The Ramsey Effect
February 24, 2021 10:45 PM   Subscribe

The Ramsey Effect is an essay written by philosopher Kieran Setiya in the London Review of Books on (or at least inspired by) Cheryl Misak's book Frank Ramsey: A Sheer Excess of Powers. Previously on Metafilter.
In time the world will cool and everything will die; but that is a long time off still, and its present value at compound discount is almost nothing. Nor is the present less valuable because the future will be blank.
posted by Jonathan Livengood (26 comments total) 20 users marked this as a favorite
 
A key variable in the mathematics is the ‘social discount rate’: the rate at which we discount the wellbeing of future generations, giving their lives less weight in our calculations as time goes on.

Add to that the discount rate that certain people right now apply to other people right now, but who live in other locations or have different physical characteristics. That discount is quite a bit steeper than 1%, and it is quite possibly the most dangerous impediment to progress that we face.
posted by hypnogogue at 11:13 PM on February 24, 2021 [2 favorites]


Hynogogue - this is known in philosophy as Problem of Distance, as put forth in this 1971 paper by Peter Singer.
posted by Spacelegoman at 11:48 PM on February 24, 2021 [2 favorites]


In time the world will cool and everything will die; but that is a long time off still, and its present value at compound discount is almost nothing. Nor is the present less valuable because the future will be blank.

This reminds me of a little poem I read once. I can't find the author to credit, sorry author:

"The world is everything that is the case.
So stop your crying and wash your face."
posted by thelonius at 4:07 AM on February 25, 2021 [3 favorites]


I am excited to read this! I know Kieran Setiya primarily from reading his book Midlife (which I adored) and his appearance on Hi-Phi Nation.
posted by dehowell at 4:13 AM on February 25, 2021


This reminds me of a little poem I read once.

That would be Donald Hall!
posted by robself at 4:22 AM on February 25, 2021 [3 favorites]


Thank you! Google search is so poor now, for finding anything like this. It returned nothing with quotes around the whole short poem, and omitting them returns results for words and phrases in the poem, like as if you searched for "wash your face".

Anyway. My hot take: Ramsey did philosophy a great service by initiating the unravelling of its most overrated book, the Tractatus, with its idiotic and barely explained ("2.1511 That is how a picture is attached to reality; it reaches right out to it.") picture theory, its pretentious, cryptic and unjustified metaphysics, and the whole apparatus of almost superstitious dogmatism that seems to follow from anyone who committed themselves to anything like logical atomism.
posted by thelonius at 4:39 AM on February 25, 2021 [2 favorites]


Kim Stanley Robinson's The Ministry for the Future is basically a thought experiment in fixing our abysmal social discount rate so that we have a future to look forward to.
posted by Foosnark at 5:12 AM on February 25, 2021 [5 favorites]


I read the LRB article yesterday. It was interesting but it didn't say anything about his contributions to mathematics, so I'm going to say a few words about "Ramsey Theory". I remember when I was an undergrad I wrote a computer program that searched for "Ramsey Numbers" which are associated with a particular type of geometric problem - finding star-shapes (simplices) inside of larger shapes - when the larger shape is large enough, there's a certain "inevitabilty" that kicks in and guarantees the appearance of the smaller shape. The sizes of the figures involved are the Ramsey numbers and they are some of the largest integers used in mathematics. I wrote a program on a little 80's computer to find them, while drawing the shapes on the screen - it was fun to watch but it only was able to find the first few Ramsey numbers because you'd need a computer the size of the galaxy to go much further ...
posted by crazy_yeti at 5:46 AM on February 25, 2021 [4 favorites]


Should you sell your soul for a donut? 'Endless suffering in hell' sounds bad, of course, but as long as the level of suffering is roughly constant then under exponential discounting we can see that the present value of going to hell is only a finite negative value. Therefore, for a sufficiently mild hell, or a really good donut, yes you should consign yourself to a literal eternity of pain.
posted by Pyry at 5:59 AM on February 25, 2021 [8 favorites]


Simple example of Ramsey theory. If you draw N dots and connect them, it's called a "complete graph on N vertices" or K(n) for short. K(3) is a triangle, K(4) is a tetrahedron, K(5) is a 5-pointed star inside a pentagon. If I give you two crayons, say red and blue, you can color K(5) in such a way that no all-red or all-blue triangle occurs. (Color the pentagon one color and the inscribed pentagram the other). But for a K(6), if you only have 2 colors, you are forced to create a single-colored triangle. (Try it!) This is expressed by saying R(3) = 6. If you want to "force" a K(3) you need a K(6). R(4) is 18, and beyond that the values are only known approximately - they get very big very fast, and there's no good way to compute them exactly other than searching through enormous numbers of possibilities. But a bit is known about their asymptotic behavior. There's also a variant where you have more than 2 colors and that doesn't make the problem any easier!
posted by crazy_yeti at 6:04 AM on February 25, 2021 [6 favorites]


Pyry - I think you just demonstrated the problem with the exponential discount model. There needs to be a minumum value, so that when you integrate out from the present to the infinite future, you can't get a finite value. (Personally I don't like the idea of future discount at all - all times are real & all suffering is too)
posted by crazy_yeti at 6:07 AM on February 25, 2021


Archived version
posted by Brian B. at 6:17 AM on February 25, 2021


Quoting the original article: "The idea that future generations count for less ‘is ethically indefensible and arises merely from a weakness of the imagination’."
posted by crazy_yeti at 6:24 AM on February 25, 2021 [3 favorites]


In mathematics: Ramsey theory, Ramsey’s theorem and Ramsey numbers.

The guy was a good mathematician but come on. These are all referring to the same thing. One really good idea is one more than most of us have, but the list of mathematicians who had one really good idea in their 20s is not so short.
posted by escabeche at 6:40 AM on February 25, 2021 [2 favorites]


Hynogogue - this is known in philosophy as Problem of Distance, as put forth in this 1971 paper by Peter Singer.

This is known, in social science (and in my house) as failing the Marshmallow Test and it applies not only to future generations, but to our future selves. If you have a kid with ADHD (or if you have ADHD yourself, which -- oh hai, both!) you've probably heard the term "executive function" which is a highfalutin' way of talking about the same thing.

If you believe that most Americans cannot currently pass the Test, it helps to explain a lot of what you see going on in the country, though it doesn't make things any more hopeful.
posted by The Bellman at 7:27 AM on February 25, 2021 [1 favorite]


The guy was a good mathematician but come on. These are all referring to the same thing. One really good idea is one more than most of us have, but the list of mathematicians who had one really good idea in their 20s is not so short.

This is entirely fair. I'm curious how you think about counting good ideas in applied math. Is a good idea in applied math a good idea in math, a good idea in the application area, both, or neither? It seems to me that it's really in applied math that Ramsey is excellent: his work on the conditional, his work on the measurement of belief and the foundations of decision theory, his work in economics (on taxation and saving), etc.

So like from one point of view, the work on the measurement of belief is a really powerful synthesis of some ideas in analytic psychology and pragmatism with some ideas in measurement theory. But in another sense the math was already there -- he's just applying Holder's theorem to beliefs. So, I really don't know how to think about these kinds of case.
posted by Jonathan Livengood at 8:23 AM on February 25, 2021


The Master and the Prodigy - "There is little doubt that John Maynard Keynes fundamentally shaped economics and policymaking in the twentieth century. Less appreciated is that he owes some of his central insights to a brilliant Cambridge polymath who died in 1930 at age 26."*
For his part, Ramsey allowed that absolute truth could be achievable within a very small class of specific propositions. The problem, as he saw it, was that this narrow set excluded the vast bulk of our beliefs. Standing in direct opposition to his contemporaries, his was a quest not for “truth” but for beliefs that would work for human beings.

Ramsey’s lasting influence was not independent of the context – Cambridge in the 1920s – in which he lived and worked. His father was a mathematician and a senior fellow of Magdalene College; his mother was an influential figure in Cambridge political life, both before and after the split-level success of the campaign for women’s suffrage in 1918 and 1928. In this setting, an evidently brilliant young scholar had potentially direct access to the likes of Keynes, Russell, and Wittgenstein. The doors to their studies were there, waiting to be opened. And when Ramsey walked through them, the history of twentieth-century thought was fundamentally changed...

Ramsey wanted to determine how much of its income a country should “save in order to maximize utility over generations.” The conventional economic view is that the present value of future benefits and costs should be discounted to account for uncertainty about their realization, lack of sympathy with unborn others, and the intrinsic salience of immediate consumption. Ramsey accepted that this might well be the case for an individual, but not for a society. That is to say, he rejected the idea that we should value current wellbeing over that of future generations...

Ramsey recognized the central question of just what – that is to say, whose – utility is supposed to be maximized, even if his formal papers abstracted from it. Minutes from the Cambridge Quintics Society, to which he presented a paper (now lost) on “mathematical economics” in 1927, record Ramsey as asserting that, “in arguing about the welfare of the community as a whole,” we have to assume either that utility is the same for everyone or that averaging utilities provides a fair representation. But, he hastened to add, “this is quite unfair except when all the people are quite well off. … For the rich man, it makes little difference how the last sixpence in his pocket is spent.”
posted by kliuless at 8:55 AM on February 25, 2021 [2 favorites]


Fascinating stuff. I'm a little obsessed discount rates, time preferences, and how they shape society but I've never actually heard of him before.

Fun fact: the real time preference rate as expressed through low risk sovereign borrowing has a carefully constructed data set that goes back to the 11th century and shows a very definite sustained decline. As a society we now have much weaker time preferences and that allows us to contemplate very long term investments. Things like wind turbines and nuclear power plants have all their costs upfront and last a long time and societies with stronger time preferences / higher social time preference rates could not contemplate building them.
posted by atrazine at 9:15 AM on February 25, 2021 [1 favorite]


I would have thought multi-generation cathedral building projects indicated a fairly weak time preference (reasonable sub specie aeternatis).
posted by clew at 10:05 AM on February 25, 2021


Yeah good point I actually did think of that as soon as I posted it. Maybe it doesn't quite work as neatly as I thought it did.
posted by atrazine at 1:06 PM on February 25, 2021


Multi-generational projects: what if the perceived benefit is the employment/prosperity of the people building it, rather than the structure left for posterity?
posted by maxwelton at 1:37 PM on February 25, 2021


I don’t think that can describe the confraternities of the cart, maxwelton — people volunteering to do heavy unpaid labor for the cathedrals when many people lived at subsistence.

I have read that it’s a plausible description of the Egyptian pyramid building, though.
posted by clew at 8:10 PM on February 25, 2021


Maybe the discount rate is only a good description of how societies think about purely temporal pay-offs? So you might build a cathedral as a multi-generational project but not a water mill for profit?
posted by atrazine at 2:44 AM on February 26, 2021


To return a little to the central point, there is thinking around how to treat irreversibility in long term discounting and I think it's related to whether a particular process is ergodic or not.

If it is, then the expected value over time is the same as the expected value over an ensemble. In other words, if I simulate a system 10,000 times for a year it is the same as simulating a system for 10,000 years. This is actually true of many physical systems which it why it has also been used in economics where it is often less true.

If I have a bet with the pay-off structure that 50% of the time, I gain $50 and the other I lose $40 at the end of each year, the expected payoff is positive. This actually is an ergodic process because simulating the outcome for one person 100 consecutive times is the same as simulating the outcome over 100 years for one person.

On the other hand, what about either doubling my money or setting it to zero each year? The ensemble average of this is flat, if sufficiently a large number of people start off with $100 and they each gamble, the average is an outcome of $100. This is not an ergodic process because if instead I personally play the game 100 times, I will end up nothing more than likely. You can't recover from zero, no matter how many doublings you subsequently have.

Processes which are ergodic, like small gains or losses can be modelled using time preference. We might collectively prefer a little more money now rather than infrastructure in 50 years. Processes which are not ergodic cannot be modelled this way because they assume that you can "play through" the zero point. This is also the problem with some seemingly fool-proof gambling "systems" because they end up ruining you at which point you cannot benefit from future gains because you no-longer have any capital. Risks which potentially lead to large scale extinction cannot really be modelled using discount rates at all, it's just not mathematically consistent.
posted by atrazine at 9:09 AM on February 26, 2021 [4 favorites]


I will endeavor to not take thelonius' comments above personally.
posted by wittgenstein at 3:58 PM on February 26, 2021 [1 favorite]


Ha! I probably don't even understand it. No, not probably.

Some comments above suggested that Ramsey was overrated as a mathematician in the piece. But I think the reason he's held in awe is that he was making first-class original contributions to philosophy and economics, too. They gave him the Tractatus to translate, while he was an undergraduate! He did it over winter break. Admittedly, they had long winter breaks. Keynes said he was afraid of sounding dumb when talking to him! And he was evidently a wonderful person to know, too. Good at parties, all that stuff. I first heard of him from a Cambridge educated philosophy professor, so of course I am predisposed a little to the hagiographic take on Ramsey. If a shaft of light had come into the room while he was telling us about Ramsey, I would not have been surprised.
posted by thelonius at 6:22 PM on February 26, 2021


« Older Non-Metro Congress-people mover under the Capitol   |   Black history of science Newer »


This thread has been archived and is closed to new comments