1/ Failures of Pandemic Central Planning (Magness)

"Public health is acutely susceptible to knowledge problems, which foster the conditions for a public choice trap that causes proposed policy measures to become ineffectual or even counterproductive."

2/ "The criticisms of Snow – that he bucked scientific consensus, that his water-borne transmission theory contradicted a self-evident effect of widely observed odors, that he threatened to undermine the Health Board’s solutions to cholera – parallels the 2020 pandemic response."
3/ "The governments of both the U.S. and U.K. effectively endorsed the catastrophic forecasts of the ICL model along with its prescriptive approach of forestalling massive casualties by enacting a series of lockdown-style NPIs intended to control the replication of the virus."
4/ "It predicted catastrophe in the absence of NPIs for every country.

"The ICL team didn't paused to examine whether its prescriptive NPI measures would perform in reality as their models maintained. The measures' effectiveness was implicitly presumed by the model itself.
5/ "Prior to Covid-19, a substantial body of epidemiological literature warned about the dangers of relying on theoretical modeling as a basis for respiratory pandemic mitigation.

2006: “Would closures decrease contacts between people and so retard the spread of the epidemic?
6/ "Or would those affected spend more time in malls, restaurants, and other settings that might result in more contacts and more rapid spread of influenza? No model, no matter how accurate its assumptions, can predict the secondary and tertiary effects of mitigation measures.”
7/ "Interventions are highly susceptible to political distortions that shift their purposes away from public health or even to counterproductive directions.

"Ferguson’s chosen assumptions for NPI-induced contact rate changes display no evidence of their own empirical validity."
8/ "By following a policy built around the prescriptive focus of the ICL model’s scenarios (which intentionally omitted care homes and prisons), government bodies almost certainly made the outbreak in nursing homes significantly worse."
9/ "The unmitigated spread scenario, which dominated the headlines, overstated mortality by between 137% in hard-hit Mexico to 1,300,000% in Taiwan, which largely avoided a major outbreak for the first year even though it did not employ lockdowns or most NPIs during this period."
10/ "ICL’s projections severely overestimated mortality irrespective of the policy interventions taken. Even the unrealistic “suppression” scenario (heavy-handed lockdown for over a year until full vaccination) resulted in a median overestimate of 535% vs. what really happened."
11/ "Flaxman's study implicitly assumes the validity of Imperial’s own mortality projections. These are then treated as a “counterfactual” to compare to observed data.

"Similar claims of unwarranted causality appears in other studies testing the effectiveness of lockdowns/NPIs."
12/ "Of the 39 papers considered, 26 claimed to show a causal relationship between lockdowns and pandemic control, yet the majority of those papers (65%) lacked an appropriate causal inference technique to support such conclusions."
13/ "Pro-lockdown studies that lacked appropriate causal inference testing (including Flaxman) tended to be cited in epidemiology journals at a much higher rate than those with more suitable statistical techniques."
14/ "Deeply incomplete information creates a classic knowledge problem, but the absence of knowledge stokes precautionary alarmism and the attempted implementation of a central plan, which would require the very same missing knowledge to execute even under ideal conditions."
15/ "In pandemic central planning, the pronounced knowledge problem itself creates the opening for a public choice trap.

"In the absence of reliable information with which to fine-tune the model, decisions on how to proceed advance on almost exclusively political terms."
16/ "Though Bowman attributes the apex and decline of Britain’s death numbers to lockdowns kicking in and working, the successive waves followed a nearly identical pattern without the implementation of these policies in Sweden."
17/ "The case surges predicted by Fauci and the modelers in Florida, Georgia, Texas, and similar open locales simply failed to materialize. When cases did rise in these states, they appear to have followed regional and seasonal patterns, unconnected to any lockdown or other NPI."
18/ More reading:

Three Myths about Federal Regulation

First Literature Review: Lockdowns Only Had a Small Effect on COVID-19 (Herby)

Covid Lockdown Cost/Benefits: A Critical Assessment of the Literature

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with Darren 🙏

Darren 🙏 Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @ReformedTrader

25 Nov
1/ BST Holdings, LLC. v. OSHA (United States Court of Appeals, Fifth Circuit; Nov 12, 2021)

"This case concerns OSHA’s most recent ETS requiring employees of covered employers to undergo COVID-19 vaccination or take weekly COVID-19 tests & wear a mask."

2/ NOTE: The tone and degree of expressed certainty in this document runs counter to most of the academic papers I read (one reason I stay away from material sourced from politics and law).

As always, when policy makers are concerned, take everything with an extra grain of salt.
3/ "In its fifty-year history, OSHA has issued just ten emergency temporary standards (ETSs). Six were challenged in court; only one survived.

"We reaffirm our initial stay.
Read 13 tweets
25 Nov
1/ When face masks signal social identity: Explaining the deep face-mask divide during the COVID-19 pandemic

"We found strong evidence of in-group favoritism among both mask & non-mask wearers when information about the partner’s mask usage was known."

2/ "Instead of simply asking what mask & non-mask wearers think of each other, we look at how information about others’ mask-wearing behavior influences cooperation when real money is at stake.

"We recruited an even distribution of subjects from different political affiliations.
3/ "Our experiment consisted of two main stages of incentivized Prisoner’s Dilemma (PD) games with & without subjects knowing their partners’ mask usage.

"Each subject’s answer to a hypothetical scenario was used to classify her as a mask wearer (MW) or a non-mask wearer (NMW).
Read 18 tweets
25 Nov
1/ Effectiveness of Adding a Mask Recommendation to Other Public Health Measures to Prevent SARS-CoV-2 Infection in Danish Mask Wearers: A Randomized Controlled Trial

"95% confidence intervals are compatible w/ 46% reduction to 23% increase in infection."
2/ "During the study period (3 April to 2 June 2020), Danish authorities did not recommend masking. Use was uncommon (<5%) outside hospitals.

"Participants were randomly assigned to mask/control groups.

"Those in the mask group received 50 3-layer, disposable surgical masks."
3/ "Participants received follow-up surveys to collect information on antibody test results, adherence to recommendations on time spent outside the home, development of symptoms, COVID-19 diagnosis based on PCR testing done in public hospitals, and known COVID-19 exposures."
Read 8 tweets
24 Nov
1/ Institutional Investors and Stock Return Anomalies (Edelen, Ince, Kadlec)

"At the one-year horizon, institutions have a strong tendency to buy overvalued stocks (anomalies' short legs) and which have particularly negative ex-post abnormal returns."

2/ SIH = sophisticated institutions hypothesis (agents trading in a way that exploits anomaly return predictability)

"We assign anomalies to seven broad categories, choosing the one from each category with the most reliable alpha (highest t-statistic) during our sample period."
3/ "Because our central hypothesis centers around how institutional investors modify their portfolios as stocks take on anomaly characteristics, we examine changes in holdings rather than levels. We focus on the number of institutions holding a stock rather than % shares held."
Read 18 tweets
22 Nov
1/ Combining Value and Momentum (Fisher, Shah, Titman)

"Simultaneously incorporating both value and momentum reduces transaction costs and better utilizes the information from unfavorably-ranked value and momentum stocks in a long-only portfolio."

papers.ssrn.com/sol3/papers.cf… Image
2/ Ranks are based on percentile of aggregate market cap that a stock is above for each factor.

Turnover = total amount bought or sold each month / portfolio size

Bid/ask spreads have fallen dramatically in the past few decades, so two estimates of transaction costs are used. ImageImageImageImage
3/ "Even though the percentage of aggregate market cap is similar for the small value and momentum portfolios, the small value portfolio tends to hold close to twice as many securities as the small momentum portfolio due to holding stocks with small market capitalizations." Image
Read 13 tweets
21 Nov
1/ Global Assessment of the Impact of Masking on COVID-19: A Country Level Comparative and Retrospective Analyses Using the Richards Model (Mamun, Vejerano)

"Laboratory and modeling studies did not translate to a measurable difference in the real world."

2/ "We are unaware of studies assessing masks’ effectiveness using real cases from multiple countries.

"The filtration efficiency of ordinary surgical (~12%) or cloth-based masks (~10%) is inferior to high-efficiency masks (≥95%). Aerosol leakage is also higher due to poor fit.
3/ "The Richards model can reasonably estimate the time when a phase shift occurs during an outbreak (turning point, ti), the infection rate (r), the exponential nature of the epidemic (α).

"It applies best for a curve consisting of a single high peak and a turning point."
Read 17 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Thank you for your support!

Follow Us on Twitter!