I'd like to ask your thoughts on how to measure the impact in QALYs from advocating a rational response to the Paris attacks and ISIS.


So let's ground this out in a specific example. Here's my op-ed that uses long-term orientation, probabilistic thinking, numeracy, consider the alternative, reaching our actual goals, avoiding intuitive emotional reactions and attention bias, and other rationality techniques to suggest more rational responses to the Paris attacks and the ISIS threat.


This is part of my broader project, Intentional Insights, of conveying effective altruism and rational thinking, to a broad audience. This belongs more to the area that conveys rational thinking, and here I am specifically oriented toward the issue of rational thinking about politics and public policy. My probabilistic estimate is that many here will agree that having a more rational response to the Paris attacks, and avoiding a rash military response, will raise QALYs (or WALYs to use Ben Todd's suggestion). On a meta-level, adopting the strategies I suggest for more rational political engagement, by individuals and politicians alike, will substantially raise QALYs.


Yet it's really hard to measure the impact in QALYs. Here are some numbers to help do a rough estimate. It's published in the Sunday edition of The Plain Dealer​, a major newspaper, 16th in the US, with its Sunday edition distributed to over 424 thousand and the website having over 5 million visitors monthly. Currently, the article has over 240 comments, and a multitude of social media shares, indicating wide engagement. Since it's the Sunday edition, many people will read it throughout the week. So it has reached and will reach a wide number of people, it's about a breaking news event, the newspaper is in a major swing state and in a city that will host the Republican national convention. Hopefully some people will not only engage with the specific content, but with the broader ways of thinking about politics itself.


What's your rough estimate of how valuable this is in terms of QALYs? To enumerate it more clearly, how much money would you have donated to have this article exist as opposed to not exist? More broadly, what kind of strategies would you suggest for enumerating the QALYs generated by this kind of article? Thanks for your thoughts, it will help me and Intentional Insights target our activities more precisely so that we can improve our positive impact on the world and contribute to human flourishing most optimally!

4 comments, sorted by Click to highlight new comments since: Today at 5:14 PM
New Comment

I think in general the issue with EA political interventions is that (a) it's very difficult to predict what political interventions are desirable and (b) even conditional on the correct prediction that a particular intervention is desirable, it's very difficult to persuade people.

For point (a), I'll play devil's advocate and try to think of some reasons that the approach you outlined could be the wrong one even though I tend to agree with it.

  • From a game theoretic perspective, it's important to present a credible threat of retaliation, which emotion can play a role in. Retaliating against defectors even when it hurts you personally could be seen as equivalent to one-boxing in Newcomb's Problem: you would like to be the kind of person who does it even if in individual instances you may lose from it. To be more concrete, if the West's response to terrorism is too thoughtful and measured, that sends a signal to the authoritarian governments of other world powers (Russia, China, Iran) that they will get a thoughtful, measured response to their own acts of aggression, rather than credible retaliation.

  • It seems plausible that isolating all the young Muslims with latent radical tendencies in a single geographical area (as opposed to having them spread evenly throughout the world's Muslim communities) is a good thing. See this and this for more.

The CIA is full of professionals who collect intelligence and analyze these issues full time. Despite this, they made the wrong call in Iraq, and it was a disaster. If CIA analysts can screw up doing this as their full-time job, you can certainly screw up from your armchair. Specialization of labor is a thing.

Analyzing issues full time while working for an EA organization is one thing, but I'm not persuaded that EAs have the ability to decide issues more effectively from their armchairs than other people.

For point (b) I'll note that you don't seem to have persuaded that many of the commenters on your article.

Thanks for playing a devil's advocate, really helpful to have someone do that!

For point A:

  • I agree on the need to represent a credible threat. However, being seen as irrational and manipulated easily by terrorist attacks is not really a credible threat. Especially if doing so plays into an enemy's hands.

  • Clearly, these young Muslims are then going out into the broader world, including into Paris, and there's not really a good way to stop them.

For point B:

  • The commenters do not represent the actual readers. The readers of this newspaper live in a liberal-leaning city, Cleveland. If you look through the actual comments, you will see it's the same people making the same criticism again and again. It's a group of activists who go on newspaper sites around the country and try to persuade readers to share their opinions. I've been getting positive emails from actual readers for the last few days. Also, other places have been interested in republishing the op-ed, such as The Huffington Post, etc.

I'm not a typical EA in analyzing political issues, actually. I'm a historian, and have quite a bit of expertise in political analysis. This is the only reason I was able to get this publication into the paper, namely my credibility as an expert.

I'm still curious about your response to the question of QALYs and money.

Some simple observations.

To perform such a QALY estimate you need

  1. A credible model for predicting the consequences of possible responses
  2. An estimate of how likely your advocacy is to effect policy

1 is something you need to even know what the best response it (and I'm currently not sure whether you have it).

2 sounds like something that should have been researched by many people by now, but I'm far from an expert so no specific suggestions.

Good ideas!

1) Hard to predict the consequences, of course - so many possible ones. Lots of noise and variability. Probably best to focus on the most likely ones, namely a smallish shift toward a pro-peace stance and a small increase in the rationality of political decision-making by readers.

2) I think this is the tougher question. How to estimate this is really hard! That's why I was suggesting thinking in terms of intuitive gut reactions might be helpful.