Fat taxes, food subsidies

Geoff Simmons is right that we oughtn't mess with New Zealand's clean GST system in pursuit of healthy eating initiatives that aren't likely to do much good. But I'll focus on the part where I disagree. He writes:

For the same cost as removing GST, every family could be given $5 for each child to spend on fruit and veges every week. This would make a much more sizeable difference to the food bill of the poor, not to mention their health.

But ultimately there is only so far that the "health-by-stealth" approach can go. Subsidising good food is certainly the most politically acceptable place to start, but it won't do the job alone.

Energy-dense, micronutrient poor food will continue to get relatively cheaper, and so the subsidy bill will have to grow to keep pace. Meanwhile, the health bill for obesity and diabetes will continue to grow.

The only way to arrest this shift is through a whole raft of other measures, the most unpopular of which will be taxing foods on the basis of the energy they contain.

After all, the biggest threat to our health now is no longer smoking, it is that we eat too much. This will no doubt raise even more fervent opposition than my mother faced at my 5th birthday party.

However, such strong actions will be the only way to deal with our biological programming to eat ourselves to death.

Astronomical excise taxes are now normal for cigarettes, and have played a huge part in getting smoking rates down in this country. A similar approach with fatty and sugary food is only a matter of time.

Where to start? First, I still don't get where the market failure is that justifies government intervention in individual diets. Bernard Hickey tweeted a potential one:

Do buyers of cheap fatty/sugary food have perfect information on the long term health costs? Is that info reflected in cost?

Nobody has perfect information about anything. So here's a short list of reasons why the information failure argument fails:

1. Information-based intervention in this kind of consumption behaviour doesn't really make sense unless information problems are greater here than elsewhere;
2. If information is the problem, interventions subsidizing information are a more direct solution;
3. It's unclear that people are choosing to eat tasty fatty things because they're ignorant about long term health consequences; it's at least as plausible that they're just weighing current consumption and making rational decisions trading off health against consumption benefits. This is consistent with the existing literature that providing health information and calorie counts has negligible effects on consumer choice

Geoff seems to be pointing to the fiscal externality argument for intervention. But most of these health-related fiscal externalities are just a transfer; further, the more efficient solution to any technological externality induced by cost-shifting would be setting actuarily fair health insurance premiums.

I'll agree with Simmons that pretty invasive measures would be needed to change behaviour; I'm just failing to see any good (by which I mean market failure) reason for them. I'll also agree with Simmons that such measures may only be a matter of time, though we may have different views on the normative aspects of the prediction.

And everything that Seamus said about GST on food also applies....