for Australia in the early 2020s would be 3,000 deaths per week; it is shown by the black ‘predict 2020’ plot on the chart. The pattern of increasing deaths from 2015 to 2019 is mainly due to population increase; particularly, increased numbers of people aged over 70. (Baseline pre-pandemic weekly mortality for New Zealand is about 600.)

The scale of these charts is (approximately) with the top number (eg 5,000 for Australia) being double the bottom number (eg 2,500). In the first year of the pandemic, 2020, Australia’s weekly death tallies were much like what would have been expected in a non-covid year, but with lower winter deaths. That 2020 experience reflects the public health restrictions which were imposed, with the international travel restrictions probably being most important.

The demographic cost – in deaths – really begins in Australia in April 2021, and clearly persists to the end of (and most likely beyond) 2022. (A look at my previous set of chartssuggests that, based on northern hemisphere experiences of late 2022, mid-2023 – winter – could be particularly bad in the southern hemisphere; although taking advantage of the seasonal time-lag to get suitable vaccination programs in place may help the ‘lucky south’.)

New Zealand data reveal a broadly similar pattern to Australia, though the 2020 shortfall of deaths was more accentuated, and the more general appearance of excess deaths began seven moths later than in Australia, in November 2021. New Zealand, having a much smaller population than Australia, shows more exaggerated peaks and troughs; and generally more ‘random noise’, more random variability, from week to week.

Australia’s pre-covid estimated life expectancy (source: IMF database) was 83.4, with a median population age of 37.9. For New Zealand, the comparable numbers were 82.3 and 37.9; so, both countries have very similar demographics.

I have calculated the increase in quadrennial (ie four-yearly) total deaths for 2019‑22 compared to 2015‑18. For Australia, with December 2022 data still missing, it’s an estimated 7.9% increase in deaths. For New Zealand it’s 8.2%, slightly higher. For both countries, 2022 is the year that most contributed to the 2019-22 excess deaths’ toll. In addition to Covid19, increases in these countries’ populations of elderly will have contributed to higher mortality this decade.

Finland and Sweden

Finland has just a few hundred thousand more people than New Zealand. Its 2019 estimated life expectancy was lower than Australia’s and New Zealand’s, at 81.9. And its median age, at 42.8, suggests that Finland should have a significantly higher base rate of deaths per week. Indeed, the chart shows a base weekly tally of about 950, substantially higher than New Zealand’s 600.

Finland was the most restrictive of the Nordic countries, with respect to public health management of Covid19. It’s pattern of weekly deaths looks very similar to New Zealand’s, with New Zealand having a six-month (seasonal) lag. This means that the extent that Finland’s deaths are above baseline from July 2022 seems likely to be a good predictor for New Zealand in the first half of 2023.

Finland had 7.2% more deaths in 2019‑22, compared to the previous four years. The main explanation for its lower mortality increase (than Australasia) is likely to be due to a lower rate of increase in the size of its elderly population.

Sweden has close to double New Zealand’s population, and had almost the same pre-covid life expectancy as Australia and New Zealand (82.8). Its median age was 41.0; lower than Finland’s, probably due to both more immigration (especially refugees) and (given Finland’s membership of the Eurozone) less emigration than Finland to places like Berlin and Brussels. As Sweden’s chart shows, Sweden’s pre-covid baseline mortality was about 1,600 deaths per week, equivalent to 800 in New Zealand; again, higher due to Sweden having relatively more older people than Australia or New Zealand.

Unlike the other countries mentioned so far, Sweden had almost all its pandemic excess mortality in the year commencing March 2020. One important reason for this was Sweden’s unexpectedly low mortality in the year-and-a-half prior (especially in the winters of 2018/19 and 2019/20) to Covid19, meaning that many of the covid deaths of the frail elderly in Sweden in 2020 were people who, based on statistical expectation, would normally have died before the pandemic struck. The other main reason for Sweden’s high early toll was its classically liberal public health policy approach.

2021 in Sweden looks very much like a normal pre-covid year in Sweden. In the second half of 2022, all west European countries had significantly elevated mortality, mainly it would seem associated with renewed Covid19 outbreaks. Comparing Finland and Sweden, both had December mortality peaks about 50% above baseline weekly death tallies. But Finland clearly had much worse autumns than Sweden, in both 2021 and 2022.

In an important sense, many people died in Finland in 2021 and 2022 who would have died in 2020 were it not for the public health mandates then operating in Finland but not Sweden.

There’s more to it than that though. Sweden’s increase in deaths for 2019‑22 compared to the previous four years was just 2.2% compared to Finland’s 7.2%. It looks as though Sweden’s population had a significantly higher level of general immunity in 2021 and 2022 than did Finland’s. Finland experienced a greater overall demographic cost from the pandemic – presumably in large part because of its different policy choices – than did Sweden.

Both of these Nordic countries (and the other Nordic countries too) had a lower increase in deaths than did Australia and New Zealand. But the Australasian countries probably had bigger increases in the elderly populations (defining ‘elderly’ as over 70); nevertheless, Australia and New Zealand most definitely should not be congratulating themselves on having ‘done better in the Covid19 pandemic’ than Sweden.

Greece

Greece is a European Union country with essentially the same population size as Sweden. It has the same pre-pandemic estimated life expectancy as the other countries charted so far (82.2), but has a significantly older population structure (pre-covid median age of 45.3), almost certainly due to a substantial depopulation of young adults in the wake of the Eurozone crisis which peaked in 2012. So we expect baseline weekly deaths to be substantially higher in Greece than in Sweden. Indeed, that is so. Pre-pandemic baseline weekly deaths in Greece stand at 2,200 compared to 1,600 in Sweden. (We note that much of this depopulation of young adults from Greece will have occurred before 2015, so we have no reason to believe that there has been a significant change in the age structure of Greece that would contribute to the pandemic-era death experience. Hence, the dominant reason for increased deaths in Greece will be Covid19.)

The chart for Greece shows no impact arising from Covid19 until October 2020. While this will be due to restrictive public health measures in the first half of 2020, it may also be due to high levels of influenza that seem to have been generally the case in southeast Europe in February 2020. This will mean that more vulnerable older Greeks will have died just before the pandemic, rather than in its early phase. And it means that Greeks will have had higher levels of general immunity in those early months.

The second European wave of Covid19, in late 2020, substantially impacted Greece, as it did all other European countries. In general, it had a disproportionate impact on vulnerable Eastern Europeans, with less natural immunity than their western co-Europeans. Greece had 11.1% more deaths in 2019‑22 compared to 2015‑18.

We see that Greece has had an unusually high level of summer epidemic death, almost certainly due to its status as a tourist destination. In 2021 Greece was a substantial victim during the ‘Delta’ outbreak; based on comparative analysis this toll was due mainly to compromised immunity (including inadequate vaccination) and not to any specifically deadly features of the ‘Delta’ variant of the Covid19 coronavirus. We see Greece with the same 50%-above-baseline mortality in December 2022 as other European countries. Southern hemisphere countries can expect this sort of peak around June 2023, though preventable by mass vaccination this April.

Chile

Chile is of interest as an alternative southern hemisphere country. It has almost exactly four times New Zealand’s population. It has a pre-pandemic baseline death rate of 1,900 which would be equivalent to 475 in New Zealand; that is, well below New Zealand’s baseline weekly death tally of 600. While this is due to Chile’s lower pre-pandemic median age (35.4 compared to Australasia’s 37.9), it also reflects Chile’s high – for the Americas – pre-pandemic life expectancy estimate of 80.2.

We see that Chile was largely successful at keeping Covid19 at bay until May 2020. But it was always more challenging in a country with land borders to three other countries, with the Peruvian and Bolivian border most significant at that stage. And clearly the international airport in Santiago was not as well sealed as those in Australasia. Unlike Australia and New Zealand, Chile was hit hard in the first half of 2021, and had a much bigger (‘Omicron’ variant) peak mortality in early 2022. Other than that, its experience in late 2021 and late 2022 was similar to Australia’s. Overall, Chile saw a 20.1% increase in mortality in 2019‑22 compared to 2015‑18, accentuated by its surprisingly low pre-pandemic base. Low pandemic mortality for South America, but high for the world as a whole. Undoubtedly Chile has also had a substantial increase in its older population, though maybe not as marked as in those countries whose birth rates were most affected by World War II.

United States

The United States is like a mix of Australia and Chile. It had a pre-pandemic estimated life expectancy of 78.9, well below both Chile and Australia; indeed a ‘third world’ life-expectancy. The median age pre-covid was 38.3, very similar to Australia and New Zealand; so having a similar age structure to the Australasian countries.

USA has a population 66 times that of New Zealand. It’s base pre-pandemic weekly mortality of 52,000 is equivalent to a high 790 for New Zealand, much higher than New Zealand’s actual 600 baseline. Given the similar age structures of United States and New Zealand, this has to be due to a generally higher pre-pandemic risk of death from all causes in the United States.

Overall, the chart for USA is comparable with the chart for Chile. Deaths in the United States increased by 16.8% over the two periods; like Chile but without the low base that might have inflated this statistic for Chile. Covid19 was clearly a big disaster for the United States, though I suspect that high levels of comorbidity – including obesity, diabetes, drug misuse, mental unwellness – contributed more than public policy. Indeed, USA had a younger profile of Covid19 mortality than most other countries.

United Kingdom

Finally, we look at the United Kingdom, with a baseline pre-pandemic death tally of 10,500 per week. That translates to 775 in New Zealand. The United Kingdom had a median age of 40.8 pre-pandemic, similar to Sweden. And a life expectancy of 81.3, similar to Finland. This older population structure largely explains Britain’s higher (than Australasia and Chile) level of baseline mortality.

Mortality patterns in the United Kingdom were most like that of Sweden, except generally higher, and significantly worse in 2021. We should also notice Britain’s huge peak in January 2023, probably partly due to Britain’s unique way of reporting death statistics. (Many late December deaths are recorded in the following year; this seems to be true for each year.)

Overall, the United Kingdom’s increase in deaths for 2019‑22 over 2015‑18 was 7.9%, same as Australia and less than New Zealand. However, we cannot claim that the demographic impact in the United Kingdom, so far, is less than in New Zealand. In the absence of the Covid19 pandemic, New Zealand’s deaths would have increased by more. This is because New Zealand’s increase in the number of people aged over 70 is more pronounced than is Britain’s. It may also be due to a worrying rate of increase in New Zealand of sub-70 non-covid mortality; that is, of mortality of people aged under 70 with American-style comorbidities. We note this already, in that the average age of people in New Zealand who die with Covid19 is significantly younger than the average age of those who die of Covid19.

Brief Conclusion

Sweden, with its more classically liberal public health approach, has had easily the lowest pandemic and post-pandemic mortality increases so far.

Data from other countries suggests that factors around general immunity and comorbidity are the major determinants of pandemic mortality; that is, neither infection rates nor viral mutation rates. And we should note that Covid19 vaccinations also probably improve general immunity, and not just immunity to the covid virus.

Re Australia and New Zealand, the lessons are two: the severe Covid19 winter of 2022/23 in (especially western) Europe will likely be replicated in Australia, New Zealand and Chile if mass revaccination does not take place this April. And that these southern countries may have unaddressed comorbidity problems, not unlike the problem in clear sight in the United States; in New Zealand this problem is by no means confined to people of Māori and Pasifika ethnicity.

______________

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

The first of the above charts shows excess seasonal deaths in the United Kingdom as winter approaches. While the huge mortality peaks of the pandemic in Britain are long past, we do see significant excess mortality in the United Kingdom since April. And October 2022 excess mortality rates are higher than those of 2020 and 2021. The situation in Netherlands is similar.

The story in Germany is similar, since June 2022. While the recent unseasonal mortality peak – due to Covid19 if reports are accurate – is now waning, it seems likely that Germany will face another mortality peak comparable with its influenza peak of 2016/17. In Germany, none of its Covid waves had as much peak excess mortality as the influenza peak of February/March 2018.

The data is not as up-to-date in the following three countries.

For New Zealand, we see the Covid19 deaths as a period of rising seasonal excess from February to July this year. Then there was a sharp drop-off, meaning that from August New Zealand experienced just its normal late-winter seasonal mortality. While some of this was due to Covid19, it was offset by lower-than-expected deaths from other seasonal illnesses. My hunch is that New Zealand will see a summer Covid19 mortality peak; not as high as the July peak, but unambiguously Covid19.

The other New Zealand story is the unexplained winter mortality peak of 2021. All the New Zealand public knows about this is that it was neither Covid19 nor Influenza. It might have been due in part to RSV, which hospitalised many young children in 2021. In the United States at present, we are getting reports (eg from ABC News this week) of a “tripledemic”, which includes a nasty ‘flu’ and an RSV outbreak that is hospitalising older Americans as well as children.

Finland is one country in Europe which New Zealand likes to compare itself to. Finland avoided the dramatic Covid19 peaks experienced by United Kingdom and Netherlands. But it has had worryingly high excess mortality since June 2021, and continues to do so.

Sweden, unlike Finland, had two big peaks of Covid19 mortality in 2020. Since then, Sweden has generally looked much better than all other European countries. Nevertheless, Sweden did have a problem, presumably a Covid19 problem, from June to October 2022. Though not as bad as Finland.

Overall, the pattern seems to be that populations are becoming more vulnerable to respiratory illnesses. If people who have previously had Covid19 are dying more, then damage already done by the SARS-Cov2 virus is likely to be the main culprit. If people who did not get Covid19 previously are facing a higher risk of death from respiratory illness, then the main problem is likely to be compromised general immunity arising from reduced general community contact with these types of viruses.

The post-covid mortality problem is slightly worse than it appears, especially if we consider United Kingdom and Netherlands, both countries with high early death tolls from Covid19. In these countries, many of the people most vulnerable to Covid19 have already died. So the denominator populations are, disproportionately, covid survivors (meaning either they had it and recovered, or they avoided it). Typically, after a demographically-significant epidemic, subsequent death rates should be below the historical average.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Chile is an important comparator country to New Zealand; in terms of physical geography, political economy, and demography (including conflict between indigenous peoples and ‘colonisers’). In terms of physical geography, it is mostly temperate; and, most importantly, it is in the southern hemisphere.

The above chart shows excess deaths in Chile since January 2020. (While I have a similar chart for New Zealand, it’s too ‘noisy’ to show here; this is due to New Zealand’s small population. Chile has four times as many people as New Zealand.)

This chart shows weekly peaks for excess death, by age group. (The data is slightly smoothed. To reduce random noise, it uses a kind of three-period centred moving average, in which the featured week has a 50% weight, and the weeks before and after each have a 25% weight.)

For Chile, we see two very sharp peaks of covid mortality, in May/June 2020 and February 2022. Neither of these should really be characterised as ‘winter outbreaks’ of Covid19. We also see a longer lasting outbreak in the first half of 2021, also not winter.

Chile did its best to suppress Covid19, but it was always harder to do this in a continental country than in an archipelago such as New Zealand. Chilean covid policy was more particular than most countries in trying to protect the elderly. This shows in the initial 2020 peak; and also in 2021 when Chile was one of the early vaccinators, and gave significant priority to its oldest citizens.

By 2022, much of the immunity – from vaccination, from the 2021 epidemic, and from other respiratory viruses – had clearly waned. So, when covid reappeared in the summer of 2021/22, there was a large cohort of particularly vulnerable older people, who died in huge numbers. We also see the second 2022 wave in Chile as in New Zealand, with winter deaths peaking in July; again mostly older people.

To understand what happens in winter, we need to take account of the other winter illnesses, of which influenza is epidemic and common colds are endemic.

The above charts show this more contextualised picture, for Chile and for New Zealand. (Again, these charts use the centred moving average technique, which flattens the peaks in the charts.) Chile typically has 2,000 deaths per week, whereas New Zealand typically has 600. We also note that the plots for 2016 to 2019 in Chile are closer together than for New Zealand; suggesting that Chile is not aging as fast as New Zealand, and also that Chile’s annual population growth is less than New Zealand’s (reflecting New Zealand’s higher net immigration in the late 2010s).

In both charts the black line shows ‘expected deaths’ for 2020. (I have calculated this myself, and have shownthat my calculation is credible.) One problem with the age-based ‘excess deaths’ data from ourworldindata.org is that it is hard to check the validity of their estimates for expected deaths. I have also calculated expected (ie predicted) deaths for 2021 and 2022, though not shown these in order to keep the charts ‘clean’. (The annual percentage increase in predicted deaths for New Zealand was more than double the equivalent calculation for Chile, on account of New Zealand’s faster aging and faster general population growth.)

Chile’s first mortality peak, in 2020, was delayed until early winter thanks to public health measures. In Chile, June has been the peak month for mortality linked to influenza and common colds. Excess deaths (eg for 2020) is the difference between actual deaths and predicted deaths. Thus, in June 2020 in Chile, ‘excess deaths’ attributable to Covid19 understates actual ‘Covid19 deaths’, because we know that covid to a large extent displaced influenzas and common colds. This was different this year, however, with the peak of excess deaths being July rather than June. The most pertinent measure is ‘excess deaths for all seasonal illnesses’. By this measure we can see that excess weekly deaths, normally about 500 in June, grew to about 2,000 in June 2020.

In New Zealand, excess deaths were substantially negative in 2020, in the absence of deaths from influenza and covid. And common cold related deaths were less than usual. 2021 seasonal mortality was about normal, with certain unspecified ‘cold’ conditions being the main culprit that winter. December in 2020 and 2021 saw more total deaths than would normally have been expected. The 2022 (belated) outbreak in New Zealand saw substantial though not dramatic excess covid deaths from March to July.

Mortality data for New Zealand released to international databases shows a sharp decline in excess covid deaths from August this year. Some of this data may be upwardly revised in coming releases.

We should note that the actual weekly death peak for New Zealand is 965 (week ended 26 July 2022); it looks less than this on the chart due to the moving average ‘anti-noise’ technique used. That is the highest number of deaths in any week in New Zealand’s history since November 1918. (Before 2022, that record was 863 deaths in the week to 2 August 2017, a week of record influenza-caused deaths.)

The final two charts show excess deaths in both countries as percentages above ‘normal’ for each of the three covid years; we note that ‘normal’ is adjusted for each year, and it represents the absence of both covid and seasonal increments to the death rate. For New Zealand the main feature to look out for is whether the rapid drop-off in reported deaths from late July is confirmed.

Chile’s excess deaths this winter have been significantly higher than New Zealand’s, with a July peak of 56% (covid and seasonal) and an August peak of 38%. New Zealand is not exceptional, however. My sense is that, this coming summer, excess deaths in New Zealand (in percentage terms) will exceed excess deaths in Chile.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

A few weeks ago this article from Stuff (24 Sep) was drawn to my attention: The shocking stats that prove Covid19 does not kill equally. While I have some problems with the article’s interpretation of the data, by and large the article itself was reasonable. Less so the headline, which is different in the online version of the story. (The print version omits the sensationalist word ‘shocking’. I wonder if there is a general pattern to present online versions of stories in a more sensationalist manner?)

In fact, there’s nothing surprising about the Covid19 statistics indicating more Māori and Pasifika deaths. The charts in the article, however, perpetrate the erroneous interpretation in the headline, that Covid19 is in some evil sense discriminatory. (In particular the chart that shows death rates of people aged over 90.) The facts are that, among population subgroups, Covid19 mortality reflects differences in general mortality.

With respect to recent data (ie since the beginning of August), just two Pasifika people have died of Covid19 (as the ‘underlying cause); one in their 70s, and one aged under 60 (though 20 others have died ‘with’ Covid19). Ten Māori have died of Covid19 in that recent time period (32 others ‘with’ covid), only one of whom was over 90. In that same time period, 176 people of ‘European or other’ ethnicity died of Covid19, 56 of whom were aged over 90.

While the article was about the whole pandemic, not just its recent phase, it remains no more useful to focus on deaths of Pasifika aged over 90 than it does to focus on Pakeha aged over 100. Regardless of Covid19, the probability that a New Zealander of Pacific Island descent will reach the age of 90 is similar to the probability that a Pakeha New Zealander will reach 100.

Life Expectancy by Ethnicity

Ethnicity statistics in Aotearoa New Zealand should always be treated with caution. For example, a person with just one Māori great-grandparent would typically be classified as Māori, regardless of the ethnicities of the other seven great-grandparents. (This suggests that the mortality and morbidity statistics are even worse for people whose predominant ethnicity is Māori, especially for people who are perceived as Māori [either due to their name or to their appearance].)

If we try to compare socio-economic ‘apples’ with socio-economic ‘apples’, we have almost no data which can give the true picture. My suspicion is that a dapper Māori man such as Scotty Morrison has a similar life expectancy as his Pakeha equivalents. And my suspicion is that a Pakeha solo-mum with three children in emergency housing has a similar life expectancy as a Māori solo-mum in the same situation. Generally, we are very light on evidence that Māori and Pasifika people have lower life expectancies than their Pakeha neighbours.

The problem is that, in proportion to their sub-population totals, relatively more Māori and Pasifika are in impoverished or facing other stressful life circumstances. The telling Covid19 statistics are shown in the table below. The median ages of death for the different ethnicities, estimated from Ministry of Health covid mortality data, are:

We may regard the final row of the above table as valid estimates for life expectancy for each sub-population. It’s not a pretty picture for Māori. What it says is that Māori have significantly more health comorbidities than other ethnic groups. In particular, this represents the socio-economic circumstances that many Māori face; this reflects significant inequality among Aotearoans in general, and also within the Māori sub-population. It reflects the substantial problem in New Zealand of rural poverty. It reflects historical circumstances faced by indigenous peoples throughout the ‘new world’. Some of that is due to biological circumstances which go back into deep history. These include being at the wrong end of the immunity gradients for most diseases during the globalisation phases of world history; and it reflects – for example – an evolutionary context which, among other things, makes Māori (and Pasifika) people comparatively intolerant to alcohol and sugar. (In Eurasia, tolerance to alcohol was a biological adaptation to the problem of water-born diseases. In England a culture developed, especially among men, of drinking ‘small beer’ instead of water. In other places, wine and other fermented drinks were partial substitutes for water.)

The above ‘life-expectancy’ table (especially the second row) also suggests that people with substantial comorbidities are more likely to get Covid19 as well as being more likely to die of Covid19.

Getting back to the interpretation of the Stuff article, because life expectancy for Māori and Pasifika is so much lower than for Pakeha, it means that we should be comparing covid mortality rates for Māori/Pasifika in their 60s with Pakeha in their 70s, comparing Māori/Pasifika in their 70s with Pakeha in their 80s, and Māori/Pasifika in their 80s with Pakeha in their 90s. A randomly chosen Māori person of a particular age, in essence, is as likely to die (from any cause) within twelve months as a similarly chosen Pakeha person ten years older.

While discrimination is certainly part of the ‘historical circumstance’ problem many Māori face, Covid19 doesn’t add to that problem; it simply reflects it. (And we should note that historical discrimination is more nuanced than intellectually-lazy words like ‘colonisation’ or ‘imperialism’ convey. Subsequent to the era of industrial capitalism which began at scale around 200 years ago, life expectancies have increased, with the life expectancies of ‘white’ people (and latterly East Asian people) increasing the most. While non-imperial historical counterfactuals might have had a smaller life-expectancy gap between Polynesian peoples and (say) Anglo-Celtic peoples, it is unlikely any such counterfactuals could have achieved a higher life expectancy for Māori than Māori have now.

Co-mortality and critical states

The final issue of importance to note is that ’cause of death’ is not a simple discrete matter. In an important sense, probably most deaths are due to ‘old age’, but few other than the Queen of England have the privilege of having ‘old age’ listed as their sole cause of death.

The reality is that most deaths have more than one ‘clinical’ cause, and environmental events such as pandemics can kill in non-clinical as well as clinical ways. (Poor quality, under-resourced, or inaccessible health services count here.) In the fable of ‘the straw breaking the camel’s back’, ‘straw’ would never be listed as the sole cause of that camel’s subsequent death. The camel was in a critical state before the straw added, fatally, to its burden.

It’s interesting that ‘comorbidity’ is listed as a word in the dictionary, but ‘comortality’ is not.

Co-mortality is the reason why an ‘excess deaths’ approach is the best indicator of the scale of epidemic deaths. The sadness is that demography is the poor cousin of social science. Much core demographic data – births by sex, deaths by age/sex and place of birth, arrival/departures by age/sex and place of birth – is hard to find in even the rich world. In many countries it remains largely absent. Population censuses are required to make up for poor record-keeping; but too often they are under-resourced, and the value of the core demographic information is under-understood. In New Zealand we remain substantially ignorant about intra-national population movements.

Almost all the Covid19 deaths tallied – whether ‘with covid’, of ‘covid’, or ‘as a consequence of covid’ – are in fact co-mortal deaths. Very few people have died of Covid19 without some other vulnerability being present.

What matters most is an understanding of critical states. Typically, when things go wrong there are multiple causes. Today we use the increasingly popular (and indeed overused) phrase ‘perfect storm’ to indicate the problem. A person is in a critical state when just one additional factor will kill them. (High blood-pressure is one oft-cited factor that can contribute to a person being in a critical state.)

Surprisingly, a person in a critical state may to all intents and purposes be healthy. One person may be much more vulnerable than another to a particular fatal illness, but not obviously so. One seemingly small trigger event may have a fatal impact on that person, but may have no impact on the other person. For a person in a critical state, a trigger event may be sufficient to cause death. (Or, as in ‘chronic fatigue syndrome’ which incorporates ‘long covid’, a trigger event can initiate a long period of chronic unwellness for some people but not others.) An otherwise healthy individual with, say, the Huntingdon’s gene may in fact be in a critical state; one small trigger may unleash the uncurable disease.

‘Old age’ is a critical state; a state which some people – and some peoples – reach earlier than others.

Systems may or may not be in a critical state. (The name ‘critical state’ comes from the nuclear sciences.) A highly stressed population is likely to be in a critical state; especially a population having to constantly negotiate with unsympathetic bureaucracies, permanently raising the levels of cortisol in the bloodstream. A homeless person will likely be in a critical state, meaning that a trigger such as a covid infection could have elevated consequences.

If Māori are more likely to be homeless, or have no socially-approved source of income, then Māori are more likely to die in a pandemic. And die at younger ages.

Is Covid19 a trigger that’s causing human existence to unravel? I suspect that a keyword search on the word ‘existential’ would show a big uptick this decade. Growth capitalism, the way we practice it, is a system that places most people in a near-critical state. (If a critical state is when one more aggravation is fatal, then a near-critical state is when two more aggravations are fatal.) There are signs that this comparatively mild infectious disease has triggered a turning point in global history, and that’s partly because the climate system was already in a critical state.

While Covid19 has been a disease of the rich, and spread mainly by the rich, it is a disease that has revealed the widespread comorbidities – critical states and near-critical states – which our economically vulnerable populations experience. So while the (often oblivious) privileged sub-populations spread covid more while suffering less – not unlike the environmental consequences of careless human behaviour – it is the sub-populations in critical and near-critical states who die the most.

Our systemic problems are our systems in critical states; superficially they may have looked healthy before 2020. Since then, for those willing to see, Covid19 has become the highlighter, not the central problem. If we are homo sapiens – wise ‘men’ – we will look to solutions which destress our systems. What we should not do is aggravate our systemic problems by converting near-critical systems into critical systems; into systems that can become ‘perfect storms’, destroyable by mere straws.

Is Covid19 returning?

My recent statistical analyses suggest ‘yes’. See my Tourist Europe again: Covid19 Waves compared (Evening Report, 17 October 2022). Spain has just had its worst round of Covid19 deaths since April 2020. Covid’s arriving now, in those full aircraft coming here. October is peak season for people flying from Europe to New Zealand. And of course, covid is circulating domestically, and is known to resurge as immunity wanes. It is far too soon to think that Covid19 has become just a winter problem.

If New Zealand’s ability to cope in 2020 and 2021 was weak, and required drastic emergency measures then, New Zealand’s ability to cope this summer may be even weaker. Aotearoa New Zealand is now facing a ‘cost-of-living crisis’, in an economy with severe labour shortages. The New Zealand economy is in a critical state; it is ‘supply inelastic’, meaning it has no surge capacity to respond to a new imported crisis.

Policy is now focussed on creating a recession, the only way today’s policymakers believe they can respond to recent increases in the price level. (Refer to these RNZ stories: Domestic inflation rise ‘a shocker’ – economists; Analysis: Inflation rate higher than expected; Unemployment next challenge for economy – experts.)

New Zealand as a whole was not in a critical state in February 2020. It’s closer to being in such a state today. Marginalised sub-populations will be affected most from any trigger events this decade. Many, but by no means all, vulnerable New Zealanders are of Māori or Pasifika ethnicity.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

Here we see charts of the same ten western tourist countries of Europe (refer to my Travel Warning: Excess Summer Deaths in Europe, 5 Oct 2022). It is important to note from the outset that ‘reported death’ figures have become increasingly unreliable. And reported cases are now unreliable indicators of the size of a covid outbreak; they do indicate the timing of covid waves, however, albeit with increasing time lags relative to the date of infection. The most reliable data is ‘excess deaths’. (Indeed Michael Baker has recently taken to citing this measure; see NZ Herald 16 October Covid19 Omicron outbreak; Another Wave.)

The Mediterranean tourist countries most popular with English-speaking visitors are shown first. All charts are to the same scale. We start with Greece this time.

Greece is one country that, using reported covid data, did not seem to be impacted much until the very end of 2020. However excess deaths in 2020 do suggest that there was a summer wave of Covid19; in addition, Greece was suffering excess deaths from influenza in the beginning months of the covid pandemic. The chart for Greece is quite dramatic-looking in 2022, with a significant death-wave in July (almost certainly continuing into August). We note that, re cases, July is worse than August; this represents the peak of American tourism into Greece, with intra-European vacations peaking in August.

Moving west to Italy, we see again the substantial covid death wave in July 2022. And we see, more clearly than Greece, the most recent (autumn) wave. This may be the wave of greatest impact, worldwide. Moving on to Spain, we see that the July death wave there is Spain’s highest covid death-wave since March 2020. And, by looking at reported cases, this July wave of excess deaths is clearly attributable to Covid19. (Note that 1,000 daily deaths per 100 million people is the same as 10 daily deaths per million; equivalent to 50 daily deaths in New Zealand. Indeed some of these deaths may be New Zealand tourists in Spain.)

Finally for South Europe, in the far southwest we see that Portugal had a bigger autumn wave of Covid19 (May) than elsewhere, while also catching the July wave early. Its July 2022 peak is as big as (or bigger than) its three 2020 peaks.

Looking at West Europe’s ‘big three’, the UK has had smaller summer peaks than the Mediterranean countries, probably because the ratio of visitors to residents is much lower than in (say) Greece and Italy. But we do see a clear rise in cases in June, meaning that the actual rise in cases will have been in May, when the summer tourism season in Britain gets underway. And excess daily deaths have averaged three in a million since May, equivalent in New Zealand to 2,700 pandemic deaths over a six-month period. We also note that these deaths occur in a context where many of the most vulnerable people had died in earlier covid waves.

France represents the high point of the immunity gradient, where many cases translate to relatively few domestic deaths; but where visitors can expect to be especially vulnerable. We see clear covid death peaks in April (when the French tourist season is well under way) and again in July, the peak month for English-speaking tourists. And, like Italy, we see that the latest wave is well underway.

Germany has seen a bigger covid death wave in the later summer; and we note that reported cases in Germany precede deaths, which is what happens in a country with an efficient reporting system. The latest wave of cases is prominent in Germany; deaths will come this month and next, not unlike 2021.

Now the mountain tourist areas, attractive in summer to people who like to walk (and to travel about by train) rather than to swim and sunbathe. Austria shows a pattern very similar to Germany, with a very marked October wave of cases. Austria, like Germany, was a leader in the autumn wave of 2021. Switzerland shows the same patterns as Austria, though with less amplitude. Norway is the most muted of these destinations. Yet its excess deaths are higher relative to reported cases than most of the other countries shown here. Reliable once for its excellent record-keeping, Norway seems to have lost interest in Covid19. My sense is that their autumn deaths will be much as they were in 2021. For Norwegians, October and November 2021 represented their biggest death-wave of the pandemic.

It is important to note that October is a peak month for air travel from Europe and North America to Aotearoa New Zealand. As I have previously noted (see Covid Waves in Europe, Lesson for New Zealand 26 July, and Covid Demography: New Zealand in context of North and West Europe 19 August), New Zealand can expect a late spring or summer covid wave. New Zealand, which has no regular summer seasonal pattern of high deaths, may well in a few months see its highest peak of excess deaths for the whole of the pandemic. Indeed, as has been announced by New Zealand’s Ministry of Health today (RNZ), Covid19 is waxing once again.

As has been the case in 2022 so far (but not in 2020 and 2021) New Zealand’s covid deaths most likely will be nameless and faceless statistics. Out of sight – as reporters, politicians and public servants take their holidays – and out of mind.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.

In many ways the world has moved on from the Covid19 pandemic. And excess mortality from Covid19 mingles with excess mortality from other causes, from heatwaves to other diseases. The summer is useful because it is generally the low season for infectious respiratory illnesses.

Covid19 is a disease that is spread through air travel, and through air; not so much in airplanes or even airports, but by travellers and the people who mingle with them and who provide services to them. West Europe is the world’s preeminent tourist destination, especially in summer.

People come to the Mediterranean coastal tourist destinations in particular, but also to the places with spectacular mountain scenery. In addition, the United Kingdom, France and Germany are strong all-round summer tourist destinations. Of the Mediterranean destinations, Spain in particular has a strong ‘shoulder season’, especially for people travelling from or via the United Kingdom. Travel from North America to Europe surges in late June and peaks in July.

Travel by northern Europeans to southern Europe tends to peak in August, long regarded as the ‘silly season’ in Europe. We should note that, as a result of this ‘silly season’, the releases of mortality and other data are delayed, and data released in September may be revised upwards by more than data released in other months. So deaths in August in particular may be higher than is shown here for those countries which have released August data.

Looking at the Mediterranean countries, we see Spain with significant and unusual mortality since April 2022, with a substantial peak in early July, just before this summer’s big wildfire peak. The excess deaths look, in the main, more like Covid19 deaths than weather-related deaths. Portugal shows a similar (if less dramatic) pattern to Spain.

Italy is as dramatic as Spain, but peaking at the end of July, and without the earlier plateau of deaths from April to June. It has the look of a wave starting in Spain and moving on to Italy. The chart for Greece reinforces this perception. (And note how different Greece is from Spain and Italy in 2020 and 2021; Greece has been very vulnerable in the later part of the ongoing pandemic.) We should also note that Malta and Cyprus show the same patterns as Italy and Greece.

This pattern of west to east travel mortality suggests, in particular the impact of tourism from North America, and also tourism from places like Australia and South America. Trips to Europe from Australia and New Zealand tend to start in the United Kingdom in April and May, and then move to Spain (before it gets too hot) and then to other destinations in Europe.

Looking at the mountain tourist countries, Switzerland certainly shows a July peak, and it looks (as in 2015) to be a bout of epidemic illnesses. Austria looks much like Switzerland this year. Finally, Norway also shows those same patterns (more like Switzerland than Sweden or Denmark). And we should note that Iceland, now a tourist darling, shows a similar pattern to Norway.

Looking at the big-three, the United Kingdom shows significant excess deaths from late April. United Kingdom data can be a bit hard to read, because it seems to show dates of registration of deaths rather than actual dates of death. The data shows lulls in the registration of deaths around public holidays, and especially the lead-up to Christmas. France shows the July mortality peak seen elsewhere (covering its peak tourism from United Kingdom and United States); otherwise, 2022 in France has been mainly a normal year. Germany also shows that peak, while showing signs that late summer deaths will also be higher in 2022 than in 2020 and 2021.

The charts overall suggest that travellers themselves may have been dying away from their home countries in significantly greater numbers than usual. And we should note that travellers include seasonal workers (including ‘backpackers’). Travellers face danger when populations in their countries of origin have reduced immunity from diseases endemic in destination countries.

This raises the issue of one big hole in demographic data, affecting the world as a whole, and especially countries (such as New Zealand) with large flows of people – as tourists and as migrants – relative to the normally resident population. For example, New Zealand’s ‘life expectancy at birth’ estimates (based on mortality data) cannot take into account many of the many New Zealand born people who die overseas. And they do include people – many people – not actually born in New Zealand.

*******

Keith Rankin (keith at rankin dot nz), trained as an economic historian, is a retired lecturer in Economics and Statistics. He lives in Auckland, New Zealand.