[lanfeust6]

My point is that every approach has second-order effects, there's no free lunch. If you pick one approach, then you're dealing with the externalities.

> In the context of all those secondary and tertiary impacts, it seems like a direct tax (like a sugar tax) is preferential

Not to voters. Taxes are unpopular, ending a subsidy to a small powerful cohort would be relatively more popular (in terms of messaging I mean, the end result would still be that consumers pay more for sugar, but of course the govt spending less frees up spending for other things). However, farmer support is right-coded which would lead to opposition by right-wing pundits and media.

It's a toss-up. A tax could be effective, but I don't agree that it's necessarily more viable or palatable. It's probably less-so. Hence I would pitch ending or curbing the subsidy.

[bumby]

We agree that it's always about tradeoffs. I just think there are probably more complex and less transparent (and potentially negative) tradeoffs with ending subsidies if the goal is reducing obesity. It doesn't mean subsidies are good, but just that they are more loosely aligned with obesity than you let on.

I just don't see how it's a more effective strategy given the fact that it's a much more complicated apparatus to do the same thing (raise prices on food). Your position seems to be, stated differently, that higher prices lead to a deterrent to overconsumption and that reducing subsidies is the best way to increase prices. Logically, I can’t find a way that is a better mechanism than affecting prices directly and in a more targeted manner with less tangential effects. It reads to me as a way to find a rationale to go after a particular policy one doesn't like, rather than being focused on the problem at hand (obesity).