#FromMyReadings: Issue 2, 2021
COVID-19 relief in Karnataka; Growing global divergence; And why “you should have read this before” is not the greatest argument nowadays.
[Before you read on, I urge you to take a look at the beginning of my last post and donate to those causes if you can. India can do with more COVID-19 relief efforts. Thank you.]
1. “Collaborate, Catalyze, Cultivate”: Learnings from COVID Relief Work (Anjana Balakrishnan, CSEI)
In my last post, I’d briefly mentioned the COVID -19 relief work being done by the Centre for Social and Environmental Innovation (CSEI) and the Institute of Public Health (IPH) in Chamarajanagara district, Karnataka.
I won’t describe it in detail (it’s best to read directly from her own account), but I want to highlight something her piece points out — that merely having domain expertise is not enough to make changes on the ground. Often, networks (and infrastructure to enable networks, like WhatsApp), skills like communication, goodwill, and long-term commitments can play equally important roles in enabling change, if not more.
On the flip side, this highlights a growing challenge for researchers. On one hand, researchers definitely need to build these skills and networks if they hope to make any kind of improvement to society. On the other hand, research also needs an arm’s-length distance from on-ground practice for building longer-term, larger-scale perspectives. Arm’s-length distance isn’t popular — it’s too often associated with ivory towers, arrogant academics, and abstract notions with no relevance to ground reality.
But the pandemic has also highlighted the importance of sometimes air-gapping research from daily practice. For example, on-ground doctors in India were quick to prescribe drugs like Remdesivir or steroids to provide COVID-19 relief, despite reservations expressed by more distanced medical researchers. Now, we are grappling with the consequences of not thinking through our actions.
In the near future, researchers in many fields will have to contend with this challenge more closely*. I don’t think it will be possible for researchers to isolate themselves from on-ground practice entirely — in fact, I do not recommend this. However, from time to time, they’ll need the courage to walk away (temporarily) from the day-to-day quick responses and fast thinking that on-ground practice involves. I say ‘courage’, because slow reflection will definitely be unpopular, subject to tighter resource constraints, and likely to be misused. It’s not going to be easy to balance the two in a world that’s accelerating.
*PS — Just to clarify, I’m not saying this is a new challenge (research versus practice is an old old debate), but it is a challenge that’s becoming harder to navigate.
2. A More Unequal World (“The Great Acceleration”, McKinsey & Company)
Speaking of acceleration, it is a key theme in a couple of publications (a post and a related podcast) brought out by McKinsey earlier this year, about how our world is going to shape up, post-pandemic.
Unfortunately, the outlook is grim. The posts highlight how, unlike the 2008 recession, the COVID-19 collapse hasn’t changed much of our existing market and economic structures. On the contrary, what we are seeing is an acceleration of existing characteristics. Countries and organisations doing badly before the pandemic are likely to do worse, while those performing well are likely to do even better. McKinsey isn’t the only organisation to note this. There have been other pieces this year which highlight how western billionaires got wealthier during the pandemic, and how emerging economies are faring worse than developed ones.
A lot will depend on what’s going to happen over the next couple of years. It’s possible this divergence is a temporary aberration (or even a miscalculation) and the world will begin to correct itself as the pandemic recedes. However, neither is this provable, nor can it be taken for granted. A divergence in economic growth for countries could very well have much longer-term effects and we must prepare ourselves for such a scenario.
What I’m concerned about is how this can translate to a hardening of practices at smaller scales. As resources and opportunities shrink for ‘losers’ in this divergence, they may revert to more conservative practices. Businesses may stop experimenting, universities and think-tanks may become more cautious about their research, and critically, governments may revert to tried-and-tested politics. For emerging economies, who were just beginning to experiment with new pathways to prosperity, nothing could be more disastrous.
In the Indian context, a part of me wants to rant about how we squander opportunities during times of growth while refusing to experiment in tougher conditions. Before the 2008 recession, Indian businesses went on a profligate borrowing and acquisition spree, wasting valuable reserves on dead assets which in turn created our Non-Performing Assets (NPA) banking crisis in the 2010s.
The 2010s gave us another chance to rebuild ourselves, with a boom in the startup economy that promised new and innovative ways of increasing employment while boosting growth via consumption. Our results on these have been mixed. While we did build some interesting models like Swiggy/Dunzo, the startup scene was eventually captured by people trying to game the system. There was too much focus on valuations and sell-offs rather than actually building infrastructure and services on the ground.
With businesses and consumption now hit by the pandemic, there’s a real danger that negative sentiments could translate into cynicism about new ventures. Furthermore, the stress of the pandemic upon household incomes and savings is likely to dent consumption. I sincerely hope I am wrong. It’s still early to say what the post-pandemic effects are likely to be in India. However, it’s clear that we have already lost two years and may lose a third. That’s not good news.
3. “Why Didn’t You Read This Before?” is Not the Greatest Argument Nowadays
Psychologist Daniel Kahneman, in collaboration with Olivier Sibony, and Cass Sunstein, has released a new book called Noise: A Flaw in Human Judgement. According to its Amazon page, Noise is about:
“the detrimental effects of noise [variability in judgments that should be identical] in many fields, including medicine, law, economic forecasting, forensic science, bail, child protection, strategy, performance reviews, and personnel selection.”
The book has generated quite a bit of…well, you know…noise. Some reactions are excited, some are cautiously skeptical and some are critical. My post here is not a response to the book itself, but to one of these reactions.
Columbia University Statistics Professor Andrew Gelman provided a critical response to Noise. Reacting to an author’s claim that:
“…while correlation does not imply causation, causation does imply correlation.”
Dr. Gelman quotes econometrician Dr. Rachel Maeger that this argument is “just outright incorrect”, explains why it’s wrong, and then asks how the authors managed to make such a basic mistake.
Fair enough, this argument is something I can agree with. However, when diving deeper into reasons for such a mistake, Dr. Gelman ends up in a rabbit hole about expertise.
The authors of this new book are a psychologist, a law professor, and some dude who describes himself as “a professor, writer and keynote speaker specializing in the quality of strategic thinking and the design of decision processes.”
Between them, there’s no reason to think they’d have any particular expertise in correlation, causation, or statistics. You might as well ask me to have an opinion on the non-accelerating inflation rate of unemployment or the theory of operant conditioning.
If I were to write a book and include categorical statements about such things, I’d check with the experts first.
Later on in the post, he goes on to say:
I mean, really, what the hell?? I’m reminded of that scene in one of David Lodge’s books where the professors of English are sitting in a circle, playing a game where they take turns listing famous books that, embarrassingly, they’ve never read. And one of them lists Hamlet. A bit too embarrassing, it turns out!
Similarly, it’s kind of admirable how open Sunstein is about his former cluelessness, but it makes you wonder whether he was really the most qualified person to write a book about a topic that lots of people know about, but which until five years ago he’d never thought about.
Personally, I have mixed reactions to this. On one hand, I agree with Dr. Gelman that the authors should have cross-checked their findings. “Causation always implies correlation” is a bad argument to make, even if you aren’t a statistician. How do you deductively prove that something ‘always’ happens, unless you have some underlying assumptions or boundary conditions? Basic mistakes like these are embarrassing.
On the other hand, I think Dr. Gelman goes a bit far, later in his post. Take that point about an English professor who’s never read Hamlet. Question: Why should an English professor necessarily read Hamlet? Sure, it would be useful if they had, but unless the professor is teaching a course that relies on Hamlet, they are not obligated to read the play. The English language has millions of works and it’s easy to build a teaching programme from another sample that doesn’t use Hamlet or Shakespeare at all. Do we really want to argue that a course using Chinua Achebe or Jhumpa Lahiri is not as good as one using Shakespeare?
It comes back to the point of what we consider “expertise”. Does expertise necessarily require the knowledge of a ‘canon’ of works, or familiarity with the giants who came before you? Again, I’m not saying this is a bad thing, but I’m asking if it’s vital. Expertise often requires you to venture into the unknown, and dealing with new problems not encountered before. Knowing what others did before you sharpens the boundaries of your knowledge about what you can and cannot do. But is this always necessary to work with an unfamiliar issue?
This question will become more important in coming years. People without training are already venturing into fields that are new to them (Eg: Data scientists with no formal training in Statistics). Many are already solving old challenges in new settings (Eg: That moment when Lyft reinvented the bus).
Furthermore, academia is notorious for siloed work, where academics reinvent each others’ research because they haven’t kept track of other fields. I’ve lost count of the times sociologists and anthropologists have berated economists for coming to the same conclusions as them (without acknowledging sociology or anthropology). Dr. Gelman’s own point on whether statisticians should comment on employment is a bit rich, given that economic issues are now regularly tackled by math, stats, and engineering students.
However, asking people to keep track of work outside their own fields has its limits. There’s just too much going on, and unless you’re in an environment where you’re directly interacting with people from other fields, it’s impossible to keep track. Even if you are in such an environment, you’ll most likely learn about other fields in the contexts of problems that you’re studying together. For example, I’ve worked with architects, sociologists and environmental scientists, but there’s no way I can tell you what’s happening in those fields, except in the very narrow context of Indian urbanism.
This is not a defence for any mistakes made by Kahneman, Sibony, and Susstein. I agree with Dr. Gelman that in this particular instance, the bar for quality is certainly higher, especially given the authors’ credentials. However, I’d caution against turning “Why haven’t you read this before?” into a more generalised standard for judging the quality of work.
As an economics graduate now working in interdisciplinary urban practice, I am aware that there are always going to be books or papers I haven’t read that others consider seminal (and vice versa). That doesn’t mean my own work automatically becomes useless. Everything depends on the arguments I make and the evidence I provide. Canons do matter and they can enrich a study, but as to whether they’re necessary? I don’t think we can have a universal judgement on that, only in specific contexts.
