I received an intriguing book from Dr Kerry Vaughan called Doing Good Better by Will MacAskill. I don’t often take the time to review books here, but this one is sufficiently aligned with the project that I am undertaking on this site that I thought I would dedicate a post to it, rather than merely a footnote at the base of my other posts. MAJOR DISCLAIMER: If you have read nothing about Effective Altruism, then this book is a great read, and I would encourage you to read it. Most of my critiques are for those who have experience in the community. Let me further highlight that while most of my comments below are negative, they are not designed to undermine the enterprise that MacAskill is undertaking, nor do I believe that the book is not worth reading. I found the book to be a good popularization of some of the toughest questions that we should grapple with, easy to read, and with a good balance of elucidating anecdotes and overarching statistics.
The book starts off by trying to compare which endeavors are most likely to do good in the world. Will searched for a fungible metric and lands on the Quality Adjusted Life Year (QALY) as the appropriate metric. This metric tries to not merely count the number of years of life added by and intervention, but also the subjective value of those years. I have three major issues with this:
- The circle of care: I spoke about this in my ‘Why’ post. It assumes that the right metric is human beings. As someone who just found out that I have more Neanderthal DNA than 97% of the people tested by 23andMe, this is concerning. If there were two species of humans on the earth, or if other animals can suffer, or if we stumble upon/create sentience then we must grapple with whether human life is the metric we should all be aiming towards. In an overly-simplistic critique, it could be considered bigoted. This is especially the case when much of the rest of the book extolls the virtues of investing at the areas of highest marginal-utility. By any measure, philanthropy directed at each individual non-human species is much less than the spending on humans. I will be the first to admit that I don’t feel compelled to chain myself to a tree to prevent its destruction, but then I too, could be a bigot. A bit of worldview diversification, or at least an explanation why this anthropocentric view is taken would be worthwhile.
- The method of obtaining a QALY is suspect, in my opinion. Macaskill says as much. Those who are actually in dire straights (e.g. those with long term physical disabilities) do not see their lives deteriorate as much as the QALY predicts. People seem to have an emotional set-point for happiness that they tend towards, irrespective of circumstance.
- The QALY does not take into account the time-value of life. That is, there is a very real risk that like 99% of the species that have ever lived, we will go extinct. We don’t know when or why this may happen, but we need to discount the future to some degree. This is not as critical for interventions taking place now that will affect the people currently alive; however, for those long-term interventions that could save people many generations into the future, this uncertainty about the future needs to be taken into account.
Stylistically, those already familiar with the Effective Altruism movement will be disappointed that many of the same examples of irrationality are trotted out.
I find few issues with the next few chapters Will:
- Reminds us that essentially everyone reading the book is in the 1 percent globally.
- Goes through the basis of utilitarianism, the greatest good for the greatest number, a framework that can be applied to nearly any philanthropic endeavour.
- Covers the value of foreign aid, and how to think about non-normal distributions
- Explains that the most salient events are not always those where aid is most needed, the concept of marginal utility
Will then asks us to consider the counterfactual. That is, what would have happened if something had not been done/left undone. I found this chapter to be weak. As I covered in my Chaos post in most cases, it is not possible to determine ex-ante what our actions will do. Who would have guessed that a bad painter who ran in and out of the trenches in WWI would have ended up leading the 3rd Reich? Much of our society could be characterized as a second-order chaotic system. The ability of any individual to determine the impact of her actions in this type of system are limited. There are certainly commonalities, but as with investing, once those commonalities are seized upon they tend to disappear, or have ironic consequences. A bit of epistemic humility (there are limits to what we can know) and an overview of how we try to isolate the effect of independent variables through regression analysis would have gone a long way here.
Will inexplicably takes on the irrationality of voting hypothesis. This is a bit tragic because there has been a lot of ink spilled on this already by many economists who give a great deal of thought to marginal utility. Will, without sufficient explanation, changes the traditional value of a vote from an individual value to the value for everyone in the polity. Economists who view rational utility maximization as an individual objective will take issue with this, he will likely turn many of them off. I personally don’t have an issue with this, but Will inexplicably stops at the nation’s borders. The effects of a vote go far beyond a nation’s borders. For instance the 2006 House election and its effect on the Iraq/Afghan wars. Better treatment of human’s poor ability to deal with low-probability high-impact events is covered by Nassim Taleb, and the text version of the Big Short where the Cornwall Capital investors explain their investment strategy. I will personally be covering investing at this level in a post in the coming weeks.
In the second half of the book MacAskill covers the operationalization of some of these strategies.
This section has some good anecdotal evidence, but lacks some narrative structure. I assume that the reason MacAskill separated this section from the first section is that the first half of the book would be too unwieldy without it. Indeed, the first half does benefit from this separation, but the second half is little-improved.
The content of the second section will be familiar to most who have spent time in the effective altruist community. MacAskill talks through:
- The typical methods used to evaluate nonprofits vs their effectiveness
- GiveWell, and some of the top charities to donate to
- The ironic consequences of ethical consumerism
- I have several critiques here:
- MacAskill extrapolates from a few anecdotes about ethical consumerism gone wrong, and extrapolates this to all of ethical consumerism. He misses cases such as Vanguard, Credit Unions, and other entities that exist only to enrich their members and can cause a great deal of good. The choice between working for Goldman, and working for Vanguard has some moral content.
- MacAskill assumes that the choice is often between the ethically produced goods and donating. For many people, this is not the mental framework they use to address their giving. Their everyday giving is often nothing, or through one of these organizations.
- The sum-total of this chapter points towards a ‘Band-Aid’ method of addressing the world’s problems where we address the negative externalities of various businesses instead of the way they are conducted. In any case where the costs to repair > costs to avoid, then only the latter choice makes sense. If this approach were taken with, for instance, with childhood vaccinations, then the number of QALYs would be greatly reduced for preventable disease. The same logic could be applied to CFCs.
- MacAskill reveals that he is a vegetarian, and talks about comparing the suffering of animals with something like QALYs. Unfortunately, he equates the suffering of chickens with that of pigs and cows. He does this interspecies comparison without justification, and makes no original tie back to the human QALYs. In principle, I am not against these interspecies comparisons, it merely adds fuel to my original critique that the original focus of the book on QALYs is unjustified.
- I have several critiques here:
- An overview of the work that has been done at 80000 hours, an organisation that tried to help people selection impactful careers.
- Finally, an overview of cause-selection, how to compare nuclear proliferation to childhood health to animal welfare. As I have already expressed, this should have been up front
More information about much of this thinking is at http://effectivealtruism.org/
Doing Good Better by William MacAskill is available on Amazon in every format imaginable.
Please, tell me what you thought before reading this, and let me know if this changed your mind.