Which governments have been most restrictive?

Guess which countries in Europe have had the lowest average restrictions on individual behaviour from March to December according to researchers at the Blavatnik school of government in Oxford? Guess which countries in the world have had the most or least stringent government restrictions? Hint: Sweden doesn’t make the top 5 of any list.

The Blavatnik researchers have put in a huge effort to compile an ongoing index of restrictions in all the countries of the world from January till now, counting such things as school closures, travel restrictions, bans on gatherings, and stay-at-home orders. They also look at economic and health policies, but I am here most interested in their findings on the stringency of government when it comes to the ability of people to socialise since that is the bit that most strongly affects wellbeing: enforced social distancing means enforced loneliness for many, leading to mental health issues, physical health issues, and the breakdown of social relations. It is a form of violence, particularly when kept up for months.

The table below show you the average Stringency index from March 1st to December 10th (or nearest to that date) for selected countries, adding the number of reported covid-deaths per million from the worldometer website on December 20th. It tells a few surprising stories.


Stringency C-death pm Stringency C-death pm
Estonia 38.0 136 United States 66.2 979
Finland 42.7 91 Nicaragua 15.3 24
Sweden 56.3 789 Bolivia 82.0 769
Denmark 54.4 181 Costa Rica 64.8 193
Norway 47.9 74 Honduras 88.5 303
Iceland 44.4 82
Italy 65.5 1139 Tanzania 26.6 0.3
UK 65.8 990 Burundi 13.5 0.2
Spain 65.5 1046 China 71.7 3
France 63.8 927 Libya 84.0 198
Germany 59.6 319
Switzerland 49.5 774 Japan 36.8 23
Belgium 60.1 1604 Cambodia 43.6 0
Belarus 16.2 141 S-Korea 52.5 14
Croatia 49.4 796 Taiwan 25.4 0.3


The first surprise is the three countries with the lowest stringency of government restrictions: Burundi, Nicaragua, and Belarus. Equally surprising are the three countries with the highest restrictions: Honduras, Libya and Bolivia. All six are basically dictatorships or something close to it. What is going on?

The story of Libya is perhaps the instructive one: it is consumed by a civil war between rival warlords vying for control over the oil trade. It has no clear government in charge of the country. Yet, what passes for the government at a certain moment during the first wave declared the highest level of restrictions anywhere in the world (100%, ie their restrictions define the maximum of the index). The likely purpose had nothing to do with the virus but everything to do with trying to make organisation difficult for the opposing forces: if your opponents are dumb enough to socially distance and stay indoors, one meanwhile can organise one’s own forces to isolate them and shoot them. Something similar was probably going on in Honduras, which is also close to civil war.

Burundi, Belarus, and Nicaragua are also dictatorial, and they have had the worlds’ lowest level of restrictions. It is not entirely clear why, but my best guess is that in their case the governments were more worried that restrictions would hamper their own movements and economic fortunes.

Then Europe. What might surprise the reader is that Sweden is the country in Scandinavia with the highest average government restrictions in this period. How is that possible when Sweden is the reluctant poster-child for sensible covid policies in the whole world?

The story is a simple one: the other Scandinavian countries, Denmark and Norway in particular, had more stringent covid policies at the start (March/April) but repented. In the case of Norway, the health authorities literally repented and apologized to the population for needless restrictions. Meanwhile, in Sweden, the wish among the politicians to be seen to do something and fall in line with international expectations has created conflict with the low-restriction opinions of their health authorities headed by Anders Tegnell. The end result is that Sweden has in fact had higher average restrictions than any of its neighbours, though still lower restrictions than any of the more populous countries in Europe, and (thus) better economic and wellbeing outcomes. What is true, but not shown in the table, is that Sweden’s restrictions have been more stable than elsewhere.

Otherwise, the restriction levels in Europe are all over the place, with a middle-of-the-road rating for Belgium that has the world’s highest claimed covid death rate, and the highest average restrictions for the UK (just my luck!).

The next clear pattern in the data, which the Blavatnik researchers have written on extensively, is that the more covid deaths in a country, the higher the restrictions on average. The causality is largely in that order: when there are more measured covid-deaths, restrictions almost invariably increase, partly due to public pressure and partly because the deaths give governments an excuse to expand their control. Policies have not followed the science (which warns against the path followed and has failed to find any clear evidence for the effect of restrictions), but the need to be seen to do something.

Finally, there is the interesting issue of low average restrictions in much of East Asia combined with very few deaths: Taiwan’s restrictions are on a par with those of Tanzania. South Korea and Japan have similarly been light-touch in their restrictions relative to Europe and the Americas, probably because that whole region might have high levels of prior immunity such that they don’t get large waves of deaths that then lead to calls for more restrictions. Only China seems to have insisted on both high levels of restrictions alongside low levels of deaths, though one should be very suspicious of the statistics on either.

This entry was posted in Coronavirus crisis, Death and taxes, Democracy, Politics - international, Science, Social Policy. Bookmark the permalink.

29 Responses to Which governments have been most restrictive?

  1. I am and will always be Not Trampis says:

    Australia and New Zealand are not on the list so it aint worth a squirt.

    Just a thought but what sort of lockdowns did Europe have if people could still travel.
    Any counties have plane travelers doing 14 days quarantine and how many countries had police etc on their borders?

  2. Dick Whittington says:

    Death rates and stringency measures should be correlated with demographics relating to age and obesity/diabetes and other factors predisposing to death from Cobid-19. Cf Israel, which has very high infection rates but relatively low death rates of which 90+% were among those over 70 or with pre-existing morbidity. Those statistics may reflect Israeli population/health demographics and/or selective precautionary regimes for vulnerable populations.

  3. Conrad says:

    1) I’m not really sure Taiwan can be used as a useful point here, as they basically got rid of it early and had a lot of quarantine requirements to stop further transmission like Aus. Alternatively, they were clearly better at doing quarantine properly than Aus.
    2) Rather than an overall measure, it might be smarter to have some measure including variance. Places like Korea have quite variable measures based on who is catching what, and Norway had quite stringent measures and then reduced them when they had almost no cases. France had this too, but they had start making up new levels when they got so many cases. I doubt any measure is going to be worthwhile unless it accounts for this, otherwise, as you note, a lot of what you a measuring is what the population wants done, not what might be effective like going in hard early and then relenting. Was the simple one-dimensional scale just used for public consumption?
    3) I don’t mind the idea that there might be different levels of resistance with groups, but it would be good if you could get some actual evidence for this other than just say some people in Asia have it because they don’t seem to spread it as much. It’s a bit like dark matter at the moment and doesn’t explain, why, for example, Malaysia is getting a lot of cases recently (and the Phillipines has had a lot too) — now 1000+ cases a day or not especially high rates of testing (although few people are dieing yet). Given their location and genetics, shouldn’t they be protected and spreading it less according to your theory? There are lots of other things going on in parts of Asia too. For example, in the first wave in HK, no medical workers got infected, unlike many places like Aus and in Euroland where they initially formed a large group and presumably one that spread it a lot.

    • paul frijters says:

      you should write your own posts on this sort of thing, Conrad. Why not?

      On the possibility of high prior immunity in East Asia, I am certainly not claiming this is known, only that I think it looks likely. There have been different studies in different countries, one saying (from memory) that 50% of Singaporeans had T-cell cross immunity, whilst in Europe it was more like 20-30% (see below a reference). I have not heard of new studies on the issue for months though, and certainly nothing in East Asia, which is actually a bit suspicious because these studies are not that hard to do and obviously important information. Let me clarify the text to make clearer that the ‘possibly’ applies to the whole argument on regional immunity.

      Per million individuals the numbers are tiny throughout East Asia. One wouldnt expect prior immunity to be exactly the same though: presuming the big cities in China are the epicenter of many prior variants, immunity would be higher the more linkages with those cities. You’d actually expect quite a bit of prior immunity in Australia given its strong trading ties to E-Asia (which might explain why outbreaks havent ‘caught fire’ to the degree they have elsewhere).

      Lipsitch, M., Grad, Y. H., Sette, A., & Crotty, S. (2020). Cross-reactive memory T cells and herd immunity to SARS-CoV-2. Nature Reviews Immunology, 20(11), 709-713.

    • ianl says:

      I’m not really sure Taiwan can be used as a useful point here

      I’m quite sure it can – and should. Since SARS-1, Taiwan had developed distinct protocols for contact tracing, international border controls, specific quarantining (not the thuggish VicGov type) and clear unambiguous information supplied to the public. They had shared these protocols with the rest of the world but were largely ignored, especially by Aus “experts”, and now by the MSM.

      Taiwan has a population of about 23 million within a land area that largely equates to an elongate shape bounded by Wollongong-Newcastle-Lithgow. Yet their protocols worked, to the point where their death rate attributed to C-19 is less than 1/10th of Aus, they did not smash their entrepenurial class up, and they had no lockdowns apart from 2 weeks of karaoke bar closures.

      Yet this obvious success is deliberately avoided from being discussed in Aus. Embarrassment ?

      • Embarrassment yes
        ,but a very narrow focus that the only things-ideas worth considering are those that emanate from Europe- America is deeply entrenched in our elite and academic culture it’s as though they can’t see anything else. Mind I feel that NSW is really doing much the same stuff as Taiwan it’s just that nobody is talking about it.

  4. Saupreiss says:

    Interesting topic.

    I agree with Conrad for the most part.

    I am not convinced that the good health outcomes in parts of Asia are due to high prior immunity in these countries (and certainly not in AU/NZ). (That does not mean that it might be a contributing factor.)

    Most of what I have read about Taiwan (by reasonably informed people) suggests strongly that the Taiwanse learned organizationally from previous virusses (in particular the 2003 SARS outbreak). A similar story seems to apply in Hongkong, Singapore, South Korea, and Japan.

    As to Straya, the Melbourne situation has demonstrated how quickly things can run out of control. The fact that the current Sydney situation might be contained (as seems currently likely), is really due to NSW Health being competent and Victoria Health (still) being a basket case.

    Methodologically, a key problem with the Blavatnik data is that it takes as sole criterion covid-19 related deaths when, as you point out correctly, we really shold be concerned a whole slew of other criteria such as various well-being consequences.

  5. I am and will always be Not Trampis says:

    enforced social distancing means enforced loneliness?
    how can this be taken seriously. you are 1.5 meters away from another human and you are lonely.
    Most people can have a conversation at that distance even share a meal at that distance.
    Paul becomes Kates

  6. I am and will always be Not Trampis says:

    you attempt to explain this rather silly sentence then.

  7. I am and will always be Not Trampis says:

    As I thought no-one can

  8. I am and will always be Not Trampis says:

    still no-one can

  9. Christopher Hood says:

    Months have passed.
    The assumptions behind the Frijters position have turned out to be unsound, all of them in ways that have undercut his contention that distancing is too expensive to be justified.

    Consequences of COVID:
    Remember 0.02% that Frijters insisted was the likely upper limit of death rate? Now we have several large countries with leading health systems where the deaths as a proportion of the whole population already exceed 0.1%…and that’s with infection that doesn’t reach or exceed 20% of the whole population, and with deaths continuing to rise.
    Remember death as something that happened quickly or not at all? Now we know that
    significant COVID deaths happen months later: the death tail is long.
    Remember how the ‘crude’ death rate comparing deaths to confirmed cases could be disregarded because undiagnosed cases were likely to be so large? Now we know that as testing has greatly increased and the proportion of undiagnosed cases in many countries keeps dropping the ‘crude’ death rate is holding up or even increasing.
    Remember death as the only significant health consequence of COVID? Now we know that continuing serious health impacts are widespread; 40% ongoing health impacts is likely.

    Consequences of restrictive measures:
    Remember how all consequences of COVID itself could be disregarded, as it was so relatively benign? Now we have multiple studies, in many areas of economic impact, showing that responses to outbreak apart from restrictions are large, and economic impacts substantial. Travel that drops by more than 80% apart from any imposed restrictions; employment collapses; other economic impacts.
    Now we have to disentangle the economic impact of restrictions from the economic impact of the virus. But Frijters won’t do so. His recent piece on WELBY impacts using two high-COVID countries allocated all economic impacts in both to restrictions.
    Remember the much higher suicides that would flow from restrictions? Despite higher use of psychiatric support services, suicides have not risen or have fallen in many jurisdictions.
    Remember the elective medical procedures that would be forgone forever? Many in Australia have already caught up, and the view that all will do so if control continues is too strong to be answered by mere hand-waving.

    Benefit of restrictions:
    Remember how delaying COVID impact didn’t really matter overall since it would spread through the whole population eventually whatever we do? Now we have experience of how effectively some countries have suppressed spread, and now we have vaccines, very much earlier than expected. Restrictions don’t just stop medical treatment from being overwhelmed, though that’s significant as we thought (and several USA states are close to being overwhelmed now, as Trump burns the joint down on his way out). They keep economic activity afloat, as country comparison now clearly demonstrates. They buy time for vaccine rollout.

    Discrediting WELBY:
    The WELBY is a great enhancement to cost-benefit analysis, weighing as it does many substantial consequences other than death. Failing to apply it to COVID is merely discrediting the project, rather than showing its application.

    Time to get real, and revisit the assumptions of many months gone in light of experience to date. We have plenty of continuing uncertainties; let’s now take account of things we know now, and give uncertainty only its remaining scope.

    • I see the straw man techniques are alive and well.

      Its not 0.02% but 0.2% that I estimated mid march 2020 to be the likely number of fatalities if the whole population got exposed. Let me simply quote what I said March 20th: “So my own best-guess for the total fatality rate for the coronavirus should the whole population get low-level exposure is 0.2%.” You are hence making a streightforward mistake on a key statistic you attribute to me.

      Hasn’t that prediction turned out amazingly accurate? The Ionanidis survey on the WHO site puts the IFR around 0.23% for the world, so a bit higher here, a bit lower there. At the time of speaking, no country in the world has claimed above 0.2% and its now clear that at the world level 0.2% is far too pessimistic, though for Europe and the Americas as a whole it looks about spot on.

      The collateral damage I was predicting in March 2020, including mental health, economic damage, etc., has also been spot on. I predicted a wellbeing drop of 0.5 for the UK, which turned out to be 0.7. I predicted a 50 trillion drop in the world economy cumulatively the next 10 years. That’s still what the IMF projections roughly point to, perhaps a shade lower depending on the speed of the bounce back.

      As to unimportance of policies for covid outcomes, that was both the explicit assumption of most of the models used early on for the ‘flatten the curve’ arguments, and has been the exact theme of many recent articles failing to find effects of policies on total numbers of fatalities from covid, though finding strong negative effects on other outcomes. The experience with multiple waves in Europe and the Americas is also exactly what the standard immunology expected to see: suppression, if possible at all, is merely postponement.

      As to the alternative scenario that I sketched in Mach 2020 (ie, to rely on personal responsibility), it is now clear that that was both the blue print scenario of the WHO and national agencies before the panic, and is now the essential position of over 50 thousand scientists in the Great Barrington Declaration. So not a fringe opinion but actually the dominant position before the panic and still the dominant position by a fair slice of scientists. Within Europe and the Americas the regions and countries that followed that advice better did much better on average with other outcomes (economic, mental health, freedom).

      Etc. You are simply displaying the narrative of your obedience. I know its unrealistic of me to expect gratitude and some amazement at how much you have been misled, but you could at least try to open your mind a tiny bit.

      • Saupreiss says:

        Plus, if you are so sloppy as to confuse (or falsely recall) 0.02% and 0.2%, which seems to be the key argument/statistic, then maybe you should just shut up.

    • Truth is there are nations such as Belgium that followed the official line had extensive severe mandated lockdowns etc and had really bad outcomes ,both health and economic and then there are nations such as Japan that did not follow the official line and have had pretty good outcomes .
      The proposition that severe mandated restrictions actually produced benefits is absurd.

      • I am and will always be Not Trampis says:

        and which country had a porous border and which one did not?

        • The virus is definitely in Japan yet it has not run rampant despite the fact that they did not follow the official playbook.

          • Saupreiss says:

            Japan is one of the countries that curiously is rarely mentioned by the lockdown fanatics, as are Taiwan, South Korea, Singapore, Vietnam, and Iceland.

            Seems not to fit their narrative.

            Meanwhile, in Australia a few cases have led again to massive restrictions … and lockdowns of sorts on the Northern Beaches. Proportionality has been shot for good.

            • murph the surf says:

              Greater Sydney should lock down hard now.
              The state government is not listening to scientific advice but fudging its decisions to better appease political posturing.
              The populations of East asian countries are entirely capable of locking themselves down in answer to why they are rarely mentioned in discussions.During Sars 1 in HK and Taiwan this was how people behaved without specific government direction and the high population densities enforce face mask wearing every winter to cope with normal less lethal viral infections.
              At the beginning of Sars1 it was not known what was causing the disease for several months but populations locked down , shops closed , landlords dropped rentals.Expat fled.
              How do I know this – like Conrad I lived in HK at that time , after a couple of years in Tokyo.

              • It’s plain that bottom up community grounded and supported responses work best and top down mandated measures can be worse than useless .

                • I am and will always be Not Trampis says:

                  one could go further,
                  Unless you have a lockdown that is entirely similar to Australia’s it won’t work.
                  Allowing people to fly hither and thither as occurred in Europe is stupid. allowing people to enter your country and not go into quarantine is equally so

  10. I am and will always be Not Trampis says:

    Alas Paul is Katesying yet again.
    I actually put casualties in Australia from herd Immunity at 175,000 which is 1% however this was poo pooed without giving a reason of course.
    The fact people who die would be much higher because of the pressure on the hospital system and the people manning it ,
    Unless of course you legislate people suffering from the virus cannot be treated in a hospital.
    Fritjers is naturally silent on this.
    what about other suffering from the viruas. The MJofA recently had an article on this. There have been others. Fritjers ignores this.
    how about the legal implications? without any change in the laws here implementing a herd immunity policy would simply involve an informal lockdown. People would work from home. adopt social distancing etc because the specialists would advocate this.
    Fritjers has never addressed this key issue. Complete and utter poor scholarship.
    how about implementing a policy few specialists support?
    the economic consequences of not having a lockdown have been shown by people such as Quiggin, Wren-Lewis or Cowen to be poor. In our case worse because of little government support.
    Lockdowns are finite and one of the major reasons for implementing them is to ease the pressure on the health system. Fritjers never acknowledges this at all
    Imagine the pressure of herd immunity on the system. Look at the USA and then increase it by a factor of at least ten. That would be pressure on the health system. Now imagine you are the person who decides who goes into the ICU and who does not and therefore dies. it would be intolerable. Fritjers uses trumpism to say black is white

    Fritjers never wants to talk about suicides not rising as predicted. Indeed Ms Foster stil has not x come to term with this and is still saying suicides have risen.

    Nor can he come to terms with his thesis being one large wave which lasts perhaps one year probably two at a minimum. Just how mental welbeing would be during and after that is highly problematic but one thing is for sure. It would be very negative.

    Lastly we have yet to be told what is involved in those who are vulnerable to the virus in terms of death. The elderly who mainly live their own premise ( whether owned or rented ) not aged care homes nor retirement villages. what about those with disabilities, aboriginals?? Again complete silence.

    Of course we then come to possibly the silliest statement made af gain with reasons or evidence that enforced social distancing enforced loneliness.

    Funny enough in our national lockdown we had increased community consciousness where I lived. For the first time in yonks we saw our neighbours and talked to them.
    It was easy to get out through exercise or going to to supermarket or the fruit and veg

    Hard to have a debate when neither Fritjers nor any of the acolytes here can answer very simple questions!

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.