Harms (Part 4: Inequality)

Global inequalities seem to be about as great today as they were at the peak of Western imperialism in the early 20th century. Indeed, the share of income presently captured by the poorest half of the world’s people is about half what it was in 1820, before the great divergence between Western countries and their colonies. In other words, there is still a long way to go to undo the global economic inequalities inherited from the very unequal organization of world production between the mid-19th and mid-20th centuries.

World economic inequality report, 2022
Listen to this post

1. Introduction

This is Part 4 in my series Harms. Risk mitigation has potential harms, as well as benefits. This series aims to chronicle some of the harms that existential risk mitigation may bring about, so that the value of risk mitigation efforts can be properly assessed.

Part 1 looked at the risk of distraction. Part 2 looked at surveillance. Part 3 looked at the cost of delayed technological development.

Today’s post looks at the propensity of longtermism to increase global inequality. Whereas short-term interventions were primarily inequality-reducing transfers of wealth from the richest to the poorest nations, longtermist dollars often flow to the wealthiest societies, and indeed to some of the most advantaged individuals within those societies.

I will look at the state of global inequality today (Section 2) and reasons to be concerned about inequality (Section 3). Then I will discuss how short-termist philanthropy tended to reduce inequality (Section 4) whereas longtermist philanthropy tends to exacerbate inequality (Section 5). Section 6 concludes.

2. Inequality today

We live in a deeply unequal world. Some of us have a great deal of resources. Others have rather less. Here, for example, is the global income distribution.

If that was not bad enough, the figures for wealth inequality are far worse. Here is the World Inequality Report‘s estimate of income and wealth inequality in 2022

One of the most concerning facts about global inequality is that there is high inequality between nations. While there is some evidence that economic inequality between nations may finally be beginning to fall, the past few centuries have left us with a good deal of catching up to do. Here is how per-capita GDP has evolved across the world over the past two centuries:

Two centuries ago, inequality between nations was much smaller than it is today. This inequality has real consequences for the quality of people’s lives. Here, for example, is an illustration of the largest between-country inequalities in child mortality, life expectancy, schooling, and income.
These differences are quite stark. Importantly, many of these differences are quite smoothly related to GDP. Here, for example, are twelve representative measures:

I hope that these figures motivate three claims. First, global inequality is high today. Second, much of global inequality occurs between nations. And third, inequality has a dramatic impact on the lives of people today.

3. Why care about inequality?

On some views, inequality or its correlates are intrinsically bad. For example, prioritarian views in population ethics hold that it is better, other things equal, for a benefit to accrue to poorer members of society than for the same benefit to accrue to richer members of society. This is not (merely) because richer members of society might appreciate the benefit less, but instead because it does not always seem so valuable to concentrate benefits among a small few. For example, those moved by Robert Nozick’s utility monster may think it is not so good to concentrate a great amount of utility within a single person, leaving other members of society badly off, and that it would be better to spread that utility more equally among society members.

On other views, inequality is bad because of what it communicates. For example, Andreas Mogensen has argued that unequal distributions of healthcare resources may be bad because they communicate greater respect for the value of some lives than the value of other lives. In the same way, leaving residents of poorer nations to make do with so little when many others have so much may communicate that little value is placed on the lives and well-being of residents of poorer nations.

Inequality also matters on instrumental grounds. Most obviously, the marginal utility of additional money diminishes rapidly. A software engineer earning $400,000 does not enjoy eighty times as much welfare as a farmer earning $5,000. This means that unequal distributions of wealth are, other things equal, highly inefficient from the point of view of promoting welfare.

Inequality breeds a number of further instrumental harms. Within-society inequalities erode social cohesion, which can lead to a range of harms up to and culminating in revolution or civil unrest. There is some evidence that inequality dampens economic growth, though the jury is still out on this question. And inequality both reinforces and is reinforced by structures of social domination that may be intrinsically or instrumentally bad in themselves.

None of this is to say that inequality is always wrong. But we pay a heavy price for inequality, and that price is worth taking seriously.

Let’s look at two ways in which the turn to longtermism exacerbates inequality: first, by incurring an opportunity cost in terms of foregone inequality reduction (Section 4) and second, by directly exacerbating inequalities (Section 5).

4. How short-termism reduces inequality

Short-termists favored causes such as global health and development that were excellent tools for inequality reduction.

Consider cash transfers as an example of work done to combat poverty. There is good evidence that cash transfers reduce income inequality by providing recipients with the financial means to pay immediate expenses and invest in their financial futures. Cash transfers also reduce inequality across many other measures, such as health inequality, providing individuals with the means to acquire needed healthcare.

Alternatively, consider malaria nets as an example of global health work. For example, the Against Malaria Foundation estimates that every dollar spent yields $12 in economic improvements in some of the most economically disadvantaged regions on Earth. And a well-known review estimates the economic costs of endemic malaria from 1980-1995 in many countries in the range of 15-20% of overall income. For brevity, I’ll reproduce roughly the first half of the full table below:

Source: Sachs and Malaney (2002)

To the extent that short-termist programs tended to robustly reduce inequality, longtermist interventions may incur an opportunity cost of foregone inequality reduction.

5. How longtermism exacerbates inequality

Whereas short-termists primarily aimed to transfer resources from the richest to the poorest nations, longtermists are, if anything, transferring resources the other way.

Most directly, longtermist grants increasingly go to members of advantaged groups. To illustrate, let’s look at the last 5 grants approved by Open Philanthropy (I write these words on April 5, 2024):

  1. $60,000 to Good Impressions “to run Amazon advertisements for books relevant to global catastrophic risks”
  2. $150,000 to UK Day One “to support the development of crowdsourced science and technology policy ideas”
  3. $91,100 to People for Animals Uttarakhand “to support its work on corporate cage-free campaigns in India”
  4. $220,000 to cFactual “to support its fellowship program for strategy and management professionals.”
  5. $100,000 to Asociación para el Rescate y Bienestar de los Animales “to support its corporate cage-free campaigns in Peru”

The short-termist grants are to organizations in Uttarakhand and Peru. Here, directly from the grants page, are pictures of the grant recipients:

What about the long-termists? cFactual is a consultancy founded by two alumni of the Boston Consulting Group. They received money to award to strategy and management professionals. UK Day One is a UK-based policy initiative. Good Impressions is an organization that funds advertisements and SEO for nonprofits. Their founder is also an alumnus of the Boston Consulting Group, and I believe that they are headquartered in Toronto.

So far, we have seen that many direct longtermist grants go to organizations headquartered in wealthy nations, whereas direct short-termist grants often go to recipients in a wider range of nations. The same can be said for community building funds, which are preferentially concentrated in wealthy nations.

For example, here are (as of today) all of the upcoming EAG and EAGx events listed on effectivealtruism.org.

These are not exactly located in the world’s most disadvantaged cities or nations.

6. Taking stock

Today’s post explored the state of inequality in the world today and gave instrumental and intrinsic reasons to care about inequality. From there, we looked at two ways that longtermist funding may exacerbate inequality. First, longtermist funding exacerbates inequality through the opportunity cost of foregone short-termist funding, much of which was directed to the world’s poorest nations. Second, longtermist funding may directly contribute to inequality by directing resources to the world’s wealthiest communities.

To say that longtermism exacerbates inequality is not, in itself, to say that longtermism must be wrong. But it is certainly to point to an identifiable harm of longtermism that should be figured against the value of longtermist interventions.


Posted

in

by

Comments

11 responses to “Harms (Part 4: Inequality)”

  1. Andrew Wu Avatar
    Andrew Wu

    You write “Most directly, longtermist grants increasingly go to members of disadvantaged groups” — it seems you mean the opposite?
    (Also a minor typo – should probably be campaigns, not capaigns)

    1. David Thorstad Avatar

      That’s embarrassing! Thanks, I’ll fix this now.

  2. Benjamin Avatar
    Benjamin

    I wouldn’t consider Open Philanthropy animal-focused grants “longtermist”, and I don’t think most other people would either.

    1. David Thorstad Avatar

      Agreed! The vast majority of animal-focused grants are short-termist. Can you help me to see how this connects to the discussion in the post?

  3. Benjamin Avatar
    Benjamin

    Ok, I think I’m a little confused by your wording- you say you’re looking at longtermist grants, then you cite a list of grants, but then you clarify that some of them are shorttermist?

  4. Jason Avatar
    Jason

    Almost all of this seems contingent on where the resources would have gone but for longtermism. I can think of at least one donor to some longtermist-associated causes whose counterfactual uses of the money might include things like overpaying for a social-media site and donating to political candidates in the US who are unlikely to combat inequality. So I think longtermists can seek money from that dude without giving much thought to the opportunity costs you describe.

    On the other hand, US educational reform (while having some anti-inequality effects on its face) carries many of the downsides you indicate here if the resources are coming from the Gates Foundation. So it is the funder’s identity that is doing much of the work here.

    Other than those coming from Open Phil, I am not sure how many resources currently devoted to longtermist pursuits would otherwise be going to inequality-reducing efforts. But of course Open Phil funds a large fraction of that work at present.

    1. David Thorstad Avatar

      Thanks Jason! It’s good to hear from you.

      You’re certainly right that assessing counterfactual impact is tricky. There are some donors, such as SBF, who probably would not have been interested in short-termist causes, except for PR reasons. It might not be worth trying to talk them down.

      At the same time, one thing I like about the effective altruism movement is that it does not take its donors’ preferences as set in stone. Effective altruists are unusually willing to tell their donors that they are wrong about how to do the most good and should change their donation behavior. And although effective altruists are far from perfect, they have had a good deal of success with this approach. So I do not want to be too quick to judge that longtermist donations could not have been redirected.

      It’s also important to acknowledge that donors’ interest in funding longtermist projects was not created in a vacuum. Their interest in longtermism was largely driven by movement-building, publication and outreach within the effective altruism movement. (For example, to hear him tell it, Sam Bankman-Fried was converted to effective altruism after a meeting with Will MacAskill). In these cases, I don’t know that effective altruists can entirely shift the blame for donors’ preferences, even if those preferences are now set in stone. After all, effective altruists had a good deal to do with those preferences coming about.

      I think you are also right to acknowledge that Open Philanthropy is one of the largest funders in this space at present. They have also, as you know, substantially reduced their allocation to GiveWell and put pressure on other short-termist causes. I do hope that Open Philanthropy will think seriously about the inequality-increasing effects of this decision.

      Could you say more about your concerns about the Gates Foundation and the funder’s identity? I want to make sure that I respond well to this concern, and I’ll do a better job if you can share a few more details.

      1. Jason Avatar
        Jason

        Gates’ website lists six focus areas, one of which is “US Programs” and seems to be mainly education-based initiatives. Spend is about $692MM, roughly 10% of their total spend. [https://www.gatesfoundation.org/about#faq] Although I haven’t studied those programs in depth, I think it’s safe to say that I would be over the moon if Elon Musk funded them, generally positive on the median US secular donor funding them, but pretty negative on Gates funding them (because many of those funds would have counterfactually been used on Gates global-health programs I feel very strongly about). So I’ve covered the bulk of my emotional gamut, even though the absolute merits of the US Programs are the same in all hypotheticals!

        To me, there seems to be some significant parallels between longtermists lobbying Open Phil for money (allocated at the cost of its best-in-class global health work) and US education advocates lobbying Gates for money (allocated at the cost of the same). I don’t think the relative difference in value one places on longtermist work vs. US education work ultimately contributes that much to the analysis as long as both are 1-2 OOM under the global-health work that is being displaced. (Suppose, for instance, one thought the US education work was 5% as good as global health, while longtermism was only 0.5% as good. The reduction in value is 99.5% versus “only” 95%).

        As you know, I am saddened by the extent of Open Phil’s shift away from global-health interventions. But I try to cross-check my reactions against generally analogous situations for consistency in an attempt to avoid being too hard or too soft on people.

        I agree that some EA leaders bear a measure of responsibility for certain donors becoming (or remaining) locked, and that factor could help me distinguish the longtermism/Open Phil situation from the education/Gates situation. That is particularly so to the extent that leaders are *explicitly* pitching that their longtermist causes are better than the other causes in that funder’s portfolio. My assumption is that outside education advocates aren’t making that kind of pitch at Gates. Then again, I don’t think most grant applicants at Open Phil are making those arguments either; they take Open Phil’s allocation between causes as a given and “compete” within their cause.

        My sense is that the number of people who have the kind of power / responsibility I’m describing here is fairly low (perhaps in the dozens?) And it would need a donor who would be counterfactually open to pro-equality work. I have zero inside knowledge, but I would not be surprised if many recent donors are there because they are concerned about AI, rather than being convinced by EA and then led to fund longtermist work.

        1. David Thorstad Avatar

          Ah, I see!

          Yes, you are certainly right that many others, including the Gates Foundation, often fund US-based interventions when interventions elsewhere could do more good. Even if the Gates Foundations’ educational interventions do tend to reduce within-country inequality in the US, that is likely to be far outweighed by the potential for reduction of between-country inequality from global health interventions, or even from education interventions abroad.

          One thing I appreciate about effective altruists is that they spend a lot of time thinking about cause prioritization. This means they are uniquely receptive to thinking about the upsides and downsides of cause areas, and this makes it worthwhile to write about some of the neglected harms of longtermist causes. If a funder is dead set on funding a certain cause area, there is, as you say, less point in trying to talk them out of it. For example, I don’t know that I ever would have tried to convince Bankman-Fried to stop funding longtermist causes.

  5. titotal23 Avatar

    I think these are good points.

    One correction: “Two centuries ago, inequality between nations was much smaller than it is today” might not be true when you consider it in term of percentages. This source, for example, has the share of wealth held by the top 10% at 50% in 1820 and 55% now: https://wir2022.wid.world/chapter-2/

    1. David Thorstad Avatar

      Thanks! It’s good to hear from you, as always.

      We’re both reading the same report (this is the report that heads the article), and I think we’re largely agreed. The figure that you cite is that the global income share of the top 10% of income earners has remained relatively steady (50% in 1820 and 55% in 2020). I think there are also a few points worth emphasizing.

      First, the same chart shows the share of incomes held by the bottom 50% of income earners falling from almost 15% in 1820 to about 7-8% in 2020. That’s a very large drop – almost a halving. This part of the chart (Figure 2.1) is consistent with the story that opened the chapter: there was a strong increase in inequality for about a century after 1820, and much of that inequality persists today. (For example, the income share of the bottom 50% of income earners was about the same in 1920 as it is today).

      Second, it’s important to distinguish between income and wealth. While income inequality is merely bad, wealth inequality can be quite extreme. We saw in Figure 1 in the World Inequality Report that the bottom 50% of income earners bring home 8.5% of global incomes, but own just 2% of global wealth.

      Finally, I do think that when we look at inequality between nations, rather than income or wealth categories, the story does not look great. Many measures of between-country inequality rose steadily for two centuries, until very recently. This isn’t great for those who care about between-country inequality as well as within-country inequality.

Leave a Reply

Discover more from Reflective altruism

Subscribe now to keep reading and get access to the full archive.

Continue reading