I'm trying to track down a 19th century French polymath -- engineer, economist, possibly public administrator. Did work on infrastructure, particularly sewers and/or water supply, bridges, and railroads.

I _think_ he also made an observation of the economic logic of multi-class passenger railroad service and amenities of accommodations (intentionally poor lowest-class service).

Likely active ~1840 - 1860.

My Web-fu is gives nil.

And here's Jule Dupuit's argument for why third-class rail service is so poor, in a nutshell:

"It is not because of the several thousand francs which they would have to spend to cover the third class wagons or to upholster the benches. ...

"Its goal is to stop the traveler who can pay for the second class trip from going third class. It hurts the poor not because it wants them to personally suffer, but to scare the rich."


Show thread

@dredmorbius Just like in the UK where the companies also did things like arrange no through 3rd class services to major destinations and hooked them onto freight trains. Then the Midland railway decided to do 3rd class properly and the game was up

@EtchedPixels Right.

My sense is that this is a general case of any network-based service or activity. It may even be an inverse of the Tyranny of the Minimum Viable User, now that I think of it.

If the fundamental capability is connecting nodes, if different service levels can be offered, then customers will tend strongly toward the minimum tolerable option. Trying to force revenues up-market is pushing on a string.

Options are worsened base tier, social signalling, or surcharges.

@EtchedPixels So long as you're working in a market context, given low/zero marginal costs, those are the options.

(By social signalling: pitching higher class service as a social signalling / wealth / merit mechanism, or of lower class service as something shameful.)

Context of this thread is an NYT item on "premium" less-annoying information services.



@dredmorbius Yes - like having an aol.com account in the early days 8) Social signalling is hard in a network though because if you create a 'superior' brand then you need to make it exclusive which also makes it less useful. Easy for tins of beans where you make the cheap own brand look cheap and obvious but harder IMHO for a network.

@EtchedPixels Facebook began as an exclusive network (it was Literally Harvard, a point I made in an earlier, now vanished AFAIU, tootstream).

In growing, Facebook lost its exlusivity.

The Tilly-Odlyzko refutation of Metcalfe's Law posits that network value ~ n(log(n)) rather than n^2. That is, there are higher and lower-value n's. Start with a high-value-n network and a small network delivers value.

Excxlusive golf / social clubs deliver this.

@EtchedPixels If the distribution of value (or, say, just as an example, individual wealth) is great enough, then both net and per-unit value can be hugely improved through exclusivity.

There are also incidental supply-demand elements that come into play by constraining supply and amping demand.

Alumni networks, professional societies, wealth-based networks, limited-neighbourhood clubs, etc., all segement along these lines. The concept is the same.

@EtchedPixels Early technical adopters also frequently self-segregate into wealthier and/or more intellectually capable cohorts.

The telephone, up through about 1950, was effectively an exclusive club. High subscription and per-minute costs limited use to wealthier citizens. This was the downfall of the "Dewey Beats Truman" poll, but for advertisers / marketers / high-end shops & businesses, was a feature.

The is brought about by low-costs, excess access, and spam.

@dredmorbius @EtchedPixels Fascinatingly different from the telegraph which had had nearly the opposite history, with it being so ubiquitous and approachable that folks were wiring up their fences to talk to their local communities.

@endomain That's what happens when you've got a nonexcludable invention.

If all you need is wires and through-transit (and you're interested in local comms), then yes, stringing your own wire (or repurposing already-strung wire) was a viable option.

It's exludability that is a key part of the perverse dynamic.


@dredmorbius I can't think of any billion dollar examples of those. You can build them but you don't get the same scaling. (And my experience of Alumni networks is that they most exist to spam ex students who throw it all in the bin)

@EtchedPixels Something like the World Economic Forum might fit. I don't know that that's a true billion-dollar example.

Various business consortia might also fit.

Though with those, you end up with the conflict that the individual members are also very often in competition with each other.

My sense is that the more exclusive such clubs would tend to be fairly underground.

The DeBeers group, the US House, Senate, or British houses of parliament might also be examples.

@EtchedPixels Gyms / gym memberships are another case. There are highly exlusive gyms, but your local heavy iron box is very nearly as effective. The business model is subscriptions based, but from the gym's perspective, the best thing is for you to _not_ actually visit. Lots of perverse incentives.

It makes the business inherently scammy, and it has been for years. Though exceptions exist. Often as public / community fitness centres.

@dredmorbius you missed deliberate overcomplexity (so its cheaper to pay over the odds than work out what you should pay or how to get the state mandated base package), and artificial risk (eg cheap rail tickets bound to one specific train) -both favourites of airline and rail

@EtchedPixels Good points.

Obfuscating otherwise simple concepts is a common trope in many domains.


@dredmorbius @lrhodes

Sorry, just calling attention since cross-instance hyperlinking is really hit or miss in fedi. This kinda thing struck me as someone else's jam

Sign in to participate in the conversation

Everyone is welcome as long as you follow our code of conduct! Thank you. Mastodon.cloud is maintained by Sujitech, LLC.