Words of wisdom

Referring to the forthcoming ban on “plain-vanilla 100-watt incandescent bulbs” (California is already there), Virginia Postrel writes:

If they’re really interested in environmental quality, policy makers shouldn’t care how households get to that total. They should just raise the price of electricity, through taxes or higher rates, to discourage using it.

Instead, the law raises the price of light bulbs, but not the price of using them. In fact, its supporters loudly proclaim that the new bulbs will cost less to use. If true, the savings could encourage people to keep the lights on longer.


Raising the price of electricity (or gas) , through taxes also has the effect of reducing the standard of living more for people who work for low wages than it does for people who write columns about economics..

So does banning low cost light bulbs

Not by as much. (if at all)

Actually a lot. (and every little bit counts for the poor)

That must surely depend on by how much the tax is increased?

Poor people will buy low energy light bulbs anyway if the price of electricity is high..

It is true that taxing energy/carbon effects the poor but this is hardly an intractable problem. It would be possible to combine such taxes with reduced payroll taxes or increased EITC such that most poor people would still be better off.

I have no reason to accuse you of anything but I do know that there are conservatives that believe income taxes are punitively progressive but also become suddenly concerned about regressive taxation when it comes from pricing the externalities of pollution/greenhouse gases.

If I believed government would stop taxing with the left hand, to make up for what it starts taxing with the right hand, then I would agree with you. I suspect it's more likely that government will take with both hands and spend over 100% of what it takes.

If I believed government would stop taxing with the left hand, to make up for what it starts taxing with the right hand, then I would agree with you.

This is why it's important that it be precisely a transfer tax.

Fund the expenses of running the program some other way. The tax that is collected goes 100% to voters. Not 100% minus program costs. Once they get to start subtracting things from the transfer, they will eventually take the whole thing.

So, you tax electric companies a very high rate and let them pass the tax on to their customers. Ideally they should pass on the tax equally, but it might cost too much to enforce that. Collecting the tax is reasonably cheap.

Then you divide up the tax money and divide it evenly among voters. Once the system is set up that's pretty cheap too.

100% of the money goes to voters, and they each get the same amount. Try to siphon off some of that money and you are taking it straight out of the pocket of every voter in the country. That ought to delay the government taking it about as well as anything else could.

Transfer money from people who use a lot of electricity to people who use less electricity. If we want people to use less electricity, isn't this the obvious approach?

The studies show consumers just do not rationally switch to CFLs, even in situations where issues like the 'quality' of the light wouldn't matter.

(and studies also show, with the newer CFLs, on blind tests, consumers cannot tell the difference).

It's an old problem in the study of energy conservation, the 'money on the table phenomenon':
- consumers and businesses apply high discount rates to energy savings (around 30-40% pa). Irrationally high for an essentially risk free activity
- people do not react to changes in the prices of small items (like the electricity burned per month by a lightbulb) the way they do large ones
- people overreact to the upfront cost of things eg more efficient refridgerators. That is consistent with models of significant borrowing constraint-- consumers cannot easily borrow to buy more efficient lightbulbs, insulate their houses etc.
- a related point is hyperbolic discounting. Small savings over a long period of time, however large in total (that Starbucks a day costs you over $10k a lifetime) just don't weigh in the human mind the way the cost of a lightbulb does. I *know* my savings, and I build spreadsheet models of them, but it's still a gulp to throw a perfectly good 50 watt halogen lightbulb to replace it with a 4 watt or 7 watt LED costing 10 times as much, even though I have a payback period of less than 2 years. Such is human nature, even for the rationally fully informed
- there are significant costs to obtaining information on energy efficiency and embodying it. An informationally imperfect market (try to figure out which appliance in the UK is most energy efficient, some time). Even now, most people will not believe that replacing a 1985 fridge with a 2010 one can save $200 a year (the former burns 2000 kwhr pa, the latter can burn 550)
- 'split incentive' problems prevail. Landlords don't invest in energy efficiency, because consumers pay the bills. Consumers do not, because they are not planning to rent for that long
- network effects/ externalities are ignored. For the utility, cutting peak electricity demand (5pm on a weekday) has a disproportionate savings in grid costs, generation costs and in fuel CO2. The end user is indifferent to this, although we all suffer from the costs of having to spec the electricity network to handle it.

As ever it is low hanging fruit. Abolishing the 150 and 100 watt sizes will save a lot more energy than abolishing smaller sizes of incandescent lightbulb. At least that is likely.

Although CFLs are only about half of lightbulbs sold (number in the article is 25%?) it's probable that they generate more than half the light out there, because they are used more.

On the Jevons effect ('rebound effect') generally there is no evidence it is 100% for most examples. We don't have 4 times as many fridges, nor do we run them 4 times as many hours in year, just because they use only 25% of the energy they did 30 years ago.

"In fact, its supporters loudly proclaim that the new bulbs will cost less to use. If true, the savings could encourage people to keep the lights on longer. "

Guess we should not go for more fuel-efficient vehicles either ; people would only drive longer.

I agree with RR. This sort of argument is getting increasingly (and annoyingly) common. Let's not make cars safe because people would only drive faster? Let's not mandate helmets because motorcyclists would only drive rasher. Let's not provide insurance because people would only live riskier.

Notwithstanding the Peltzmann effect would people always modify their behavior to totally nullify the original intentions?

It is what happens.

The other alternative is to be the slave of idiots.

The takeback effect is a real phenomenon in energy conservation; see for example this article from Wikipedia. The only disagreement is over the size of the effect. In general, people do tend to drive longer distances when they switch to more fuel-efficient vehicles. The question is, do they drive enough to cancel out the effect of the efficiency increase.

Efficiency compensation does not equal efficiency nullification.

Exactly. Why isn't anyone accounting for the expanded value to consumers here? For one thing, people probably wont change their habits too much - I know that my preference with lights is 100% to do with what I am doing at the time - and little to do with my own financial concerns.

If consumers are leaving lights on longer, it stands to reason that they actually want to have light for longer. Longer trips seem to point to expanded consumer options in where to go. If the same amount of gas is being guzzled, then that is an improvement - despite the fact that more will still need to be done to shift our energy consumption.

What I was thinking. In fact, lets just halt innovation altogether because, after all, people will want to use their new things.

"Guess we should not go for more fuel-efficient vehicles either ; people would only drive longer."
Right. I believe the data show that something like three-quarters of the benefits of increased mileage go toward more driving. So if you increase mileage by, say, 20%, you reduce gas consumption by 5%. (It is also doubtful whether airbags save any lives at all. Motorcycle helmet laws do save lives, but only by making motorcycling less attractive, so fewer people ride.)

Remember when we referred to our country as "the land of the free"? Way back when.


That's the point. There's a difference between developing fuel efficient vehicles people can choose vs. forcing people to buy fuel efficient vehicles. Hence CAFE is felt to be a stupid, inefficient policy that does little for the environment but makes regulators very happy. Once more, dumping the regs and raising the taxes do the job better and more simply.

So, to convince people to save money, and make sure the market supplies money saving products, the price of electricity for all uses, washers, driers, heat pumps, stoves, Energy Star TVs should be taxed to raise the price?

Here is a question for those who think the market works best without government regulation:

Why don't the new technologies flood the market at low prices before the regulations are proposed and set to go into effect?

For example, CFLs are very old technology, long ago replacing incandescent lamps in industrial and commercial markets, with the electronic ballasts being installed in mass conversions in the early 90s. Even at $20 each, CFLs were cost saving over their lifetime, and within a few years for heavily used lamps. Fluorescent lamps have been common in home kitchens for decades for certain well lite design styles. As marketing programs and utility incentives increased sales, the prices dropped under $10 and then to about $5, and now you can find some pretty cheap ones, with a significant variety of color temperatures.


I think the argument was that if you were to choose between the two "evils" of taxation and banning certain products then taxation is the least bad of the two

Candles are even cheaper and they use no electricity, yet few use them as their main light.

Gee whiz, people are so stupid, not saving money by using candles.

For example, CFLs are very old technology, long ago replacing incandescent lamps in industrial and commercial markets, with the electronic ballasts being installed in mass conversions in the early 90s.

That's not a compact flourescent. External ballasts and separately replaceable tubes are indeed old technology - but they're not made to screw in to an existing incandescent-designed fixture.

And they likewise don't have the troubling reliability issues that CFLs still have.

(And to answer your question - why should anyone have moved to CFLs at home? Most people spend very, very little on lighting power, as far as I can tell - and it's only very recently that CFLs have been able to produce light that isn't, well, ugly.

Why would they want to do that?

Then again, how many 100 watt bulbs do most houses have? I'm not sure I have any...)

On the main topic: First thing that happens after a ban on 100 watt incandescents?

The rise of the 95 watt incandescent.

CFL's are the worst ****ing thing ever. I mean just absolutely awful. The light quality is absolutely horrible- about on par with a warehouse or hospital waiting room- and I haven't noticed anything that resembles "cost savings" since all the ones I have had burnt out within a couple months. That last bit is likely poor luck, but even if they lasted as long as they were meant too they still look awful.

Taxes are the proper solution. The obvious solution to anyone with any knowledge of economics at all. All of the nanny state sort of liberals (mulp etc.) need to move to one state so they can focus hard on ruining it instead of forcing everyone else to live with their nonsense.

Ah, but most of the fun is in ruining things for other people.

Skip -- Most of us sensible liberals believe in raising the taxes on the things we think need to be controlled, like motor fuels and electricity, because the marketplace is, in fact, a very effective mechanism. Unfortunately, the knee-jerk "no tax" Tea Partiers (not a sensible person such as yourself, of course) shout NO TAXES so loudly that our legislators are shocked into submission, and the knee-jerk NO TAX Republicans block any such legislation. So nothing ever gets done on that front.

Just think, if we added a few cents to the gasoline tax, with the proceeds dedicated to highway maintenance, we could help alleviate several big problems at once. If we added a few cents to the electricity tax, we could help alleviate several big problems, too. So why don't you join me in trying to promote that solution.

the things we think need to be controlled

You probably have no idea how grating, how silly, how hubristic that sounds to the ears of many 'round these parts.

You're right, I have no idea.

I choose words carefully. I don't say "the things that need to be controlled." Instead, I say "the things we THINK need to be controlled." I freely admit that others may disagree with me, and I suggest that the marketplace is a good resolver of those types of things.

If you find it grating that I disagree with you, and you think that I'm filled with hubris because I disagree with you, then perhaps you have a problem that goes deeper than my carefully chosen words.

Do you really think that taxes have gone down because of the TP crowd? Absurd, ignores the massive graft that the tax-phobic rich have engaged in for decades.

Focus on government taxes are just an obfuscation of the issues. Otherwise, how can you explain why the rate of exploitation on the company side is ignored?

Can you re-phrase that in English?

I mean, I grasp that this is somehow a complaint about them evil "corporations" (at least, that's all I can assume company can refer to), and something about the horrid and abominable "rich" who, unlike everyone else, don't actually like taxes (those monsters!).

But could you say it, plainly and by making your assumptions explicit, so I don't have to read things in that might not be accurate?

Say, to explicate the nature of this posited "exploitation"? And this "massive graft"?

I am not an economist but my limited of experience of human behaviour tells me that Ms. Postrel is not correct. There is a disconnect between using electricity and paying for it so when the bill comes around at the end of the month it is quite hard for the average customer to make the connection between leaving that upstairs light on three weeks ago and a extra few cents on the bill.

On the other hand when that same customer goes into a shop to buy a new light bulb they will immediately notice if light bulb A is cheaper than light bulb B.

Don't you think you are overestimating the elasticity of demand with respect to price? Without smart meters the connection between electricity bills and light switches is simply too remote for most people to bother with. Lights get left on because of this disconnect. Also, raising the price of electricity to the point where lighting becomes something you need to worry should the sake of your household budget would make many other activities (like watching tv or running complex statistical algorithms) off-the-scale expensive. Surely better to make lighting more efficient?

Nik -- (1) No, I don't think the elasticity of demand is being overestimated. If anything, I think it may be underestimated. Look what happens to travel usage when gasoline goes up.

I think MANY consumers are aware of electric usage, but simply ignore it because it's so cheap. I think the "disconnect" you mention is definitely in play, but not because of remoteness; rather, I think it's simply because of irrelevance at today's low prices.

(2) The cost of lighting is substantial with incandescant bulbs because of their inefficiency. I think raising the cost of electricity would encourage BOTH (a) purchase of more efficient appliances (including lights) AND (b) more efficient (i.e., less wasteful) use of them.

The other day I saw a politician justify the high price of the new light bulbs. According to his theory, we should all be willing to pay more if it reduces the overall cost of electricity for everyone.


Don't know how to break it to you, but if I pay more I pay more.

Raising the price of electricity penalizes the poor. The rich will keep on doing whatever they like to do, and they use more electricity.

So does raising the price of incandescent light bulbs.

The only difference is you think you know better than the poor what light bulbs they should be using.

Prices--all of them--penalize the poor. The poor are disproportionately minorities. Prices are racist. Stop the hate!

"Prices–all of them–penalize the poor."

Yep, I would even go as far as saying that this is the very definition of being poor.

The point not mentioned so far, is that current electricity costs under price externalities. The taxes should be exactly proportional to the direct and indirect cost of the externalities, as best you can measure them, and let the economic processes work out the hows.

This is a really dumb comment from an economics perspective: raising ALL electricity prices in general, RATHER than effectively raising the price of the more ELASTIC segment of energy usage, lighting.

First, recognize that a bulb AND electricity produce light. You can raise the price of the more elastic lighting segment--electrical consumption for lighting--by RAISING the effective price of wattage used in lights OR by RAISING the price of the component which makes the finished product--light. A light bulb AND electricity produce light. If you either RAISE the price of the bulb OR mandate the use of an alternative bulb which consumes less electricity than a standard bulb, you end up in the same place, or maybe a better place. Why? Because there are economies of scale to bulb production--driving costs lower for their production--and there is an incentive now to innovate in making a more efficient bulb that you achieve without raising electricity prices in general. If you want to use a market signal, you should price or tax bulbs based on their energy efficiency.

Bill, I didn't know you knew any economics.

Still don't know, actually.

I'll ignore your judgmental ad hominem other than to say I've taught a graduate MBA course in pricing, lecture at MBA exec ed programs on the same subject, and also teach antitrust in law school and am an antitrust lawyer in private practice.

Bill, but you are ignoring the fact that electricity is also used for many other things than to light bulbs, so therefore you see you may get to the same place in terms of electricity used on lighting but that is not really the whole story now is it?

Emil, The post and the article cited was about lightbulbs. Of course one could go many places concerning end uses that are more or less elastic, but sometimes it is easier to stay on point--which is to say, stay with the post that was on lightbulbs..

The post quoted the article saying:
"If they’re really interested in environmental quality, policy makers shouldn’t care how households get to that total. They should just raise the price of electricity, through taxes or higher rates, to discourage using it."

This is exactly the same argument I am making so I can't see how I was off-topic.

Bill, in trying to decide which input to tax, you want to get at the one that will raise marginal cost. Here you have a fixed cost component, bulbs, and a marginal component, electricity, in the production of lighting. In order to reach the stated policy goal you don't necessarily want to tax the less elastic input if raising it's price has little to no effect on the cost of an additional unit of lighting, which would be achieved by taxing electricity, not bulbs.

NickL, Lightbulbs have an expected useful life. Lightbulbs depreciate with usage over time. The fixed and variable distinction you are making is irrelevant because the issue is total electrical usage over the lifetime of the bulb. You can make any cost fixed or variable based on the time period you choose.

Again, the issue is on total electricity consumption, not electricity consumption over the lifetime of the bulb but all electricity consumption (over the lifetime of several bulbs as well as for all other means)

(The rest of your statement is pretty much correct IMO though)

So, Emil, let me ask this question:

Did you sign up for that 3 year phone plan that gave you a "free" phone with the $120 month plan at 40c a minute?

Bill, that is an irrelevant question since what I am arguing with you about is not the time value of money issue but about the assumption you are making that we are only looking at the energy consumption from lighting.

While sympathetic to the argument, it isn't that good of a one. Electricity is used for many purposes and lighting is a lesser one. The price rise to encourage adoption of greater efficiency in lighting would be large in other areas and would swamp what could be achieved there leading to higher costs but little increased efficiency, instead it would be viewed as a cost of living. Targeting lighting can promote development and adoption of specific technology that raises efficiency efficiently.

Lord, I think you have just highlighted why it is stupid to target lighting when you say: "Electricity is used for many purposes and lighting is a lesser one."

For the (original) intention is not to recude lighting but to reduce energy consumption. Lighting has been chosen as a means to that end.

15% of energy used in homes is for lighting. Not a small amount.

"Electricity" not "energy".

But 100% of the electricity comes from the electric bill.

If it costs too much for the poor, use less.

Don't forget that incandescents generate waste heat.

This is an added cost in the summer if you have A/C, but saves money on heating in the winter.

(Also, EIA says 8.8% for lighting, not 15% - and that's marginally less than the expenditure for water heating. [EIA has newer data, but I can't find a comparable chart for the 2005 RECS survey, and it's harder to puzzle out - and I doubt it's changed for the WORSE...]

The other question, then, is what percentage of that is due to 100 watt or larger bulbs? I'm not sure I have even a single 100 watt incandescent in my home, myself - but lots of 40s and 60s.)

If you want economic efficiency, you ascribe a large cost to being deprived of choice so taxes are more economically efficient, but if what you want is energy efficiency, depriving people of choice is more energy efficient. Lighting is the cheapest way to save energy and one of the largest because even though lighting is not that large in itself, it offers the greatest efficiency improvements. It takes high taxes to force people to do what they do not wish to do, or even are willing to do but are indifferent to. To even approach doing this by taxes would require both high taxes and some means of rebating some of them back in a way that the taxes do not suppress activity but do suppress consumption. If you want to preserve choice than taxes are the way to go, but don't fool yourself that that is the most energy efficient, it is not. Depriving people of choice is.

"Lighting is the cheapest way to save energy and one of the largest because even though lighting is not that large in itself, it offers the greatest efficiency improvements."

Really? How do we know that it is so much more than for heating/cooling, cooking, food storage, laundering, watching TV, grid efficiency or anything else? Even if we were to know that that is the case today, how can we know that it will be the case tomorrow? What if someone comes up with a great idea for saving on something else?

For some reason, Ms. Postrel has written an infinitely higher proportion of columns (1/0) about how the government shouldn't ban incandescents (I agree with all her points) than she has about the need to incorporate the externalities of carbon production into the cost itself.

No one thinks banning incandescents is the perfect policy. But the reason it happens is that she and her ilk have at every turn done their damnedest to undermine correct policy, meaning we now have to turn to second-best or third-best policy measures.

That makes what she's doing now trolling, not informed discussion.

There is a big literature on "rebound effects" for more energy efficiency, meaning lower marginal cost increases use. For vehicles, the most recent estimates are 2-12%. Kenneth Small and Kurt Van Dender, "The Effect of Improved Fuel Economy on Vehicle Miles Traveled: Estimating the Rebound Effect Using U.S. State Data, 1966-2001"

Mulp has a good point. It is a continuing puzzle as to why people don't buy more energy efficient stuff, even if the payback period is short.

The biggest rebound effects aren't on light, they're on heat and cooling, closer to 30% But Postrel's point is ALWAYS left off of cost-benefit analysis by partisans, with a concomitant overstatement of net benefits. Also uniformly omitted is a utility offset for poorer quality light

Why do you care if someone makes a choice different from the choice you would make?

It is a continuing puzzle as to why people don’t buy more energy efficient stuff, even if the payback period is short.

It really isn't a puzzle at all. Value is subjective. The entirety of the argument for banning incandescent bulbs is that the people passing the ban have imposed their subjective value on everyone because they can. This, of course, is nothing new- it is how government works, but there it is. To give an example, you may as well be puzzled by someone who doesn't buy a Honda Civic instead of an Escalade, or that they shop at Whole Foods rather than Walmart. Are you puzzled, too, in those cases?

Yes, if you ask me for $1 today and guarantee me $1 million tomorrow and I say no to that offer, I find that puzzling.

You don't, of course, because value is subjective. Keen insight. But to pretty much anyone who has studied this problem. Clearly, you are not part of that set of people.

See, you clearly don't get it. You are making an assumption in the case of the light bulbs, one that is totally unsupported by any real evidence, just your own opinion. Care to take a guess what your assumption is, or should I outline it for you (though I should not have to since several other commenters have already hinted it for you).

i imagine if anyone is like me, who needs high wattage for some dark areas & doesnt mind the heat (northeast ohio) they've already laid a supply of the 100 watt bulbs in...


If you heat with electricity that is rational-- sort of. You are heating your ceiling though, the least efficient place in the room.

If you use a heat pump, or you heat with gas, then the cost of generating a kwhr of heat for you is roughly 1/4 of what it is via electric induction (aka a lightbulb). That's because natural gas per unit energy is about 1/4 the price of electricity in most retail markets in North America. And the coefficient of performance of air sourced heat pumps (down to the low 0s F temperature) is around 3 times, and ground sourced around 4 times.

Put it another way, for electricity from the plug you are paying around 10-12 cents/ kwhr in the midwest. For natural gas, assuming 90% furnace efficiency, around 3-3.5 cents/ kwhr.

And of course in Ohio you use air conditioning? Then you are paying twice for excess heat: once to generate it, once to get rid of it.

Now to the light point. I've been very happy with LED as halogen substitutes-- 7 watt for 50 watt. That gives an indistinguishable point light (you can get a 40 degrees of angle one, at least in Europe) for dark spaces. They are expensive, but at my electricity price (twice yours) payback is about 18 months.

On CFLs, the very high end CFLs are not bad, but not, admittedly, as bright as a 150 watt lightbulb. On the 'soft' CFLs (ie 2700k light, which is close to an ordinary incandescent, 3000K is 'daylight' and is harsher) I oversize one size over the recommended equivalent. Seems to work pretty well.

One of the things that amuses me about economists and people interested in economics is how passionately concerned they become about light bulbs.

"If they’re really interested in environmental quality, policy makers shouldn’t care how households get to that total."

Precisely so.

It's almost as if their true interest is putting on theater that makes gullible people feel good, and reminding businesses that they can ban any product at any time for any reason so they better keep those contributions flowing.

It is rarely difficult to find large discrepancies between their stated motives and their actions. See also:
"If the TSA was really interested in keeping us safe..."
"If the Government really thought Global Warming was a pending catastrophe...."

The legal aspects are interesting. Do the Feds use the Commerce Clause? Where does the Fed derive its power to mandate efficiency standards and bans.

I don't understand the reasoning in the defense-of-lightbulb-regulations in the comments. Higher energy prices (whether electricity or gasoline) might disproportionately impact the poor, but an electricity/gas tax could be revenue neutral, with taxes on low-income people decreased in lock-step with the revenue from the energy tax. Could also use the revenue to expand EITC, or to give low-income folks permanent payroll tax holidays.

Average people respond to changes in energy prices. The structure of demand for vehicles changed quite a bit from 2008 onward. If we used increasing energy taxes to guarantee that energy prices would be higher in a year than they are today, and even higher in five years, people could make durable goods purchases with that in mind.

OK, I can't resist it. How many economists does it take to screw in a light bulb?
Answer: it depends on the type of bulb. Apparently, energy efficient bulbs take more economists. This should be good for employment (not to mention the external benefits of keeping them out of things like policy analysis).

Misplaced environmental efforts are probably not the most pressing issue. "Small steps toward a much better world" indeed.

This is a different Bill than the post at 8:47

In fact, its supporters loudly proclaim that the new bulbs will cost less to use. If true, the savings could encourage people to keep the lights on longer.

Is that supposed to be a bad thing?

If people can afford more light and they choose to have more light, then I say that's just fine.

Similarly, if cars are safer and people choose to drive faster, paying more for gasoline and accepting the same risk of injury or death, aren't they still better off? Before they couldn't afford to drive fast and spend less of their time driving. Now they can. They are better off.

If you make life better for people and they don't choose to do what you want, still you have made their lives better. Who should complain?

An energy tax at first sight would make people's lives worse. The poorer they are, the less they can afford the energy tax. But it doesn't have to be that way!

Instead, give every voter a debit card like the Welfare people currently get. Collect a high energy tax, and every week you divide up the receipts evenly and put the same amount in everybody's debit card.

Then on average voters lose nothing. The average voter can use his debit card to pay the electricity tax and come out even. But voters who use more electricity than average pay a lot for it, while voters who use less electricity than average get a surplus.

This way everybody gets a real incentive to conserve electricity, and it doesn't hurt poor people, it hurts people who use a lot of electricity.

Businesses should not get any of their electricity tax back. They can pass it along to their customers. Anything they do to reduce electricity use is a plus for them, they can reduce prices and get more business, or keep the savings as profit. Their customers will factor in electricity use by the business as part of their buying choice -- things that use a lot of electricity will be automatically more expensive.

If we want people to conserve electricity, then make a transfer tax from people who use more electricity to people who use less. Simple and direct. The more you use indirect measures that might have some tendency to get the result you want, the more side effects you can expect and the less the effect you want.


The Wikepedia article mentions that the Jevons Paradox, as stated by Jevons, related to industrial usage. What is overlooked in the rest of the Wikipedia article, and in most of the yada-yada-yada in the above comments, is that energy usage is subject to very different price-demand functions dependent on whether it's industrial, commercial, or residential.

In an industrial situation, the energy is one of the costs going into a "black box" whose output is PRODUCT. The sale of the product, in turn, produces PROFIT All else is irrelevant. So in an industrial situation, if a slightly lower unit cost of product can move the demand for that product further to the right, it makes perfect sense for the industry to raise their TOTAL cost to produce their products, since the lowered UNIT cost will result in equal UNIT profit at the new lower sale price, and the higher demand at that new lower price will raise TOTAL profit. Hence Jevons Paradox.

In a commercial situation the model is totally different. The energy consumption is likely close to a FIXED input. A store wouldn't be expected to add or subtract lights simply because electricity got cheaper, nor would an office. Nor would they be expected to significantly modify their operating hours because of the cost of the lighting bill. So their electricity demand is essentially FIXED.

Residential situations are nothing like either of the above two, and Jevons Paradox is mostly irrelevant. The crucial factor is trying to quantify the relationship between price and demand, which is certainly not linear, but is probably monotone.

Thanks for your comments, Ken. Except the yada-yada part :)

Has nobody mentioned the possibility of an industry lobbying group being solely responsible for this outcome? I assume that's how these things usually happen. Follow the money

You didn't read the column, did you?

Here's a good backgrounder: http://www.nytimes.com/2011/06/05/magazine/bulb-in-bulb-out.html

To follow up what I just wrote, when I was in college in California, there was a "stem cell research" proposition that passed with something like 70% popular vote primarily because it was presented to the public as "abortion vs life" instead of as a mutli-million-dollar grant to biotech companies.

If you want to see the stupidity of the initial post, and why increasing electricity prices in general to reduce electricity consumption for lightbulbs, think of the following example:

1. Assume the electrical utility were the only person who could sell both electricity and lightbulbs.
2. The electrical utility sold a magic lightbulb that had a lifetime hours limit of, say, 1000 hours, and the magic lightbulb communicated to the utility whenever it was used .
3. The electrical utility sold the lightbulb and provided electricity for the magic lightbulb for "free". Well, not really free, because the magic lightbulbs price included the price of electricity to operate it. (Whenever the magic lightbulb was turned on, it communicated to the utility that it was on, and that amount of electricity was deducted from the homeowners electrical bill.
4. Now, the utility offers you two lightbulbs equal in light quality, one an efficient lightbulb for $100 and an inefficient lightbulb that uses more electricity for $400.
5. Which do you choose? Da. The $100 lightbulb.

But, then, ask yourself, why do you currently choose to buy a cheaper inefficient lightbulb today (which has listed efficiency ratings on its packaging) rather than a slightly more expensive lightbulb that is more efficient and less costly over its expected useful life?

Because consumers are not rational and they do not discount to the future very well. That's another reason we prohibit things and just do not always use a market. We rationally choose to constrain ourselves with laws, just like Odesseyus facing the Sirens.

Bill, your points may be valid, but your first sentence compromises your subsequent points. When you grow up, perhaps you will say "If you would like to see what's wrong with the logic of the original post..."

Your comment is valid regarding my characterization, but that is how I felt: that in an econ blog one would not expect to find such a "[you insert word here]" article for discussion.

Bill, the problem that apparently vexes you is that there are a large number of folks, including some pretty strong economists, who generally believe the simple approach: "If your objective is to promote conservation of XYZ, then raise the price of XYZ."

The fact that situations have complex social ramifications does not make those economists stupid, or [insert your own word], for continuing to believe that simple approach.

"If your objective ("A") is to promote conservation of XYZ, then ("B") raise the price of XYZ".

I think that is probably not so controversial if A _is_ your objective; your complete objective, not just one among many things you seek.
But some people in this discussion are raising secondary metrics (some A') in addition to this A, and suggesting B may affect these other goals in messy, possibly even adverse ways. (Example of such A': we would like not to hurt the poor; and the plausibility argument why B is detrimental to A': there are larger and less elastic uses of electricity that are important to the poor, and for which there is no clearly obvious technology (as there is in the case of lighting) they can substitute to maintain their well-being).

There are various intellectually viable responses to someone raising such secondary objectives e.g.: denying that A' is/should be part of one's goal at all; arguing why B affects A' minimally or positively anyway; protesting that you are just considering a toy problem for a text book you are writing so you'll hand-wave A' away for simplicity, and so forth.

But if someone claimed that A' was _inherently_ irrelevant; that "if you want A then you should do B" can be simply detached from other goals such as A', and then cited as a guide to real world policy ... IMO such a principle would be just obviously wrong. I'd be shocked if any pretty strong economist would or does sign up for it.

I don't buy that at all as coming from economists. The economists I work with and retain as consultants would all focus on end use elasticity markets.
Give me a citation to an economist who would say that if you have an inelastic end use market and a elastic end use market for the same input product, and you were trying to maximize the reduction of consumption, you would raise the price of the input (here electricity) rather than tax the input at the elastic segment.

I can certainly understand, however, if your goal is to raise revenue, you would tax the inelastic end use, but you sure wouldn't see much reduction in consumption.

Call your bluff.

Bill, again: you are building a straw man, the whole heart of the argument of the original post and the article it linked to was that if we want to reduce electricity consumption than we should not look at the specific input for a (small) portion of the electricity consumption but at that part that is shared among all electricity consumption.


I read the article. Here is the quote from the article posted here: "If they’re really interested in environmental quality, policy makers shouldn’t care how households get to that total. They should just raise the price of electricity, through taxes or higher rates, to discourage using it."

For the reasons discussed above, if there are elastic and inelastic end use markets, raising the price of electricity in the inelastic market is inefficient. Might be good for me if I'm a utility, or the government, though as a source of revenue, but not as a basis of reducing consumption.

Ok, so at least we have agreed that we are talking about the total energy/electricity spend and not about only the lighting?

Correction - we choose to constrain other people with laws. If you were only constaining yourself no one would have a problem with it.

There's certainly a legitimate argument that a energy or carbon tax would be a significantly better approach from a economic-efficiency standpoint, and even a cap and trade approach. It would also be nice if we could all light our houses with glowing magic ponies. (Except for the glowing magic pony dung, I suppose.) In California passing a electricity tax would require a 2/3 majority and hence is effectively impossible; in the Senate we saw how well cap-and-trade did. It's not the people on the environmental side of this debate who have created such hysteria over taxes that the word can barely be uttered in a political discussion.

California has already approved a cap-and-trade mechanism in Assembly Bill 32. Under California's regulatory regime these costs will flow directly through to utility customers who will see a clear price signal in their electricity rates.


- the light's not the same, the bulbs are not the same; I will still prefer old lightbulbs on my stairway and ancient lights even if it is more expensive
- the "price" is not the same for all the lightbulbs; there is no point in buing efficient lightbulbs when you put it to some infrequently accessed place
- the $100 now and nothing tomorrow vs. $100 tomorrow and nothing today...umm... seems to me you don't discount future?

"Because consumers are not rational and they do not discount to the future very well. That’s another reason we prohibit things and just do not always use a market. We rationally choose to constrain ourselves with laws, just like Odesseyus facing the Sirens."

Because everybody is stupid and I'm an airplane....

Andy, 1. Regarding light quality, that is a point 2. Regarding your ancient building and infrequently used space, that's a point also, but do you mind if the ancient bulb were taxed so that we don't see your consumption create an externality for others? 3. Don't understand the point. What I was saying with the free example was that the full cost of electricity was included in the bulb, and by full cost I would assume a stream of payments reduced to present value. 4. Ever see the movie "Airplane" by the Zucker brothers (great comedic movie). Yes, unfortunatley, we are in the same airplane, which means we should not pass gas or belch, vomit or wheeze if we can avoid it. We are also on the same planet.

It is maddening to hear that banning light bulbs is less efficient than taxing CO2 in a right-of-center complaint.

Yes, we should tax CO2, but right-of-center fixation on lower taxes creates this kind of inefficiency. And it will just get worse and worse. A no new taxes policy will generate thousands of new regulations.

> Instead, the law raises the price of light bulbs, but not the price of using them. In fact, its supporters loudly proclaim that the new bulbs will cost less to use. If true, the savings could encourage people to keep the lights on longer.

Well, yes. But if there were a consensus that consumers were_that_ sensitive to the cost of their lighting, and aware of their technological options, and totally rational even as to what for many households is a fairly small expense, I suspect the whole question would likely not even arise beyond the fringes. We would trust the market.

Are you suggesting that because a proponent of the new regulations tries to reassure the public that (as he sees it) they would not be worse off, indeed would be better off - that in saying so this proponent is thereby conceding the point that people are indeed very price sensitive in this arena, that indeed the market would get us to an optimum (of whatever) anyway? I hope not, that would be a real cheap shot.

The law caused the invention of incandescent bulbs that are 10% more efficient, thus legal. This is a very clear example of a market failure being fixed by government intervention.

There is no clear evidence that there was a market failure before.

Let's discount all the arguments that incandescent bulbs are better. You could for example put two fluorescent bulbs in places you wanted more light and still use less electricity, and there are various fluorescent bulbs for different spectra, etc. That doesn't matter. The customer is still right. If people want inefficient lighting that costs more and gives no benefit whatsoever, that's their right.

Similarly, people have the right to prefer strychnine in their acid. They have the right to prefer women from stripper bars who are more likely to carry STDs to less-expensive just-as-skilled women they meet at church. Even if smoking high-tar high-additive unfiltered cigarettes really does seriously reduce their lifespan, they have the right to choose to die younger. People have the right to play slot machines at casinos that they know return 40% of their money. They have the right to play pool with people they know are pool sharks, and play poker with professional poker players.

People have the right to throw away their money.

It isn't a market failure when people throw away their money. It's just customers making the choices they prefer.

If marketers find ways to hypnotise large numbers of consumers into paying large sums for garbage which does them no good whatsoever, that is not a market failure. Customers who lack resistance to marketing are still customers who have the right to make their own free choices.

Imagine that there were rich people who wanted exotic meals, who wanted to eat eagles. Imagine that they were willing to pay enough for eagle meat that eagles were threatened with extinction. Imagine that these rich people did not mind this, that in fact the man who ate the last eagle felt like he got his money's worth. That is not a market failure. That is a market acting as it should. People supply a product to people who want it. In an economic sense is it any worse to drive a species extinct than to make a limited print run of an artwork so that people will pay more for it?

Admittedly, people usually define market failure in terms of there being another way that's better on average for everybody. But there's no way to measure that in practice. The light bulb case is an example -- people say they prefer to pay for the inefficient lights, and who's to say they're wrong? If we let the last breeding pair of eagles hatch out three eaglets then three more people get to eat eagles before they're all gone. But how can you measure the satisfaction they get from eating the last three eagles next year, versus mine from eating the last eagle today? It isn't comparable.

Imagine that we are approaching a crisis, and if we don't act now the crisis will hit us hard later. And imagine that consumers choose to ignore the crisis and get hit hard later. And producers choose to ignore the crisis and get their profits in each quarter. Isn't that their choice? It isn't a market failure if people prefer to live rich now and suffer a calamity later. They get what they want. And the ones who die before the crisis, win. They get all the benefits and none of the costs.

There simply was not a large market for energy-efficient light bulbs. People preferred to use light bulbs that made their air conditioning work harder, that cost them. We could have had more efficient incandescent bulbs years ago, but the market wasn't there. How is that a market failure? If people prefer not to think about it, if they prefer to be poor than to think, isn't that their choice?

The real issue is goals that fall outside economics.

Like, suppose there are things we have to do or humanity itself will go extinct. And suppose that the mass of consumers is not willing to do those things. I don't want humanity to go extinct, but I can't afford to pay for survival all by myself. From sheer economics, that's just fine. People get what they want, that they can pay for. I want more than I can afford, my tough luck. An economy which results in extinction succeeds to the extent it provides customers with what they want while there are still customers.

I don't have an economic justification why I should get what I want instead of other people who have more money getting what they want. Any justification to avoid extinction has to come from outside.

Maybe I've stated this too strongly, but there's some truth in it.

Keep Bill out of your MBA classroom. He doesn't understand basic economics and has a Dunning-Kreuger problem on the subject.

1) The objective is to maximize economic efficiency, not to minimize energy consumption (as Bill keeps asserting erroneously). That's the whole point of externality taxes--we tax at the marginal social cost and let people reveal their private consumption preferences so they now "internalize" the externality. It would be economically inefficient to have people reducing their electricity usage by some arbitrary amount (or "as much as possible").

2) The externality at issue is some form of air pollution, so even if you accepted Bill's erroneous idea that cutting back as much as possible is desired, it's air pollution, not electricity consumption, we would want to cut as much as possible. People whose incremental power comes from dams or nukes, for example, shouldn't then be cutting back at all.

3) Even if Bill's erroneous claim about minimizing consumption were true AND it were true that the marginal watt came from burning coal, it's not smart to force people to economize in one area. Burning the lights for fewer hours, cutting down on computer or TV use, using a clothesline, etc. might well be more sensible ways to achieve a given consumption reduction for a particular household.

4) The behavioral claim about customer irrationality in bulb choice is supported by exactly zero evidence. We know that the alternatives to traditional bulbs are some combination of a) more costly, b) ugly, c) undimmable, d) mildly toxic, and e) not as durable as advertised. Before I start calling people stupid or irrational (even Bill) I would want to rule out those other objections.


1. You mistated the goal: it is energy efficiency and reduction of carbon, and minimizing energy consumption. If your goal is to reduce the externality, it is also to reduce the consumption of the end use market which is the most elastic.

2. You are wrong on this point as well. If you reduce the demand from light usage from a nuclear plant it can than be sold to displace energy from a coal plant. Electricity is electricity and grids are connected throughout most of the US.

3. You are wrong also on number three. No one is forcing. What we are doing is making sure you see the full cost of what you are buying when you buy an incandescent bulb, and you are paying for it.

4. You are also wrong on number 4 as well. (Quite a record). Here is the evidence that consumers are irrational when they purchase a combination of two items in different combinations at the same price. (Think lightbulb that consumes more energy and you purchase the bulb because it is cheap:

It is from Dan Ariely's website, and it involves "Free Shipping" . Enjoy: "In one of our projects, Kristina Shampanier, Nina Mazar, and I examined whether our reaction to Free! is just a rational reaction to a low price (a very low price) or if it is an irrational overreaction to Free! We carried out a set of experiments in which we measured not only what people chose but also what they gave up in the process. By doing so we were able to show that Free! can tempt us so much that we are willing to forgo a really good deal for a mediocre one simply because it is Free! For now let’s skip the experiments and consider the following thought-experiment:

Consider how long you would be willing to stand in line for a free Ben & Jerry’s ice cream cone. Let’s assume that your answer is 20 minutes and that the cost of a Ben & Jerry’s ice cream cone is $1.45. Now answer this: would you be willing to stand in line for 20 minutes for $1.45 in cash? No way.

This is exactly what the experiments showed; when something is Free! we get excited and as a consequence we are willing to give up better deals — not to mention our time, money, etc.

For the original paper see this link.

For a story about this in the New Yorker see this link.

"it is energy efficiency and reduction of carbon, and minimizing energy consumption"

Right, energy consumption, not energy consumption from lighting, as I've been saying all along. Ergo, you are now moving the goal post.

"Electricity is electricity and grids are connected throughout most of the US."

1) so again, what the electricity is used for doesn't matter, what matters is the total consumption
2) having said 1), it is true that we do care about how that electricity is consumed, we do care about how it is produced ...

"No one is forcing."

excuse me, how can banning the sale of a product not be forcing?

"Here is the evidence that consumers are irrational when they purchase a combination of two items in different combinations at the same price"

But I though we had agreed that it was about total energy consumption, not about energy consumed for lighting?

Your thought experiment is wrong. There is a difference between the cost of something and its value. The cost of the ice cream cone is $1.45 and I am willing to stand in line for 20 minutes to spend the $1.45 for the ice cream cone. This means that the value to me of that ice cream cone is something greater than $1.45 plus the 20 minutes of time value. This greater amount is called consumer surplus. You can't, from this example, tell how much it is except that it is greater than zero. The decision to stand in line for cash involves a simple comparison of my time value and $1.45.

You have attempted to show irrationality in consumer choice by constructing two logically identical situations and demonstrating that people will choose one over the other. But they are not identical. In one you get an ice cream cone, and in the other you just get cash. Very different.

This is what is stupid about libertarians. We're not going to get a price on carbon, so yeah, that would be better, it won't happen. Lightbulbs can still increase welfare, there is no estimate here of the rebound effect, and there is a lot of work that finds consumers aren't especially sensitive to the price of electricity since they have no idea how much it'll save them to switch lightbulbs or turn off a light when they go to the next room.

Interesting that it can be this many comments with no examination of actual evidence. This is in part because actual evidence is available on paper and with a long reporting delay. If you check sales figures for the US for the A socket luminaires (i.e., ordinary bulbs) you will find that 20-25% of the bulbs sold are CFLs during the last five years of available data. (I looked at the numbers from NEMA.)

Energy efficiency and cost efficiency has already won. The longer life of CFLs means that 20-25% sales is an installed and used dominance for CFLs.

The rest of this issue is politics about who gains power. The eco-fanatics want their politicians to get power. Their opponents want other politicians to get power. The idea that the public could freely choose their own path is anathema to both.

The much longer life of CFLs means that the percentage of in-use bulb time for CFLs is probably already over 50%. It's difficult to use sales statistics directly because sales are for:
- Replacement of existing in-use bulbs that have failed. CFLs last much longer than incandescent, but it's not a fixed ratio. It's typically between 4-10 times longer, but hot locations like poorly ventilated recessed ceiling fixtures can drop CFL life to less than incandescent life.
- Inventory. I bought a ten-pack a couple years ago for cost reasons and still have four sitting on the shelf.
- New sockets, primarily from new construction and major renovations. Again, this has changed a lot over the past decade.
Only gross sales numbers are available. Using various models of new vs replacement and lifespans, I find 40% to 80% in-use hours for CFLs corresponding to the 20-25% sales.

This says CFLs have already won.

(As a side note, CFLs come in a wider variety of color temperatures than incandescents. So those who are concerned about the color temp of white should complain to the buyers of CFLs about their bad choice. It's not an inherent characteristic of CFLs.)

Energy sure is confusing:

*Politicians tax us by forcing us to buy more expensive, "green" light bulbs - that contain mercury. Doesn't sound very green to me.

*We spend billions in taxpayer money each year on ethanol subsidies and taxes, which everyone agrees causes more pollution than gasoline. Doesn't sound very green to me.

*Al Gore leaves from one of his many mansions to fly by private jet to give speeches on how to be green. Doesn't sound very green to me.

*We prohibit US companies from drilling for oil even though they follow the strictest regulations in the world. Instead, we buy oil from Russia, Africa and other places that have free reign to devastate their environments when drilling. Doesn't sound very green to me.

Energy sure is confusing.

Comments for this post are closed