Lonely as it is to be the guy arguing for economic efficiency in environmental policy discussions, it's absolutely critical to success on both fronts - good post.

It still has to add up, nicely argued. A few well placed subsidies could go a long way.

I came, I saw, I cried

One note: I think advocates of this kind of policy (like me!) need to find a way to distinguish the kind of incentives they have in mind from "subsidies."

After all, Ken isn't talking about Congress handing out money to favored industries. Just like feebates, this would be an automatic transfer of money, inside an economic sector, based on objective performance standards. No legislator would ever have his/her hands on this money. So the incentive would be better performance, not better lobbying.

grist.org

For example, I live in northern California, where we have a mild climate and little coal power, and I don't need to drive much, so I might see my net income rise by maybe a couple thousand dollars. That would be nice, but folks back east who are paying more wouldn't like it one bit.

It may be "unfair," but living in a non-temperate climate means a bigger carbon footprint.  A direct price on carbon would signal this loud and clear - and it should...

And vehicle owners DO think about gas prices.  I think the summer's shift in purchasing patterns is evidence...

In the cited example, tax-and-dividend creates a $1000 inequity between the hybrid and gas guzzler. Tax-and-subsidy (or whatever you want to call it) would apply the same $2000 guzzler tax to create a $20,000 inequity.

OK, I've always argued for this when it comes to cars. I actually think it can be automated even more and be made technology neutral. "Tax" cars with life-cycle emissions per mile above a certain amount. "Refund" to cars with life-cycle emissions per mile below a certain amount. Size of the tax depends on how much above target. Size of the refund depends on how much better miles per emission is. For electric cars, emissions would include average emissions per kWh from grid. For all cars manufacturing and end-of-life costs would be include, including costs from manufacturing and disposing of batteries (which means having a good recycling system in place would lower emisisons per mile slighly for hybrid, PHEV, and EV makers. The nice thing is, in the (very) unlikley event that hydrogen advocates turn out to be right, their vehicles would qualify for refunds too.

Gutenberg actively resisted known ways to lower the cost of printing books - and they managed to continue doing that for a long time.  Although the printing press was obviously from day one a cost-lowering technology, Gutenberg kept the bells & whistles & sex-appeal foremost ... and costs high.

"We don't want no stinkin' cheap books!

Hybrids today are the same deal: over-built and sexed-up deliberately to keep them a high-end product - same business-strategy that Gutenberg used.

... Whereas hybrids are actually, like airplanes, an inherently simpler, more-flexible, adaptable, interchangeable ... cost-lowering vehicle 'philosophy' - like the printing press.

To keep hybrids expensive, takes work.  Hybrids are expensive, because powerful interests fear the inherent cost-reduction of the fundamental approach.

To get America out of gas-guzzlers, the Government should throw a competition to design a low-cost hybrid, aimed at the low end of society.

Hoity-toity people don't want no stinkin' cheap hybrids.  No, they want a symbol of their social superiority.  And car-companies are happy to oblige them ... for the proper 'consideration'.

Screw 'em.  Let the government hold the competition to make a po' folk's hybrid - then offer assistance (if necessary) to get the winning design up on it's feet & running.

Probably, an existing company would see the wisdom of going cheap-hybrid (once it couldn't be prevented..), "Well ... if'n you put it that-a way...".

What we have currently is a BS game.  Hybrids are Ooh-La-la!, and spendy, and all the parties in a position to change the situation are 'shining it on'.

Ideally you'd also want to structure the feebate program to increase the subsidy as fuel economy increases, and increase the tax as fuel economy decreases (i.e. the more gas-guzzling, the higher the tax, and vice-versa).

Various feebate ideas has been around for a while, but maybe now we finally have a chance to get something enacted...

Don't know how well this would scale. Past a certain point I think you would simply want to combine standard based regulation with a straight carbon price.  

Also when you talk about recycling the money back into emissions reductions:

Fine for things like automobiles. Not so good if done generally, because even then you end up with things you want to do depending on things you don't want to as a source of revenue. Even with automobiles, you might end up with a lot autos that are just efficient enough not to get taxed, followed by an abrupt drop in the subsidy for electric cars, followed by a lot of electric car companies going out of business. A reason not to overdo subsidies. You want subsidies to help break chicken/egg deadlock. You don't want to create industries so dependent on subsidies, they collapse when the subsidies are reduced or eliminated.

Also, why is always more efficient to direct revenue from a carbon tax or feebate into the same sector? Suppose you tax the electric industry on emissions per kWh. OK, well maybe a good place to spend that is  electric industry:  solar or wind or generating electricity from waste heat. But maybe it makes more sense to use that money to weather seal homes. That saves some electricity directly, and it also saves natural gas that can then be used as a replacement for coal in electricity generation.  

If you moved from specific products to sectors maybe you could generalize it. Fees for homes and offices that emit above a certain level per person or per square feet, subsidies for weather sealing and duct sealing, and insulation, and window treatments, and efficient lighting, appliances and office equipment, and subsidies for solar space and water heating and for ground source heat pumps. Feebates in transport based on emissions per passenger mile and per ton mile. Feebates in electricity based on emissions per kWh. Ironically, the one sector we do NOT have a good proxy , that is a good what in the emissions per what equation is manufacturing. I know Sean thinks emissions per delivered BTU work process is such a proxy, but he overlooks that there are a lot of ways to reduce emissions per dollar of output that this does not capture. In manufacture of high quality steel for example, beyond heat per ton of steel, there is near net shape manufacture of steel, which reduces scrap during manufacture, and thus ends up with more USABLE steel per unit of emissions. Also there is not just a question of scrappage of raw materials incorporated in products, but scrappage (as opposed to use) of materials used in manufacturing process not incorporated in product. Spills, emergency flaring and so on. Plus of course scrapping finished products due to various avoidable problems (for example not adequately inspecting raw materials, and not adequately controlling processes.) 8% to 16% of industrial emissions could be reduced just by feasible reductions in avoidable scrapping, about a 1/4th to 1/2 of savings from Tom Casten's estimate of maximum feasible savings from recycled energy.  And that does not even consider other forms of changes in material intensty: reduced use of materials, reduced use of high carbon footprint marterials, increased product lifespan. Nor does it consider changes in process once material choice and lifespan are optimized that can reduce BTUs needed to drive a particular result as opposed to just delivering BTUS to that process more efficienctly.

Subsidies can be simply implemented as a refunded tax. For example, Take the $2000 tax on the gas guzzler and the $1000 tax on the hybrid and return the tax at a uniform dollar-per-vehicle rate ($1500 each). The result is a net $500 tax ("fee") on the guzzler and a net $500 subsidy ("rebate") for the hybrid, same difference.

The general principle is that the tax is applied to emissions (at a uniform dollar-per ton rate) and is refunded in proportion to emissions-related economic utility (e.g., per vehicle, if they are in the same utility class). There should be no "abrupt drops" or discontinuities in the tax/refund formulas. Marginal incentives for emission reduction are determined by the tax rate (aka. "emission price"), but with the refund a high tax rate does not imply high net taxes. Initially, net taxes are low and net subsidies are high, but as clean tech displaces fossil fuels the revenue-neutral allocation formula automatically shifts to high net taxes and low net subsidies.

Keeping revenue neutrality within an industry, or even within sub-sectors within an industry, can in some cases avoid massive and economically destabilizing revenue flows that would make a high tax rate impracticable without providing any real emission benefit. For example, if pure output-based refunding were applied to electricity generation, most of the revenue flow would go to legacy renewables (large hydro, nuclear). That may be fine if those sectors have capacity for expansion to displace fossil fuels. But if they don't, the subsidy diverts resources from new renewables, which could be scaled up much quicker and cheaper if the subsidy shift were focused on new renewables.

Economy-wide policies (tax or otherwise) are intended to reduce the cost of achieving near-term targets and mandates, but can increase long-term costs of climate stabilization. The economic theory behind such policies essentially rationalizes procrastination. For example, investing in weather stripping might be able to achieve a marginal 1 percent reduction in coal use more cheaply than installing wind turbines, but to achieve 100 percent reduction in coal we will have to make a heavy investment in renewables sooner or later -- better sooner if there is to be any hope of attaining climate stabilization. Don't pick the "low-hanging fruit" first if the best, ripest fruit is up high.

But I'm trying to visualize this policy applied on a large scale. Are you saying we could do carbon policy with 100% feebates? Because I've always see that as something that would be nice, but is impossible in practice. Do you really mean to do it on a "goods class" by "goods class" basis? Or are you talking more a sector by sector basis? That is toaster rebate to toaster, refrigerators rebate to refrigerators? Or household appliances repbate to household appliances? Or are you you set a general rule of emissions per person and per square foot, and higher emissions than this rebate to those with lower emissions that this?

I'm trying to envision widespread use of feebates covering whole sector on a product by product basis. Past a certain point, I think you'd have to go for sectorial implementation. I can't see feebates implemented for thousands or tens of thousand of individual products.

Then we are on same page.  

If it could be made to work for electrical generation, that would be a Very Big Tool.

grist.org

Actually I've always agreed that output based standards are feasible with electricity. It is the manufacturing end that makes zero sense. Emissions per kWh are a great standard for electricity. Because a kWh is a good surrogate for the economic value of electrical generation. (That is a kWh from one source is pretty the same value as a kWh from another. It is not a perfect standard. Dispatchability, and how far it is from point of demand do factor into value, and are not not taken into account in a an emissions per kWh standard. But kWh is a good enough surrogate for practical purposes. )

Of course in terms of effectiveness I'm pretty sure straight regulation beats output based standards.   A simple ever tightening regulatory standard for total emissions per kWh within each utility district will have the same results. And most electricity is produced within utility districts large enough to provide plenty of flexibility in terms of path take to achieve reductions. If compliance within a single district did not allow enough flexibility you would allow utility disticts to voluntarilty combine into compliance superdisticts to let them achieve compliance across multiple districts.

Also straight emissions standards per kWh rather than an output based system in this case achieves the same advantage David attributes to cap-and-trade over carbon taxes: certainty in amount of reduction rather certainty of emissions price.

One problem with vehicle feebates is how to quantify "output" (i.e. emission-related economic utility) for the purpose of refund allocation. With electricity, there is no such complication: "Output" is just MWh; refunds are allocated in proportion to MWh output. The "canonical" output-based refunded tax for electricity generation is the Swedish NOx program -- see my last note to Gar. But I should again caution that this "ain't no silver bullet" -- pure output-based refunding would send most of the subsidy to large hydro and nuclear, not to new renewables, so the Tool needs to be applied judiciously.

Re "...straight regulation beats output based standards."

Why not cut to the chase and set a regulatory standard of zero emissions? Well, because we are more concerned about costs than about emissions. So how do you set your regulatory targets to ensure cost acceptability, given that you can't reliably predict future costs?  You base them on worst-case assumptions, biased toward extreme cost conservatism. And when costs turn out be lower than expected (and benefits higher), you find that your regulations are not motivating emission reductions commensurate with our willingness to pay, not to mention sustainability requirements. But because regulations are so inflexible, you need an Act of Congress to tighten them.

With monetary incentives (e.g. taxes, feebates) you set the price incentive by direct regulation. If technology turns out be be cheap, you get lower emissions, not price collapse. No Act of Congress required.

Zero is the ultimate goal. We don't set it as an IMMEDIATE goal.  A phaseout to zero over the course of 20 years with regulations that were utility wide rather than per plant, and allowed utilities to form superdistricts for greater flexiblity, would provide plenty of flexibility to attain goals at minimum price. It seems that the argument here is comparable to carbon tax vs. cap-and-trade. Except ironically I'm arguing for goal certainty rather than price certainty. The difference is when we talk standards based reg vs. output based standards we are talking about core means of lowering emisisons, not reinforcements. And when it comes to core means we do need target certaintly.

I think Hansen's position has been misunderstood, and this discussion, while well-intentioned, takes us further from where we need to be.

There are three legs for carbon policy: regulation, investment and an economy-wide cap/tax.  Hansen prefers a tax; I prefer a cap.  In a world of certainty (we know exactly how much tax equals how much emission reduction) the two would be the same.  We don't live in that world, so the burden of uncertainty has to fall one way or the other.  Hansen would put it on carbon emissions, I'd put it on carbon prices; otherwise we're in sync.  In any case, the third leg is not intended to do the whole job but just to make sure that the job gets done.

The cap/tax needs to be as simple as possible.  The goal of this leg is to fix in place as a requirement the achievement of our overall carbon target, so have a single cap/tax for the whole thing.  Just one, applied across the board.  No fiddling with this product, that technology, etc.  If it's a tax, say per ton of carbon introduced into the economy, period.  If it's a cap, specify the number of permits for introducing carbon into the economy and auction them all off.  Leave all the detailed consideration of how we save carbon where to regulation, investment and markets.

The tax/cap is a sales tax.  Sales taxes are regressive.  But give the money back to the public as a dividend, and the majority of people will come out ahead.  In particular: (1) Don't fund green investments out of a national sales tax.  Don't be regressive.  (2) Care about the living standards of ordinary people.  Lots of folks are barely making it or not at all, especially in the current economic situation.  Don't raise their energy taxes and leave them to suffer the consequences.  Give them the money back.  (3) We are talking about a carbon diet that the country must adhere to for at least the next 40 years.  It has to be politically untouchable, like Social Security.  (Bush tried to junk it but couldn't get anywhere.)  If you don't give the money back you will have to depend on the willingness of political majorities, year after year after year, to willingly tighten their belts for the green cause.  I'd rather up the odds by adopting a policy that goes out of its way not to hurt people.

Repeat: we need lots of targeted policy in housing, transportation, manufacturing, agriculture.  This is not the role of the cap/tax, although it does provide an appropriate incentive environment for policy.  (Think how much easier it will be to get voters to support bonds for mass transit when gas is really expensive.)  So we don't need fine-tuning on carbon prices, just a single, big, emission-reduction-forcing boost will do it.

Peter

Tax and refund is a comprehensive policy that would allow increasing the cost of fossil fuel to very high levels without putting an unbearable burden on the poor.

A policy built on subsidizing hybrids and similar targeted subsidies would be problematic on several fronts. Here are a few issues I came up with while eating a small, nutritious snack:

"Green subsidies" is not a comprehensive policy but a patchwork of ad hoc solutions that would not give consumers, producers and innovators a clear and predictable system of incentives to reduce carbon emissions and develop green technologies.

It is open to manipulation as various special interests would lobby for subsidies.

The benefits of many such subsidies would be hard to analyze in advance and would be open to debate. Your suggestion of subsidizing hybrids, for example, could be very problematic. E.g., it would encourage buying cars - hybrids, but still cars; replacing conventional cars for hybrids carries the cost of producing the hybrids; buying hybrids locks society into the private car transportation mode; for people spending much of their driving time on the highway a hybrid does not significantly increase efficiency.

While the costs and benefits of tax-and-refund clearly flow from people who pollute less to people who pollute more, it is hard to control how the costs and benefits of the subsidies will be distributed. For example, people who use public transportation would not be rewarded, while people who already own a gas guzzler would not be penalized. If R&D is eligible for subsidies, various industries could use the subsidies to fund their research and development projects, socializing the costs and privatizing the profits. If R&D is not eligible for subsidies, the subsidies would entrench existing technologies, deterring development of new ones.

"It is democratic for public offices to be assigned by lot, for them to be elected is oligarchic." -Aristotle