I need to learn Lasso regression.
As I’ve pointed out, the effectiveness of an intervention (regardless of the method used to measure “effectiveness”) is not the only consideration. You could achieve a dramatic decrease in R by killing 90% of the population. That in itself doesn’t mean it’s a sensible policy. Costs matter.
I’ve also pointed out that reduction in R is not, in and of itself, a useful goal. R simply describes the rate at which people become infected, and it is incredibly hard to measure; so apart from anything else, scoring any intervention in terms of R is going to introduce a lot of uncertainty. If your goal is to reduce the load on the health service, then an entirely equivalent intervention would be to increase health-service capability (which is on the list of NPIs). If your goal is to reduce the ultimate death toll, then reducing R doesn’t necessarily make any difference.
You may disagree, but the aim of any intervention should be to reduce overall harm - or, perhaps, to do the greatest good for the least overall harm. Very few papers have addressed this tradeoff; they tend to focus either on harms, or on benefits exclusively.
As for the chart, their assessment of ‘effectiveness’ doesn’t make much sense to me. The blobbogram clearly shows lockdowns having a less impressive effect than (say) crowd cancellation, with the right-hand end of the line at zero; the regression measures don’t seem to justify its position at the top, either. And again, considering the relative harms imposed, it seems to me that crowd cancellation wins hands-down. IIRC, crowd cancellation was the one of the primary interventions (to the extent that they had any at all) used in Sweden.
What’s your point here?
The point was that if nothing can be done about infection then we would see similar outcomes throughout the world regardless of policy, but the example above of UK vs. New Zealand show otherwise.
London alone has 2x the entire population of NZ. UK had 10x as many annual tourist arrivals as NZ before the pandemic.
Comparing NZ to TW, the population of NZ is under 5 million and has a population density of 18 per sq. km. TW’s population is 23 million and it has a population density of over 600 per sq. km.
What can be done successfully in NZ isn’t necessarily applicable to TW. Basically TW was NZ as long as it managed to keep the virus out. That’s not the case anymore.
Interestingly, since the beginning of the pandemic, NZ has conducted over 1.8 million COVID tests. TW has conducted less than 500,000. 
1 Like
Yes. That is precisely how it panned out. Policy did not correlate with outcome.
If there were some magic policy bullet that explained the variations between countries, then you can be sure someone would have identified it (and probably been nominated for a Nobel by doing so). But nobody has.
Well, they show that the UK and NZ are different countries. I’m still not sure why comparing these countries, specifically, is relevant to Taiwan’s predicament.
1 Like
Singapore and HK which are even more densely populated also had outbreaks and successfully fought them. Your logic is just not true
What’s not true? You compared the difference in the numbers of the UK and NZ and made some assumptions, unproven, about why the pandemic developed differently in each. It’s not unreasonable to see that NZ is far less populated, far less dense and far less transited by tourists and conclude that these might be among the reasons why.
As for SG and HK, look at the testing numbers. SG has done over 11 million tests. HK has done over 16 million tests. SG’s population is under 6 million and HK’s is around 7 million.
TW, with a population of 23 million, has done less than 770,000 tests (!!!).
To me it’s crazy to think that TW can lock people (many of whom live in multi-person multi-generational households) in their homes for a few weeks and the virus will magically die out. If you want to look at the countries that have managed to get outbreaks under control, it’s crazy to ignore the fact that all of them have done big time testing.
Edit: corrected the number of tests TW has done.
1 Like
Good, so what do we learn from this: The key to success for Taiwan to get things back under control is to implement further restrictions and hugely expand testing. I agree with this 
You seem to be having a conversation only with yourself.
Well sadly I don’t see any indication TW is going to expand the testing significantly.
From other threads, I understand that TW doesn’t even let non-doctors perform the tests and it is threatening big fines for anyone buying or selling self-testing kits.
Testing was a controversial issue for the past year. The CECC always said that there was no need and it would be a waste of resources. Fair enough. There was always the implied suggestion that large scale testing could be done if it ever became necessary but now it looks like the emperor has no clothes.
If other countries could ramp up their testing big time in the beginning of the pandemic why can’t TW after more than a year?
2 Likes
Well, there is hope on the way. I think I read somewhere that today the CDC allowed or is just about to allow big companies to do rapid testing
EDIT: Yes, indeed…
Step in the right direction but still not enough.
The rapid tests are useful but you cannot rely on them in all situations. They are most appropriate for screening symptomatic people, who shouldn’t be coming to work anyway.
Their sensitivity in asymptomatic people or people who are in the early stages of infection is low enough that false negatives are a real problem.
What I can see happening is they let employers use rapid tests as a way to appease the laoban who don’t like WFH. People who might be asymptomatic or in the early stages of infection will test negative, but everyone in the office will have a false sense of security that they don’t have the virus and they’ll behave accordingly (no social distancing, reduced mask wearing).
What TW needs is the ability for people who have symptoms or exposures to be tested and receive their results in a timely manner so that they can isolate and inform their contacts. This will also help TW identify hot zones more quickly so that they can take appropriate action.
If that gets mandated how many companies will turn the tables and just say sleep at work haha
The point was the outcomes were vastly different, so it is evidently not true that nothing can be done about infection. A better comparison would have been Sweden against New Zealand, similar population density but completely different end result (but Sweden is not an island…).
But is there something that can be done here to make a different outcome? The current policy appears to have flattened the growth, but assume it stays like this we’ll just have to wait months until we get the vulnerable vaccinated (2022?), the offices will keep going, but the restaurant and tourist sector will be hammered, what happens when fatigue sets in? Would a stricter lockdown work? It undoubtedly would slow the virus if offices were closed, but you would still need at least a month for it to burn out, and what about the economy? The other option is to just let it run wild and ‘live with it’ in which case a lot of people will be saying early goodbyes to their elderly relatives.
There’s been a whole spectrum of outcomes. Of course in some cases geography can be a factor, but why at the beginning of 2021 did Taiwan only have 10 or so deaths and UK has 100k, if not related to the difference in policy?
I happen to believe that this is not the case. The test backlog is too big. The hospitalization/severe cases rates are too high. I believe they are only detecting the tip of the ice berg at this point and that the number of cases is much higher.
1 Like
Because TW managed to keep the virus out and the UK didn’t. As you’re seeing now, once the virus gets into the population and you get to the point of community transmission, it spreads and people die.
Remember in early 2020, Taiwan didn’t have any lockdowns. It was the border control and luck.
But do you think the policy has made a difference to the growth? We don’t know the dynamics through April and early May as we only see the cases pile up when they started testing outside the original cluster, but it looks like there is a flattening in the growth, and the positivity is holding steady.
Sure but is it a meaningful difference? People are still going to work. Even though we’re not at Level 4, many people have self-imposed a Level 4 like lockdown on themselves.
But what does this mean? Right now you have multi-person households hunkering down together and we know that people spending extended amounts of time with other people indoors is where the virus spreads the most efficiently.
The positivity rate is meaningless at these testing levels and with the backlog IMO.
1 Like