4 ms·
It's an important point, that these concerns are always about careerists, not the beer money group. But there are two groups here. The beer money group could b
by brownbat 10y ago
It's an important point, that these concerns are always about careerists, not the beer money group. But there are two groups here.
The beer money group could be significant: retirees, students, people who want extreme flexibility because they have another primary job or responsibility. Half of all drivers is a significant block.
On the other hand, careerists legitimately face an impossible situation, since so much of their life is in the hands of one company that's constantly tinkering with its economics.
The common policy response across similar situations seems to demand every single job must work as a 40 per week career with benefits, killing any pure beer money jobs.
I don't know whether that's a wise policy solution or not. Maybe the benefits do outweigh the costs. But it's certainly not pareto optimal. Restrictive policy responses hurt one group of people in an attempt to protect others.
I'd feel better about such policies if proponents showed awareness of who they were hurting. Like if proponents said, "ok, honestly, we're going to hurt retirees with this jobs protection measure, but we think that's ok, because we value single moms over retirees."
Though if you get that far, you could just reduce it to a wealth transfer. Let Uber do what it wants, then tax retirees and give the money to single moms. Similar effect. One group loses, one group benefits.
Though of course then you'd stop and ask why you're taxing people supplementing a fixed income. Maybe you'd rather tax some other group first.
And that line of analysis is how you end up supporting deregulation combined with aggressively progressive taxation where the benefits are targeted to vulnerable groups.
- tnecniv 10y ago> Restrictive policy responses hurt one group of people in an attempt to protect others. That's the definition of Pareto optimality
- brownbat 10y agoOk, pareto optimal describes a state, and I was abusing the term to apply it to a change of states. Definitely could be ambiguous: Are you moving towards pareto optimality, or starting at PO? When I've heard others commit similar abuses, they use "pareto optimal [change]" or "pareto efficient [change]" as synonyms for "pareto improvement." ie, moving towards pareto optimality. ie, a change that makes someone better and doesn't hurt anyone. But you're right, "pareto improvement" is really the proper term there.
- SilasX 10y agoI think the parent was referring to how "you're at a Pareto optimum when any change has to make someone worse off", and was applying that (incorrectly IMHO) to the situation you described where a regulation makes someone worse off. I say "incorrectly" because "some regulation R makes some group worse off" (as you were saying) does not imply "all reg changes would make someone worse off" (Pareto optimality).