Hacker News Reader

Comment by anonymous5133 | Hacker News Reader

yes

Taxing externalities is a better approach than subsidising what we think might be the best path going forward.

And if you can't tax externalities, do you just throw your hands up in the air, and say: 'Welp, I guess coal is here to stay?'

Perfect is the enemy of good enough.

What do you mean "can't"? It's hard for me to picture the government being unable to tax a product. It seems like an irrelevant thought experiment.

Politicians in the United States can't accurately tax coal's externalities in the same way that they can't embark on a campaign of land redistribution. In the current climate, it is political suicide.

Subsidy is more efficient than taxes - based on undergrad economic theory - but it somewhat disregards exactly where the money for the subsidy comes from (at what cost to something else?) so reality will be different.

But I will say one thing for subsidy is it allows startups to enter a high tech space and take them as a funding source, ideally in proportion to how well the tech works.

Cap and trade can work out similarly (like Tesla selling their credits) to fund the R&D.

But the point is you want to do more than just tax the externality, if you can subsidize R&D by startups which fundamentally shifts the curve rather than just trying to move along it.

IF the government happens to choose the right technology to subsidize, then a subsidy is efficient. Suppose they pump a lot of money into wind, and then someone discovers a solar solution that's cheaper. That giant investment will have been wasted. By contrast, a tax lets researchers competing on a level playing field find the right substitute.

Also, a subsidy does nothing to reduce consumption before the subsidized substitute takes effect, which could be decades.

Externalities can be either positive or negative. For example, education has positive externalities, as well as healthcare. Subsidizing and taxing are both methods of dealing with externalities, and in this case subsidizing non-carbon emitting power sources is nearly equivalent to taxing carbon-emitting power sources.

Isn't a subsidy just a negative tax on what's seen as a positive externality? I get the argument for trying to find a 'market-based solution', but at some point a relatively small group of people (government representatives) need to make a decision that will affect the whole market/population. Whether they're taxing one thing or subsidizing something else, a decision is being made.

It is better to tax the negative outcome, because then the choices about what replaces it are not being made directly by the government, and that can reduce costs and to an extent scale up more quickly. But, US politics seems too dysfunctional for a carbon tax, so unfortunately we’re left with deciding whether to let the perfect be the enemy of the good.

Imagine we wanted to influence people to eat burritos instead of tacos. Tacos are currently very popular. The normal cost of a taco is $3 and the normal cost of a burrito is $4. Let's say I have $4.

If we subsidise the burrito by $2, then the burrito will cost $2 and the taco will cost $3. I have $4 so I can buy either of them. They are about the same price (with the burrito being just a little cheaper). However since I'm used to buying tacos, there's a good chance that I won't switch.

If I leave the burrito at $4 and tax the taco an extra $2, then the burrito will cost $4 and the taco will cost $5. I don't have $5 so I'm forced to buy a burrito.

Of course the problem is that in the second scenario everything costs more and the government is pocketing the tax. So while it's arguably going to be more effective at switching people to burritos (because a subset of people will no longer be able to afford tacos), it's easy to market against: "The government is destroying the economy by putting tacos out of reach of the everyday person. Food will be 33% more expensive because of this tax that is designed to funnel money into government coffers. There are 2 million taco makers who will lose their jobs because of this terrible tax. Vote for us and we will return the choice back to consumers."

A subsidy is easier to introduce.

That's valid as long as there are only two choices, but a more accurate analogy would be that we don't really care what people eat as long as they eat fewer tacos. You could subsidize burritos, but that's less effective for people who would have had enchiladas as their second choice. In fact, some portion of people who would have bought enchiladas anyways might end up getting burritos because they're now cheaper, and the government foots the bill for no good reason.

Another factor to consider is that the subsidy must be paid for by somebody, so you end up needing to raise the money through taxes anyway. This means that ultimately you are raising net taxes on everything that's not subsidized.

With all that being said, you're completely right that subsidies are an easier sell. They have an enormous advantage: you can seperate the spending / mechanism of action (which people love) from the taxes that pay for it (which people hate). Taxes on bad things inherently tie the two together, which makes the marketing harder.

Additionally, the benefits of subsidies are concentrated while the costs are diffuse, which makes it easier to build a strong voting block in favour of subsidies.

We know carbon emissions are bad, but we don't know that any particular alternative is the one we should use instead.

Subsidizing various things instead of taxing fossil fuels is a classic example of looking for your keys under the streetlight. A simple, obvious solution that you know is wrong is not a good one.

First, this is incorrect in general. Others will make the points about tax vs subsidy re: externalities.

But the comment also misses one of the main points of subsidies: they aren't simply to offset some externality, they are used to provide efficiently leveraged market availability to products that are at the beginning of an industrial learning curve which will reduce costs over the longer run.

They let us get the tech out of the local minima that it would otherwise be stuck in to move further out on the learning curve so that long-term savings can be had.

Reply on news.ycombinator.com