I think I read that post at the time. I agree with that, too.

I believe the argument "How can we justify making people who are already hurting pay more?" is really a red herring, designed to elicit an emotional reaction against a proposed policy and keep people focused on the story of imagined people and away from the big picture. As long as the policy in question is a good policy and improves societal outcomes, we should not worry so much about the poor's ability to cope with that particular policy, and worry more about the poor's ability to cope with being poor.

Although you mostly refer to price-based policies in that post, the arguments do not rest on the assumption that the policies are price based, just on the assumption that the policies in question are good policies. My point above was that that this (mostly emotional) argument is much easier to make against a price based policy than it is against a command & control policy.

Which is why we have low gas taxes and jerry-rigged CAFE standards.

Now that you've reminded me of just how weak the "hurting the poor" argument is, I think my comments about hurdle rates should be adjusted. If we need to use a hurdle rate beyond the simple standard "Is there a greater societal benefit from this policy than from our other policy options?" it can probably be a lot lower than 50%.

By the way, there is another interesting aside here. You said "the command-and-control approach is clunky because it treats halogens and CFL/LEDs equivalently, although the energy savings are vastly greater with the latter."

While this is widely believed to be true, the numbers do not support the assertion. I googled "60w equivalent LED" and "60w equivalen CFL" and got a 800 lumen Cree LED using 9.5 W and a 840 Lumen Philips CFL using 13w. Correcting the wattage to adjust for luminosty, we would have a 12.4W, 800 Lumen CFL.

So the energy savings for the LED is 50.5 W, compared to 47.6 for the CFL- only 6% greater. Although it is widely considered to be true, energy savings do not justify replacing CFLs with LEDs. The advantages of LEDs are
- cold temperature performance
- dimmability
- durability, especially in applications with vibration
- no mercury
and (sometimes)
- color (CFLs are vastly better than they used to be, and LEDs are far from perfect.)
- instant on (although this depends on the fixture- I have 2 LED fixtures in my house which are slower than most CFLs)

This is mostly beside the point, but it does argue for any policy designed to reduce energy usage to treat CFLs and LEDs fairly evenhandedly, as the current ban does and as most price signals would. Halogens, of course, are another matter. The ban is clunky in this regard.

(1) Agreed.

(2) The case for command and control when an "uneconomic" decision is in fact an "economic" decision when you include the cost of executive capacity is even stronger, assuming that there are negative externalities involved, as is the case with energy consumption.

Yes, a person may be rationally choosing to not expend limited executive capacity in exchange for long term savings, but the exact same can be said about applying price signals. If we think it is worth forcing the expenditure of executive capacity by applying a price signal, why is that better than forcing the expenditure of executive capacity with a ban?

The light bulb ban has the effect of forcing everyone to expend executive capacity in order to choose a new light bulb for each application (LED, CFL, or halogen?) but this is a one-off cost, borne fairly equally across society, and all of society benefits from the reduced negative externalities. Once the individual is forced to expend executive capacity in this manner, there will be an added benefit of increasing elasticity and making existing or new price signals more effective.

Regarding high discount rates and living paycheck to paycheck, this is a much stronger argument against price signals. In fact, it is the first argument we often hear against most price signals, and it usually goes: "How can we justify making people who are already hurting pay more?"

There should be a hurdle rate for the IRR of investments which we force with command-and-control techniques, but when those investments exceed that hurdle rate, we can honestly say that we are helping the poor by forcing them to make high IRR investments. I would suggest that an IRR hurdle rate of 50% should be sufficient to make this claim, since it is higher than most poor people's cost of funds. Command-and-control techniques may be appropriate in lower IRR cases, but this will need to be decided on a case-by-case basis.

Good points, Ed. I think we're agreed on most of this. A few quibbles below:

1. Agreed on the limited savings of halogens, but my point was that the "pain" from the loss of high wattage incandescent is limited by the availability of halogens. People who would not have switched without the command and control approach are not going to be complaining that halogens do not save "enough" compared to their old bulbs.

2. I'm picking on lighting because you brought it up as an example in your post. So I turn the question back to you: Why did *you* pick on lighting? I'm not picking on CAFE because I agree with you that that is an example of bad command and control.

If you'd like another good example, I will point to building energy codes. Here, an agency problem leads to the builder/landlord not making clearly economic upgrades, so it makes sense to nudge them along with building codes.

3. I disagree. Bulbs are only taxed when they are replaced. Little-used bulbs are seldom replaced, so they are taxed less. Perhaps the tax could be refined to take bulb life into account.

4. I used the term "market failure" when there is a clear economic benefit which the market fails to capture. In the case of an investment in CFLs, a $1-$2 bulb often has an IRR greater than 100% in relatively high usage light sockets. If you would like to call this a "highly inelastic market" or something like that, I'm OK.

I'll rephrase my point: there is a case for command and control in favor (or in conjunction with) price signals in highly inelastic markets. In elastic markets, price signals are clearly better.

Ed,
Yes, nudges are imperfect, too, but you have failed to address the point that a price signal (in terms of electricity prices) would not have worked as well or would have been incredibly painful.

This is especially true because the "pain" in your example is pretty painless or nonexistent.
The light in the closet or crawlspace can be a CFL, which cost less than $2 these days, and light quality does not matter in this context. And you won't have to replace it anyway, since it won't burn out if you don't use it.

You favorite lamp can use a halogen, same with your bedside reading light. http://www.lampsplus.com/products/72w-equivalent-...

I'm with you on CAFE standards, though, but that's because it's an example of regulatory capture and a bad law, and the car market is more efficient. Recall that my point was that command and control is better in cases of *market failure*.

We agree that when the market works, price signals are better.

That said, it might have been possible to design a better price signal to improve lighting than the command and control approach. Perhaps a $0.05/Watt tax on all light bulbs, with the revenue used to reduce other taxes. That would make a 100W incandescent cost $5.25, a 25W CFL cost $2.25. Then you'd still be buying CFLs for your closet, but you might splash out $5.25 for those important reading lights. (The 100W equivalent halogen (72W) would cost about $6.)

So I agree with you that a carefully designed price signal can work well, but not just any price signal. Doubling electricity prices would not have been enough to get much movement on lightbulbs.

I would like to point out that command and control strategies can be less painful than price-based strategies in the case of market failure. Your own "painful" example of banning the incandescent bulb is actually a painless- even profitable- case in point.

Using an incandescent bulb costs $2-$10 dollars a year depending on usage. This caps the benefit to the consumer of adopting more efficient technology at a fraction of that; the market failure arises because a few dollars a year is not sufficient to get most consumers to adopt new technology in the place of perfectly functional technology.

In the case of CFLs, the new technology was also inferior in a number of ways, but the ban led to accelerated innovation to make CFLs better and increase the availability of efficient halogen and LED alternatives. I do not believe a price based strategy such as a carbon tax could have raised electricity prices enough to enable market forces to accomplish these changes without at least a doubling of electricity prices. I think you'll agree that a doubling of electricity prices would have been much more painful than the bulb ban.

Exactly! Thanks.

I agree with Peak Oil Poet (despite the typos and the brusqueness of his comment)- average daily driving distance is very important in considering the cost of fuel as a percentage of disposable income, and, since commute distances generally grew until the early 2000s, this will make recent fuel prices look relatively higher.

Another factor is looking at the percentage of income which is actually disposable income- I doubt this ratio has been fixed, either. I suspect that it will make fuel costs look relatively more expensive in the early years and possibly recently.

None of this is to day that the original comment “The cost to fuel your car has never been higher as a percentage of disposable income” was accurate, only that an actual analysis is even more complex than the one you did.

I know you said not to be offended if you left out *my* favorite econ blogger, but I really think Ed Dolan deserves a mention.

That's very close to the point I'm trying to make. The difference is that I don't agree that standard microeconomic theory is correct. You ask the right question: "If we impose a tax that is equal to all external costs, and all internal costs are captured through the prices the producer pays for inputs, why isn't that the efficient outcome, regardless of whether elasticity is high or low?"

The answer is that consumers often do not have the knowledge or time necessary to respond in an efficient manner to market signals. A very low elasticity is a sign of market inefficiency.

Sticking to the example of the gasoline usage, many drivers drive around for long periods on under-inflated tires. This leads to excess tire wear, reduced road safety, and, more to the point, low gas mileage which could be easily remedied. But they don't properly inflate their tires because they don't know that they are under inflated. A 2011 white paper (http://www.theicct.org/sites/default/files/publications/ICCT_tireefficiency_jun2011.pdf) from the International Council on Clean Transportation states, "Optimal efficiency requires... maintenance of correct tire inflation... However, consumers are inadequately informed about how to adequately improve or maintain the tire efficiency of their vehicles." This lack of information is a classic market failure, and is unlikely to respond to price signals such as a carbon tax.

Suppose you add a carbon tax, and increase the price of driving on under inflated tires. "inadequately informed" drivers are not going to respond by properly inflating their tires- they are simply going to (grumpily) pay the additional tax and continue to drive around on under-inflated tires... because they don't know that there's a problem.

California's CARB mandates that whenever any auto service is performed, tire pressure is also checked, and the tires are inflated to the vehicle manufacturer's recommended pressure. http://www.cncda.org/resources/A_Dealer_Guide_to_...

This is yet one more example of a regulation which leads to a more economically efficient outcome (because it saves fuel at minimal cost) than a price signal such as a carbon tax. This problem is also being addressed by tire pressure monitoring systems which are now standard in many newer car models. Which do you think would be more effective at ensuring the spread of tire pressure monitoring systems to more vehicles: increasing gasoline prices through a carbon tax, or giving auto manufacturers an MPG credit as part of CAFE rules?

I agree with you that a carbon tax will increase economic efficiency, but I don't think that it can achieve optimal economic efficiency alone. A carbon tax is part of the answer, but it needs to go hand in hand with regulation to improve market efficiency.