People sometimes claim that speeding economic growth and development is the project on which classical utilitarians—for short, throughout this post, I’ll just say “utilitarians”—should focus their efforts. This post is a(nother) rough attempt to grapple with the plausibility of that claim, and ultimately, I think, a rough argument against the claim. The argument is on the abstract side, but hopefully the intuition is sensible and straightforward.
Hopefully it’s also clear enough that, though I’m restricting my attention to classical utilitarianism so as to avoid worrying about exactly which assumptions can be made where, the argument applies to many consequentialist value systems.
_____
The “global distribution of wealth”, without qualifiers, is the global distribution of wealth across individuals: how much wealth each person owns. Let us analogously define the “global distribution of wealth across (final) goals” to be how much wealth—including that stored as human capital and so on—is being allocated to each objective function which someone or some group is looking to maximize.
It’s a simple enough concept, but here’s an illustration. Consider a world with two Christians (A, B) and one Buddhist (C), each with wealth valued at $100. Each follows a policy of spending 10% of his wealth on a common religious project, which he cares about for its own sake, and 90% on his own consumption. The distribution of wealth across goals is then {$90 to A’s welfare; $90 to B’s welfare; $20 to glorifying God in the Christian way; $90 to C’s welfare; $10 to glorifying gods in the Buddhist way}. The resources budgeted for a goal are spent as effectively as possible with respect to the goal. There’s no sense in which global spending is optimized across goals.
Of course, resources allocated to one goal may happen to contribute to, or frustrate, the achievement of other goals to which resources are being allocated directly. If C’s medical bills impoverish him, the Christians may conclude that their $20 can glorify God more by relieving C’s poverty than by building a church or whatever. But the distribution of wealth across goals will remain as above.
_____
Consider a world W like our current world Y, but with two big differences: it has only 2% as much wealth, and this wealth is all pursuing the utilitarian objective function.
On considering that W’s (perhaps few, at first) people would be making energetic plans to populate the earth, care for its animals, maintain their civilization and values across the generations, and so on, I think it’s safe to say that the expected utilitarian value of the future of such a world EU(W) is higher than that of the world as it stands today EU(Y). In 1900, our world
did have only about 2% as much wealth as it has today, and it was very far from all being in wholeheartedly utilitarian hands. W therefore basically
is Y—albeit after 120 years’ delay—but with dramatically improved initial conditions.
Normalize current global wealth to 1, and consider a spectrum of worlds X(
a),
a ∈ [0, 1]. Wealth in X(
a) totals 0.02 + (0.98/0.999)
a, of which fraction 1 –
a is put explicitly to utilitarian purposes and fraction
a is put to goals drawn from the current global distribution of wealth across non-utilitarian goals. Thus X(0) = W, and—if as much as 0.1% of the world’s wealth is currently “fully utilitarian”, which strikes me as a high estimate—X(0.999) = Y.
When a philanthropist spends a utilitarianism-purposed budget so as to produce an indiscriminate level shift in global economic growth, she essentially does two things: (1) increase total wealth and (2) change how wealth is distributed across goals. In particular, she decreases the share of global wealth reserved for utilitarian purposes and increases the share for everything else. If the most cost-effective ways to increase total global surplus do so at a ratio of (0.98/0.999)/0.02 = $49.05 per $1 spent (as the Copenhagen Consensus roughly concludes*), then marginal efforts to increase growth, starting from world X(0.999), are equivalent to marginal increases in
a.
Since it seems that EU(X(0)) > EU(X(0.999)), the case for spending on growth—i.e. the case that
d/
da[EU(X(
a))] > 0 at
a = 0.999—requires a strange non-monotonicity in the relationship between
a and expected utilitarian value. That is, it requires that—even though it’s bad
on average from X(0) to X(0.999) to accept a trade in which the world’s utilitarians get slightly worse-endowed, and everyone else gets substantially better-endowed, at the best ratio currently available—it’s
good on the current margin… even though we’re starting from a world with comparatively few utilitarians and comparatively abundant wealth. This seems unlikely, at face value, and I don’t see a compelling reason to believe it.
_____
This is not an argument that growth is bad on account of putting resources into the hands of rival value systems, or anything like that. The judgment that EU(W) > EU(Y) is compatible with the view that making others richer at
no cost to oneself (or one’s moral allies) does damage, on average, but it’s also compatible with the view that doing so is somewhat beneficial. If someone wants to pursue some growth-related project with time and money otherwise destined for their own consumption and leisure, I have no objection here. It’s probably not the best thing to do, from the perspective of any particular value system, but the perfect shouldn’t be the enemy of the good.
Yet this line of reasoning still paints a bleakly adversarial view of utilitarianism (and, as noted at the beginning, at least much of consequentialism). We typically think that any charitably-minded person should be glad when others, as long as they’re not outright Nazis or sadists or whatever, are better equipped to achieve their ends. But broad-based growth and development projects can be framed in terms of an explicit tradeoff: just how much should one value wealth for others in general, relative to wealth in the hands of one’s own moral tribe? Often, I guess, the answer is “hardly at all”.
*
The most recent (2012) Copenhagen Consensus ranked micronutrient supplementation the most cost-effective intervention available, in the cost-benefit analysis sense, which they estimate would, under "very conservative assumptions", create $30 in total surplus per $1 spent. In a 2019 EconTalk interview, Copenhagen Consensus Center president Bjorn Lomborg gives a figure of $45 per $1, and it is the highest ratio he cites, except for birth control interventions (which are of course complicated by their immediate interactions with population ethics) and trade policy reforms (which are blocked by political opposition).