Theory of change, depression drugs, fossil fuel divestment

In much of science practice, scientists form ideas and models and then test these models in the world. The more credible the model to other thinkers and the more the model can explain empirical results and - even better - predict new results that be tested then the more useful the model and the more weight we can give it in our thinking.


Many models can not fully explain what we observe in the world. 

Many observations have no good models for them.


How should we think about decision making faced with forward looking uncertainties? I argue we should put more weight on when we have plausible models. This is harder in social sciences than physical sciences but still should be considered. That said empirical data can guide where models seem inadequate.


I’m going to think about these cases and what they might say:


  • -Anti-depression drugs and their effectiveness or lack

  • -Germ and virus theory and mask wearing

  • -Fossil Fuel divestment , social political change strategies

  • -Quantitative Easing by central banks

  • -ThenDoBetter grants - catalyzing change


Anti-depression drugs and their effectiveness or lack of.

Prozac, known as fluoxetine, is a serotonin re-uptake inhibitor or SSRI and it was designed as an anti-depression drug based on a model for depression that scientists developed.


The basic model is called the monoamine hypothesis of depression and this theory s proposes that patients with depression have depleted concentrations of serotonin, norepinephrine (noradrelanline), and dopamine. 


It was conceived due to at least one line of evidence on the effects of reserpine on serotonin and catecholamines. Reserpine, an alkaloid extracted from the Rauwolfia serpentina, was utilized as a treatment for hypertensive vascular disease in the 1950s; however, reserpine was found to precipitate depression in some patients. The depression produced by reserpine was reversed after the treatment was terminated and following either rest or electric shock therapy (Additionally, reserpine was found to produce depressive-like effects in animals Reserpine was found to inhibit vesicular monoamine transporter, and as a result, depletes brain monoamines (i.e. serotonin and catecholamines), which provided evidence for the role of serotonin, norepinephrine, and dopamine in depression.  


Since its original formulation, scientists can see that this hypothesis can not explain many other observations (for instance that healthy patients who deplete monoamines do not become depressed).

Still, SSRIs were invented  as the NHS website suggests:

“ It's thought that SSRIs work by increasing serotonin levels in the brain.


Serotonin is a neurotransmitter (a messenger chemical that carries signals between nerve cells in the brain). It's thought to have a good influence on mood, emotion and sleep.

After carrying a message, serotonin is usually reabsorbed by the nerve cells (known as "reuptake"). SSRIs work by blocking ("inhibiting") reuptake, meaning more serotonin is available to pass further messages between nearby nerve cells.”

Without going into too much more detail, we know empirically that SSRIs relieve depression in a good number of people and can help prevent relapse. But we also know that they don’t work in  a good number of people and that they stop working for a good number of people. And we are not exactly certain why.


There is much we simply do not understand about depression and brain function.


So how does relate to masks?


We have a pretty good theory about how viruses are transmitted. They travel in aerosol droplets from the nose and mouth - from sneezing and coughing and breathing - and they are then breathed in by others. They can also go from nose to hand to someone else’s hand to nose, or from hand to object back to someone else’s hand. Although there is debate as to how long viruses can survive on objects and how easily this transmission occurs.


Scientists have a consensus on this. And so when thinking about mask use to prevent transmission it’s surprising with hindsight that this did not feature more prominently.


Of course now we can look back and also note how many Asian countries had mask wearing and how European and America seemed initially reluctant. Still, I think it would have been much better if those in policy making or decision making roles could have case back to whether there was an underlying model for whether mask wearing should work and then assess risk/benefit.


Here the model was and is very useful, but didn’t seemed to make an impact.


Central Banks have engaged in what economists call quantitative easing or QE. These are large scale purchases of asserts by central banks. Economists are debating as to how QE actually works in the real economy and they are unsure. At least I can say scientists are much more sure on how germ theory works than on how quantitative easing works.


It’s perhaps in the same area of dispute as depression drugs. There’s some empirical evidence but the totality of it can not be explained by one model we have. And the three models posed by QE are not universally agreed upon.  (Segmented Market, Preferred Habitat and Signalling theory).


There’s - to my sense - maybe more agreement amongst depression scientists about what can or can not be explained then there is amongst economists. I note this as important as the Bank of England publishes an in-depth report into QE - https://www.bankofengland.co.uk/independent-evaluation-office/ieo-report-january-2021/ieo-evaluation-of-the-bank-of-englands-approach-to-quantitative-easing


This brings me to fossil fuel divestment strategies and thoughts on activism.


In biology, scientists call the SSRI mechanism as a mechanism of action. It’s how we best think the SSRIs are working biologically.


In social science, scientists often talk about a “theory of change”. 


“A theory of change is a description of why a particular way of working will be effective, showing how change happens in the short, medium and long term to achieve the intended impact.”


Supposedly a good theory of change needs to be testable. THis is in common with good theories of biological mechanisms of action - but are often much harder to assess in social science.


To my mind - a purported biological mechanism of action - such as the SSRI one - has much in common with a theory of change mechanism.


I come across some activists who have not really thought about their theory of change. And if the conversation allows, I will suggest they do think about it. I will often stress I do not know myself what theory of change is correct because many of the social science ones can not be supported either way by the evidence.


Still - when you think about the theory of change around divesting as opposed to the theory of change around engagement and persuasion of a company - you end up with very different theories.


This was summarised in Ellen Quigley et al’s report for Cambridge University (H/T Dominic Burke).    


The divestment social-political theory of change:


“The political case for divestment rests on expectations that it will accelerate the pace of legislation in favour of an energy transition away from fossil fuels. It does so both through creating a political

environment more favourable to legislation and by weakening the political power of the fossil fuel industry.”


This to use a policy term “moves the Overton window” - this changes the range of policies acceptable to the general public to be enacted by governments.


It’s quite far away from anything involved with investing (Although there are separate arguments for financial risks and theories such as influencing cost of capital).


One can argue this makes the movement more of a moral movement - one that you can align with women rights, minority rights, and slavery abolition and mperhaps more recently apartheid campaigns.


Using aparthied in South Africa is a complex and interesting parallel. Many factors (some with parallels to engagement and diplomacy and some with parallels to embargoes and divestments) are impossible to disentangle with cause and effect in the outcomes.


As strict financial theorists would argue most fossil fuel financing is not made by trading shares but by primary financial capital raising from bank or government loans or share issuance, and such that trading in secondary equities.


Thus the theory of change for engagement or within system workers is to persuade companies to change models and change strategies and this is more effectively done via share votes and engagement. And it is to invest in primary innovation and primary capital formation for companies making the most positive impact.


Neither theory can easily be proven with the empirical evidence we have. There are some companies that have changed course. Some climate policies have come into place and some evidence that the overton window has moved - although more easily for innovation policy than for carbon taxes, it seems.


But, it does bring me back to which theory of change is more meaningful or more reasonable to you.


I am asked where do I stand. If there’s time I typically outline a number of these arguments for both sides and because it is complex and unproven, I hold the theories of change lightly. 


But, my theory of change is that it falls - by a complex dance and a good deal of luck and happenstance - that a number of individuals spark the change that leads to systems altering. These individuals lead others.


And so, this is one of the reasons behind my idea on ThenDoBetter grants. The focus is on individual grant giving with a huge dose of luck and happenstance to people who will make that positive change.


As a coda thought. There are no universal theories of autism. The three main ones (central coherence, theory of mind, executive function) can explain certain aspects of autism but fail in many other aspects.  We have (in my view) poor models of autism (although not zero models, thus we can reject “refrigeration” and other catastrophically bad theories) and so we should hold lightly too much strong form advice here.


Quigley report:

https://www.cam.ac.uk/sites/www.cam.ac.uk/files/sm6_divestment_report.pdf 

How effective are anti-depressants:

https://www.ncbi.nlm.nih.gov/books/NBK361016/#:~:text=Studies%20involving%20adults%20have%20shown,within%20one%20to%20two%20years.

Brief history of anti-depressants: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4428540/


Theories on QE

https://www.stlouisfed.org/on-the-economy/2017/november/economic-theory-quantitative-easing


Quick links: climate, circular economy, grief, china, spaced repetition

Quick Links:

Where we are on climate. Wallace-Wells (a former?) climate alarmist (now just not as extreme) looks at the evidence that the central case is for 3 deg warming by 2100 (down from 4c but not 2c).

Circular economy: this business recycles chopsticks into furniture. How many more circular economy businesses are out there waiting? 

Interstitial timeson the difficulties of endings
Mark Carney on the purpose of markets, financial crisis, COVID and climate.
Ethical Progress, better today than before
Ray Dalio on grief 
Jeremy Grantham on late stage bubbles
Dan Wang on China
Using spaced repetition prompts for recall
FiveThirtyEight on US responses to protests.


Was 2020 a turning point year?

Why do I think 2020 may be a rare turning point year?
I still think there’s a decent chance that COVID will fade from memory and we will have learnt little (this would make it like Swine flu). But, I think that chance sits around 30% and moving lower. This is partly due to the length of time COVID will be affecting us and partly due to our response both innovation-wise and public health wise.

On the positive front we have had many science innovations not least on the vaccine and biomedical front:  mRNA technology looks set to prove long-term robust to make many kinds of vaccine. We have a malaria vaccine in late stage testing, Deepmind/AI has made advances in protein folding modelling, and new molecular entity drug approvals (excluding vaccines) was c. 51 this year in the US, which is in line with the last few years in terms of therapeutic innovations. With gene-editing technology and our increasing knowledge and comptuting power, biomedical advances for the next 10-20 years look promising to me.

Environment-wise: We’ve had China, Japan, South Korea commited to carbon net zero. Battery technology has continued to improve. Solar power is the cheapest form of energy in many places. Even Nuclear (mini) and fusion technology has continued to improve. Apple has joined the electric vehicle / driverless car race. 

Governance-wise: We had fiercely contested US elections that have essentially been peaceful and robustly managed given that over 161 million Americans voted. UK and EU managed to agree a Brexit trade deal. 

My guess is that certain people will be inspired by science and innovation as having some answers to our challenges that will make them place more bets here and invent more valuable things that will improve human welfare and the environment. That COVID has triggered an enhanced ability to work out of the office should help bring more productivity and people to work and develop and, hopefully, this should also bring about better welfare.

Many of these improvements are slow-moving - like our overall improvements in human life expectancy and welfare. Many of us both misjudge how far we have come, and perhaps if we understand our progress we misjudge the challenges which are still great.

But we will need both parts. To understand where we have made progress, where we still have challenges and to use the opportunities COVID has given us to do better while trying to defeat its catastrophic impact.

That's not to downplay the awfulness of COVID. That's with us. But how we react is still up to us.

I remain more worried about creative arts practitioners.While over the long-term creative industries have typically bounced back from hard times, I think 2021 will continue to be hard and I see many brilliant creatives having to leave the arts and related work. It’s hard to measure the value of arts and the financial rewards are low for the majority. There is little joy in a future generation of creative work when this generation is so hit.

COVID, why so many are mostly wrong, or only a little correct.

Summary: Vaccines are likely to give protection for at least c. 12 months and likely to reduce transmission rates, but vaccine hesitancy, mutation and maybe some amount of re-infection will mean that the virus stays with us permanently like influenza does. However like ‘flu we will find this disease manageable. We may also never know for sure why certain groups (eg men) suffer higher mortality. 

The medium to long term speculative thoughts is that this crisis will spur more innovation and creativity across several domains.

This is because many may conclude it is human innovation that has saved us and will save us. Similar thinking may be applied to climate challenges (I expect Bill Gates will double down on this in his next book). I also think - while with much pain- the creative arts will also react with more creativity, although extremely crimped near term, as people will have to find new ways of reaching audiences/consumers.


This is a long form read over why so many people are fairly wrong (or only a little correct) about COVID and why the information seems so confusing. I will attempt to touch on:

  • Predicting vaccines

  • Immunity and immune memory

  • Cross-protection

  • Different strains

  • Different genetics

  • Super-spreaders

  • Cultural differences

  • Data reporting differences

  • Complexity models

  • Re-infection

  • Narrow vs broad thinking (fox vs hedgehog)

  • Ideology

Back in August 2020, I made the point estimate judgement of an 80% chance of a vaccine by the end of 2020. Significantly above some observers estimates (although a good number of healthcare investors were making similar judgments).  I noted some of my thinking in my August blog.

What’s useful to note is why many expert observers were more pessimistic. I can summarise that those group were focused on past experiences, focused on the risks (which were clear) and anchored on previous examples. They were not willing to place faith in mRNA technology that had not produced commercial vaccine before even if much of the theory is well established.

Source: Google Finance

Source: Google Finance


Stock market prices embody future expectations that people with money (not reputation or press articles) buy and sell at. It’s very difficult typically to be ahead of this collective wisdom of the crowd. Still with in a stock price reveals a signal that can be interpreted.

If you look at Moderna’s (one of the vaccine makers) stock price - which embody many factors including politics, interests rates, etc - there was much of a run up from March to early November before the positive pivotal data in November. There are still future unknown events to come eg launch and distribution, but looking back one can suggest that investors with money were not super surprised by early November as much had already been “priced in” over March to October.


Mostly investors do not bet directly on a question such as “will there be a COVID vaccine in 2020?” But indirectly on stocks or other assets and prices which lead to money win/loss outcomes. These investors were suggesting through the Moderna stock price signal that there was a decent expectation of some success here.


I won’t rehash all the many science and socio-political points that went into my August forecast but suffice to say there are a number of people who do make and essentially bet behind these predictions.

Cross-immunity, herd-immunity, re-infection, strains, genetics and why everyone is only a little bit right.


Mostly - with rare exceptions - media articles will take a single look at a narrow domain question and present evidence in favour of a certain answer. Sometimes coloured by an ideology. (Even studies tend to look at a narrow question).


For example, if by ideas, you strongly favour individual choices you may balk at the idea of government imposed lockdowns and so you are drawn to articles suggesting Sweden or a “herd immunity” process as a way of proceeding without lockdowns. The actual data from Sweden does not matter too much - especially when you can find media articles to support your inclination.

Another example is re-infection. There are cases of re-infection, but it seems from what we know re-infection is rare but it can and does make article headlines.

[A distant simplistic parallel that people might understand is that you can get chickenpox twice (or rather, shingles after chickenpox) but it is rare.]


Still depending if you have an idea already about what we should be doing then a case of re-infection or an article about it can be used to support that view.

So you can put all of these statements together which have a little bit of truth to them.

  • There are asymptomatic carriers of COVID.

  • You can gain (some amount of) cross-protection for some (unknown) amount of time by exposure to other coronaviruses including the common cold. 

  • This level of protection will vary with strain, genetics, immune responses and memory - which in turn vary with factors such as age.

  • Different strains can act with different people’s genetics to cause varying levels of severity of disease.

  • Different people’s immune system will “remember” the virus differently (age, strain etc. variant)


All of this becomes confusing because we would like a simple answer of do I get cross-protection or not? Not the complex answer of it dependant strain, time and genetics (and perhaps environment)  and will not be static.

And from some of these simple parameters that can change we can have events such as “super-spreaders” where one person or one event (eg a sports or a night club evening) seem to cause many infections. The interplay of all those infection factors can produce those results. Or not.

In that sense - a distant parallel is with weather forecasting.  We can put together large trends to fairly accurate assess total infection cases in regions over  a few weeks or days, but predictions at the single person or event level are much more uncertain.

Other factors which interplay are cultural differences and reporting data differences. Certainly, if you have ever travelled through Japan then the cultural differences in hygiene and also in the populations general adherence to rules from authority (also see China, Taiwan) are very different from England or the US.

As an aside, I do think the politics of mask wearing especially in the early days of the pandemic in Europe and the US were surprising to me - although not in hindsight. There was (and is) a strand of thought as to how so simple an intervention could have an impact. A walk through a poorer country or even a more mixed one like South Africa would not scorn “simple” interventions so heavily (access to proper toilets and hygiene make huge impacts). I do think - again with hindsight - it is surprising that more weight was not given to first principles - in that we knew the virus was carried in aerosol droplets (and like colds, flus) and so the physical methods of transmission could well be interrupted by barriers like masks.

Putting this all together what does this mean? In my view, vaccines are likely to give protection for at least c. 12 months and likely to reduce transmission rates, but vaccine hesitancy, mutation and maybe some amount of re-infection will mean that the virus stays with us permanently like influenza does. However like ‘flu we will find this disease manageable. We may also never know for sure why certain groups (eg men) suffer higher mortality. 

The medium to long term speculative thoughts is that this crisis will spur more innovation and creativity across several domains.

This is because many will conclude it is human innovation that has saved us and will save us. Similar thinking may be applied to climate challenges (I expect Bill Gates will double down on this in his next book). I also think - while with much pain- the creative arts will also react with more creativity, although extremely crimped near term, as people will have to find new ways of reaching audiences/consumers.

Here are a mix of random thoughts and questions that I considered when thinking about COVID:

Where did SARS-CoV-2 come from?

Some uncertainty, but seems very likely that it came from animals (zoonotic, maybe bats) and crossed into humans. Evidence that is was present in China in November 2019 (as early as 17 Nov) and maybe earlier. Open question. We don’t know if the virus mutated in animals and then crossed to humans. Or crossed to humans and then mutated and crossed human-to—human.

Definitely seems NOT lab made (IMO).

https://www.nature.com/articles/s41591-020-0820-9

https://www.scmp.com/news/china/society/article/3074991/coronavirus-chinas-first-confirmed-covid-19-case-traced-back

Why have certain regions (Taiwan, South Korea, Singapore, Hong Kong) handled the pandemic better than others (Italy, Spain, all of Europe, US….)?

…Same for sectors and businesses ?

The high-performers had:

-Very prepared systems

-Responsive public health authorities

-Responsive general public

-Responsive private companies (at the request of the public health authorities)

But, they had very prepared systems + public because:

-They had dealt with the trauma and cost of SARS-classic

The actions were/included:

-Early responses (masks, restrictions)

-High testing (fast deployment + development of tests)

-Strict isolate, contact, trace protocols

-Travel bans and similar

-Tracking of quarantined people

There is a 124 point list of what Taiwan did:

https://www.vox.com/future-perfect/2020/3/10/21171722/taiwan-coronavirus-china-social-distancing-quarantine

https://jamanetwork.com/journals/jama/fullarticle/2762689

…Same for sectors and businesses ?

Some sectors/businesses:

-had more awareness on what exponential growth can look like (tech), and/or, 

-had more respect for the seriousness that China were taking (and put weight on that signal)

-more redundancy built into their supply chains (typically, as product considered critical, eg insulins, other must-have pharmaceuticals)

-more cash on balance sheets to deal with emergencies (typically these were maybe ear marked for litigation or other catastrophic events)

-ability to remote work

-business models that are resilient to COVID (eg. Video conference calls)

This has lead to:

(Parts of) Tech + Health + Utilities > most business

Big business > small business


Within countries / regions 

Some regions influenced by:

-understanding of exponential growth (Tech community in San Francisco)

-population density

-culture

-strains

-maybe weather?


Open Question: Why are death rates different across European regions, Asia etc ? Also, knows as heterogeneity.

We don’t know. 

We do know:

-Data is patchy

-Testing criteria are different

-Testing efficacy varies

-Older people, men, people with underlying diseases (eg heart problems) are more at risk

(But even here, there are regional differences with US rates of hospitalisation in the young much higher than in other regions).

-Different strains

-Different genetics

-Different cross protection

No one has a model that explains these intersecting factors.

One tentative suggestion is the difference in “viral load” or dosage of virus you get on first infection may explain part of this.

We do know viral load can have an impact with other viruses.

Open: Why are some people more susceptible than others?

This goes across many subgroups: Children, Men, but also differences in the young who do get impacted.

Open Question: Why are death rates so low in children? This pattern is consistent across regions even if rates vary. Explanations include:

  • Children’s immune system being more flexible and rapid

  • Adult immune system may over react due to priming with other coronaviruses

  • Adult immune system being slower

  • Other varieties of explanation…

See: Christakis https://twitter.com/NAChristakis/status/1243883141900763137

Open Question: Why are death rates higher in men? (also Co-morbidities)

We don’t know. Partial explanations that I have seen touted but with no evidence include:

-men being worse at hand washing/hygiene 

-men being more likely to smoke or use vapes.

But, essentially whatever your underlying risk the virus seems to magnify it (eg age, male, underlying diseases)…

Open Question: How long will immunity last? (Likely ranges, we have looking to be quite a few months, I’d would hone in on at least a year) 

Partly Open Question: How long does a person remain infectious? (We have some likely ranges)

Partly Open Question: How exactly is the virus spreading? (While we know it’s via viruses in droplets, we don’t really know if it’s surviving to infect people in open spaces as opposed to enclosed spaces. There’s tentative evidence that open spaces are safer (some outdoor mass events protests have not lead to super-spreading spikes but some internal ones have, also cf. different experiences in Italian cities, also Brazil) . Even if viruses can survive on cardboard in a lab how that works in the real world is unclear.)

Taboo Cognition, Tetlock

From Philip Tetlock (2003) paper on Taboo Cognition: Thinking the unthinkable: sacred values and taboo cognitions.

Many people insist that their commitments to certainvalues (e.g. love, honor, justice) are absolute and inviol-able – in effect, sacred. They treat the mere thought oftrading off sacred values against secular ones (such as money) as transparently outrageous – in effect, taboo. Economists insist, however, that in a world of scarce resources, taboo trade-offs are unavoidable. Researchs hows that, although people do respond with moraloutrage to taboo trade-offs, they often acquiesce whensecular violations of sacred values are rhetorically reframed as routine or tragic trade-offs. The results reveal the peculiar character of moral boundaries onwhat is thinkable, alternately punitively rigid and forgivingly flexible.

This article summarizes an emerging body of research that explores how people cope – cognitively and emotionally –with a fundamental contradiction of social life. The contradiction can take diverse forms but its canonicalform can be stated simply.

On the one hand, as economistsfrequently remind us, we live in a world of scarce resources in which, like it or not, everything must ultimately take onan implicit or explicit price. Indeed, this austere insight prompted Oscar Wilde to define an economist as someonewho knows the price of everything and the value ofnothing.

On the other hand, sociological observers pointout that people often insist with apparently great conviction that certain commitments and relationships aresacred and that even to contemplate trade-offs with the secular values of money or convenience is anathema. In the social world inhabited by most readers of this journal, to be caught calculating the opportunity costs of one’s family or professional integrity or loyalty to one’s country is to reveal that one ‘just does not get it’ – that one simply does not understand what it means to participate in these rule-governed forms of social life in the roles of parent/spouse, scientist or citizen.

When economic necessity collides with cultural-identityand moral-religious imperatives, and in the modern world such collisions are common, the resulting dissonance can be excruciating.

Finite resources sometimes require placing at least implicit dollar valuations on a host of things that society at large, or vocal ideological sub-cultures, adamantly declare non-fungible: human life (what price access to medical care?), justice (what price access to legal representation?), preserving natural environments (what price endangered species?), and civilliberties and rights (can ethnic – religious profiling toidentify terrorists be justified on Bayesian and cost –benefit grounds?).

This article explores these issues in two sections. The first section offers a working definition ofsacred values and a set of hypotheses concerning howpeople cope with secular encroachments on such values. The second section sketches the principal lines of empiricalwork bearing on these hypotheses.

Conceptual backdrop

Political philosophers – from Aristotle to Marx and Nietzsche – have long speculated that citizens are morelikely to do what they are supposed to do if they believe themoral codes that regulate their lives are not arbitrarysocial constructions but rather are anchored in bedrock values that transcend the whims of mere mortals. ‘Don’t do x because I say so’ has less impact than ‘don’t do x because God says so’. By the middle of the 20th century, prominent anthropologists and sociologists had made the complementary observation that, although there is vast variation inwhat groups hold sacred, sacredness seems to qualify as afunctional universal across societies, both primitive and modern, and that moral communities erect a variety of psychological and institutional barriers to insulate sacredvalues from secular contamination.To jumpstart social-cognitive research on this topic,T etlocket al. defined sacred values as those values that a moral community treats as possessing transcendental significance that precludes comparisons, trade-offs, orindeed any mingling with secular values. Of course, the policy a community proclaims towards a sacred valuerepresents an expressed, not a revealed, preference. Our actual choices may belie our high-sounding proclamations that we have assigned infinite weight to the sacred value.

Tetlock et al. advanced a sacred value protectionmodel (SVPM) that asserted that, when sacred valuescome under secular assault, people struggle to protecttheir private selves and public identities from moral contamination by the impure thoughts and deeds impliedin the taboo proposals. The SVPM can be captured in three interrelated sets of propositions: moral-outrage hypotheses, moral-cleansing hypotheses, and reality-constraint hypotheses.

… to conclude:

…Intuitive theologians are suspicious, and unapologetically so, of the classic Enlightenment values of open-minded inquiry and free markets. Opportunity costs be damned, some trade-offs should never be proposed, some statistical truths never used, and some lines of causal/counterfactual inquiry never pursued.

More available here and Tetlock’s site here.

A blog on his Superforecasting ideas here and putting the ideas into practise on judging success of biopharma drugs here.