October 28, 2007
How Big A Carbon Tax To Dampen Gasoline Demand?

An editorial in the Christian Science Monitor discusses the carbon tax idea as a way to reduce carbon dioxide emissions.

Economists agree that the real cost of burning fossil fuels – damage to the environment and health, not to mention the cost of replacing them as they run out – isn't reflected in today's prices. A carbon tax would directly send a market signal to reduce carbon use. And it would provide an incentive for investment in renewable sources, especially if the tax is set at the source: for natural gas, at the wellhead; for coal, at the mine entrance. Oil would be charged at the refinery because petroleum products create different levels of emissions when burned.

The World Resources Institute calculates that a tax of $15 per ton of carbon-dioxide emissions would double the costs for coal use and raise gasoline prices about 13 cents a gallon (or about 5 percent, at today's prices). Natural-gas prices would rise less than 7 percent. That would result in a 12 percent reduction in CO2 emissions.

So get this: Some carbon tax advocates who want to prevent global warming advocate a carbon tax that will increase the cost of gasoline by a mean 13 cents a gallon. Is that all we are arguing over when we argue about whether to take any expensive steps to stop global warming? I've previously read references to an even higher $30 per ton carbon tax. That's double the $15 per ton and so lets go with that for a 26 cents a gallon increase in gasoline cost. Now we need some historical perspective on gasoline costs to see if 26 cents amounts to much in terms of reduced gasoline demand.

Check out this table of refiner wholesale gasoline costs from 1998 to 2007. Back in the good old days of 1998 gasoline sold at wholesale for a mere 53 cents per gallon. The cost has more than quadrupled to $2.13 per gallon for 2007 and probably right now it is even higher. At the retail level gasoline bottomed out slightly below $1 per gallon in late 1998 and early 1999 and is around $2.80 as of this writing. So the price of gasoline increased by $1.80 per gallon just due to market forces (though by only about $1.50 per gallon if adjusted for inflation). Doesn't that make a 26 cent a gallon increase from a carbon tax seem, well, somehow inadequate for the goal of reduced consumption it is meant to achieve? Recently gasoline demand in the United States went down .2% as compared to early October 2006. The rising price of gasoline has finally stopped consumption growth. But a pretty big increase was necessary to stop consumption growth. If a far larger increase in gasoline prices has only barely stopped gasoline consumption growth why should a carbon tax of 13 or 26 cents a gallon make a big difference? Granted, it will make a difference. But the difference won't be large.

Maybe a carbon tax of between $15 and $30 per ton will have a much bigger impact on the use of coal for electric generation. I'd really like to know how much a dollar of carbon tax increases the price of coal electric per kilowatt-hour. How big a carbon tax on coal would make nuclear power cheaper than coal electric?

But if a carbon tax isn't going to make much difference in US demand far more powerful forces are already pushing up the price of oil, reducing US demand, and these forces promise to push up oil prices even higher. Can anyone guess what I'm thinking about the future when reading this paragraph?

Market Saturation -- The U.S. has reached the unusual position of having more vehicles than licensed drivers -- 1,148 registered personal vehicles (cars and light trucks) for every 1,000 licensed drivers. Britain has 700; Mexico, 208. Brazil has 137 per thousand eligible drivers, and India has 11 per thousand. while China has just nine cars per thousand eligible drivers.

What happens to the price of oil when China reaches 100 cars per thousand eligible drivers? Chinese drivers are going to force US drivers into smaller cars, public transit, electric cars, and electric mopeds.

Share |      Randall Parker, 2007 October 28 11:14 PM  Climate Policy

averros said at October 29, 2007 12:34 AM:

Carbon tax is just another harebrained socialist scheme to screw the economics (and, consequently, the technological progress) while enriching the elite. They cannot come up with consistent valuation of environmental harm from CO2 precisely because it's kind of hard to put valuation on a religious belief in Mother Nature's sacrity. How much does desecration cost?:)

Now, if someone could take a CO2 "polluter" to a court and *prove* that damage to some specific victim exists... but the reality is, they can't. It is not even clear if "global warming" is undesireable.

David Weisman said at October 29, 2007 4:27 AM:

That tax might have a tiny effect on the economy in the short term, might be beneficial in the long and medium terms as it leads to the development of new technology. It could be written so it would expire when the price of gas was too high - thus encouraging the market to prepare for future prices before they were officially reflected.

The problems are political. The coal industry would fight it tooth and nail - not primarily because of the amount of the tax. The mere idea of setting a precedent, of acknowledging that we can and should take action, is what they fight tooth and nail. After all, if the tax worked well we might increase it, and the mere passage of even a token tax would be admitting all the things the deniers deny. The only enemies would not even be those with economic interests. I think libertarian leaning Republicans have the same emotional reaction to this that the religious right has to abortion.

David Weisman

MI said at October 29, 2007 5:38 AM:

I think Worldwatch Institute might be lowballing the cost (no surprise there). See the following site:


...and in particular slides 21 & 25 of this slideshow they put together to promote carbon taxes:


They're advocating a $370/ton carbon tax, which they freely admit would raise gas prices by $1/gal, and electricity prices by 7.2 cents/kWh. They also admit that even this (heavy) tax would only reduce CO2 emissions by one third.

As far as I can tell, this is a bona fide environmentalist site, not something funded by the fossil fuel or auto industries. Which makes the aforementioned even more interesting, IMHO.

Dave Huff said at October 29, 2007 6:49 AM:

I think another interesting idea would be to take the carbon taxes collected and use them to reduce other taxes or better to provide a per person rebate every year (not one controlled by your usage, but just a per capita refund). Therefore, on average, there would be NO cost to the economy. And over time, people would adopt more fuel efficient means to game the system. Note the refund idea allows other taxes to stay constant, which is important as (hopefully) carbon consumption and thus taxes will decline over time.

I don't believe 0.13/gal will impact consumption, but if we put a steadily increasing plan in place, it will eventually drive choices, but in a timeframe that doesn't hurt too much, and if other taxes decline in response, again, there is NO COST on average!

Of course, since this is a sensible and reasonable plan without any real painful impact to the populace, it won't happen. Note that in cap and trade systems, consumers will pay additional cost without any recompense.

Tom said at October 29, 2007 10:15 AM:

Dave: Even with what you recommend, there can still be cost to the society. Taxing the cheaper, dirtier product will generally result in some people paying the tax, and some switching to more expensive, cleaner products. You're right that on net, there is no net cost in regard to those people who pay the tax. But for those who switch to the cleaner, more expensive product, there is more cost, and no tax revenue to offset it.

K said at October 29, 2007 3:24 PM:

The measurement is not good. People have little choice about the amount of gasoline they use in the short term. Over the longer term they can change jobs, or where they live, or what they drive, and their habits.

And they don't have much choice yet about what they drive. Few vehicles offer much better mileage than they did five years ago. But those that do are selling very well even though hybrids cost more. And many new models are on the way, 2009 models will be when things really change.

For several months it has appeared that US gasoline consumption has peaked and may never again be as high. This is good news. And this is precisely the time for a higher but not ruinous tax on gasoline. Either by a direct increase in gasoline taxes or by a carbon tax on oil. I prefer the former method for now simply because there are already gasoline taxes, nothing new is needed to boost the rate.

Don't fall for the nonsense that the tax must be fair. That is not the nature of taxes. Keep it simple - to discourage gasoline use raise the cost.

Dave Huff said at October 29, 2007 4:40 PM:


Agreed, there is a cost, mostly in economic efficiency due to the churn of the money. If you choose to use a greener alternative and therefore not pay taxes, you still get a refund as long as others are still burning carbon. Hence, you get to double dip, or at least make an economically indifferent choice. Eventually, as fewer burn carbon, this benefit drops as the tax rebate drops, and people who wait, for whatever reason will benefit less when they change. Some will decide to switch soon because their fuel burning assets are largely depreciated, while others will wait since the depreciated value far exceeds the savings, that is why it is a good system allowing reasonable capital purchasing decisions.

The other cost is in by switching to another energy source because of taxes, you are in theory making an economically disfavored choice (otherwise you would already be using that other source). But that is only because today the 'cost' of pollution is an externality and not born by a purchaser of, say, gasoline. In reality, society pays this cost (for instance in the costs of global warming), and efficient taxes would offset this cost exactly. Unlikely I agree, but in the ideal case, there is no cost on average individually or to society as a whole. Practical limitations do in fact introduce cost, but they are nominal under a reasonable taxing regime.

Dave Huff said at October 29, 2007 4:42 PM:

@Tom, PS, the thing you may have missed is that you receive the per capita tax refund whether or not you use carbon-based fuels...

Randall Parker said at October 29, 2007 6:12 PM:


A high carbon tax wouldn't increase electricity rates by that much because the cost of substitutes is lower than the 7.2 cents per kwh projected increase. Raise the cost of coal plants by 2 cents a kwh and suddenly you'll see a huge shift to nuclear.

The existing coal burners don't want the carbon tax because given a high enough carbon tax their plants become worthless.

If the coal burners can hold off carbon taxes long enough then sequestration technologies might give them a way to avoid the tax. But sequestration takes energy and capital which raises the price cost of coal electric. Likely sequestration will raise the cost of coal electric over that of nuclear electric.

Randall Parker said at October 29, 2007 6:13 PM:

Dave Huff,

I do not trust governments to refund the full money collected by carbon taxes.

Randall Parker said at October 29, 2007 6:16 PM:


Some of the shift to newer and cleaner products won't be net cost. Lots of people apply a wrong discount to the cost of money when considering fuel saving features in cars. If the incentives for them to buy a hybrid or diesel increase that will tend to make them make choices they already ought to make. So the point is a carbon tax will reduce market failure.

Plus, there are real externalities of fossil fuels such as particulates and oxides of nitrogen which will be reduced by a carbon tax.

Engineer-Poet said at October 29, 2007 8:03 PM:

Tom:  You forgot the third alternative, people adopt more efficient methods and just use less.

I see (still!) people speeding past me in Chevy Suburbans, weaving through traffic (on surface roads, not freeways) to get to the next red light a little sooner.  The average person can probably save 10% on fuel just by driving a bit differently, and I would not be surprised if the guzzling jackasses could save 20-30%.

I did better than that.  The last time I had a commute like I have now, I got about 18 MPG.  Today I'm getting 35-36, for about 50% savings.  Half of this is the vehicle (SHO Taurus vs. Passat TDI) and the other half is careful driving.  I shift to neutral and coast to lights, carrying speed by not arriving before the green if possible, and engine-braking when necessary to put the injection pump into fuel cut-off mode and eliminating the idling demand.

What I can do, I can teach other people how to do.  All they need is the will to change.

Fat Man said at October 29, 2007 10:12 PM:

"How big a carbon tax on coal would make nuclear power cheaper than coal electric?"

NEI claims that: nuclear generated electricity costs $0.0132/kWh and that coal generated electricity costs $0.0237.

Reference: Electricity from coal produces CO2 at a rate of about 0.915 kg CO2/kWh.

0.915 kg is about 2 lbs. A tax of $15/ton CO2, would therefor add $0.015/kWh to the price of a coal generated kWh. Given that the current national retail average is about $0.104 Reference that is a pretty substantial increase. Whether that increase alone could change consumption patterns is another question. A $30/ton CO2 tax would probably make fossil fuel uncompetitive.

Rob McMillin said at October 30, 2007 2:45 PM:

Carbon taxes are a regressive and bad way to get people off fossil fuels, not unlike funding a health care system with a tax on cigarettes. Anybody suggesting that a "surplus" would be "refunded" needs to look at the profligacy of government.

Dave Huff said at October 31, 2007 11:42 AM:

@ Randy/Rob: Ahh, well, one can dream. I wasn't talking about surplus being refunded, I was looking at it ALL being refunded. The accounting is easy, but I agree that trusting government is problematic at best. Like I said, it is a great theory. In practice, it is the only way I can think of to incent a change in behavior without impacting, especially low income folks, disproportionately. It is a serious question, as Rob points out.

Larry said at October 31, 2007 11:57 AM:

There are lots of vehicles that are substantially more efficient than today's average. People can change and will given incentives.

Changing the tax system always involves winners and losers, but somehow (even after Bush's "scandalous" 2003 cuts) we always end up with 18-21% of the GDP going to Washington (Reagan, Bush 41, Clinton, Bush 43). It seems that when we go off from that somehow, things change and we end up there again. I prefer 18 over 21, natch!

I support a predictably increasing carbon tax which would raise US energy prices over 10 years to be similar to those in Europe, offsetting those increased taxes with tax cuts and reforms elsewhere that would keep federal revenues around (you guessed it) 18%. Something like 25 cents a gallon each year for four years, then 50 cents a year for 6 years.

That would give everyone time to adjust, and result in dramatic cuts in our CO2 production.

Randall Parker said at October 31, 2007 4:49 PM:


I do not expect taxes will be needed to raise gasoline prices. I expect a world decline in oil production will accomplish this without any taxes.

The 18% to 21% taxes: Yes, I'm aware of that range of tax collection by US federal government going back decades. But if we get a federal consumption tax (and that is what a carbon tax would be) we will go outside of that historical range. A VAT or sales tax lets governments collect more with less political resistance. Taxes on income are much more visible.

Dave Huff,

The lower income folks are going to be disproportionately impacted by rising oil prices. They are going to have to live in higher density areas closer to jobs.

Engineer-Poet said at October 31, 2007 6:46 PM:

The problem with waiting for declining oil production to raise prices is that the revenue goes to oil producing nations (many of which are hostile).

This also accelerates the increase in price and impact on oil importers (us).  If all the price increase goes to producers, it increases the producers' buying power... including for petroleum and the things which use it.  Domestic consumption in producing nations will rise even as production falls, due to the increased buying power of their remaining exports.  The exports will decline far more rapidly than production, leaving importers (like us) in the lurch.  (This is Jeffery Brown's "Export Land Model".)

The only solution is to drive oil prices high enough to dampen demand and push efficiency WITHOUT letting the producers have the money:  taxes.

Jeremy Hellman said at December 14, 2007 12:28 PM:


Just curious as to where you got that data on Cars per Thousand. Can you share your source ?


Randall Parker said at December 15, 2007 10:42 AM:


The numbers for China and India are changing very rapidly. Here's a 2007 source (Cambridge Energy Research Associates) which measures it per thousand eligible drivers rather than per thousand people total:

When it comes to cars, the United States is saturated--and then some. It has reached the unusual situation among all the countries of the world of having 1,148 registered personal vehicles (cars and light trucks) for every 1,000 licensed drivers (see Figure O-4). France and Britain follow with 702 and 700 cars, respectively, per thousand eligible drivers. Japan is also up there. Despite its small land size, Japan has some 608 cars per thousand eligible drivers. In other countries, the numbers fall away. Brazil has 137 per thousand eligible drivers. But even Brazil seems well supplied with cars when compared with China and India: India has just 11 cars per thousand eligible drivers and there are only 9 per thousand in China. That tells you something about what to expect for the future and where the growth will be!

Numbers based on per thousand people total show smaller numbers for the US and probably for India and China. But they have different ratios of adults to children and so the numbers won't scale quite the same. Here is a 2006 source that puts India at 7 per thousand and China a little higher:

According to Global Automotive Financial Review, the Indian passenger car market will go up to 19 lakh cars by the year end 2010 and the market is estimated to witness an impressive growth of above 50 percent by 2010.

India's passenger vehicle sales in the fiscal year, ended on March 31, 2006, rose to record 882,094 units, which is an increase of 7.6 percent from last year. Vehicle sales have increased in 7 year out of the past ten years. India has about 7 cars per thousand people, fewer than the 10 per thousand in China, 450 in the US and 500 in Western Europe.

Robert Zubrin phrases it differently as the number of people who own cars:

It is worth dwelling for a moment on the magnitude of the rising worldwide demand for oil. Even though oil prices have quadrupled over the past six years, the rate of rise of global oil consumption is accelerating, with recent annual increases of about 1.7 million barrels per day. Only about 11 percent of this rise is occurring in the United States; most of it—by far—is occurring in China, India, Eastern Europe, and Latin America. Everywhere the same pattern is repeating: Once people become wealthy enough to buy a car, they do so. Auto sales in China doubled between 2001 and 2003, and have doubled again since. They are going to keep doubling. In the United States today, 800 out of every thousand people own cars. In China the number right now is only 8 cars per thousand. There are a lot more cars coming. These developments promise to drive the price of oil through the roof.

In the fight against OPEC, conservation is a sure loser.

I've read some place I can't find that there are 200 million cars in the United States. I think that number is correct. The "cars per thousand" figures are probably used sloppily by various writers who do not distinguish between total populations versus adults versus possessors of drivers licenses. Some who own multiple cars only drive one at a time.

But the big picture is that by 2030 China will probably have double the number of cars sold per year than the United States. There's no way oil production is going to keep up with that demand. Likely the estimates for numbers of cars sold are optimistic precisely because we won't have enough oil to power them all.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©