It's a common concept in economic theory: https://en.m.wikipedia.org/wiki/Perfect_competition

It requires a bunch of idealizing conditions that never holds in practice. But it's still useful to study what effects deviations from perfect competition has on efficiency.

Think about it like this: subsidies lets you buy something you don't actually want, that's inefficient.

For subsidies to make sense, they need to compensate some other deviation from perfect competition. For example, it can make sense to subsidize green energy because fossile fuels has external costs not accounted for by the market.

But the fact that a rich individual is willing to pay your air ticket is not a good reason to subsidize something.

An intuitive example: A rich guy enters a bar and yells "free ice cream for everyone!". That's nice. But if the same cash were distributed among the guests, some would've gotten beers and some wine. But when faced with the choice "pay for beer or get free ice cream" everyone choose ice cream despite there being a better allocation of resources.

Thanks for elaborating. It seems now I don't understand your definition of efficiency. People choosing free ice cream over costly beer doesn't strike me as a loss of efficiency.

"Efficiency" is total Utility produced with the available resources.

> People choosing free ice cream over expensive beer doesn't strike me as inefficient.

But it is!

With the total available resources we have two options:

All get beer. Or all get ice cream.

Total utility (value/happiness/whatever) is maximized when everyone gets beer.

When resources are unevenly distributed, ice cream is subsidized, and all get ice cream.

When resources are evenly distributed everyone can make the choice themselves, and they get beer.

Clearly "resources evenly distributed" had the better outcome.

Ah! I understand. I agree that if you do that once, it works one time. But as soon as you take the money they all gave to the bartender and redistribute it, to maximise happiness, and they all buy more beer, you might start to spot where this falls down as a useful model, mightn't you?

I haven't said anything about redistributing wealth.

The point is not that redistribution is a good idea. The point is that wealth inequality makes the economy less efficient, and that it is something we should take into consideration.

Redistributing wealth might also make the economy less efficient, that's an argument that can be made.

From where I stand, it seems the inefficiency caused by wealth inequality is too rarely discussed relative to its harmful impact.

The problems caused by redistribution/taxes etc. are otoh extremely often raised.

Fair point - you didn't mention redistribution.

But I do think your argument of efficiency is not persuasive, as the definition of utility relies on people making choices they never regret, or that don't harm their total lifetime utility.

E.g. some people might want to take that money and save it, and if you come back to the situation in a year's time they now have more money than the people that bought beer. Now there's wealth inequality and they can buy something the others can't. Viewing the original situation in microcosm makes buying beer the "efficient" thing to do, but it also makes the second scenario less "efficient".