Thursday, August 9, 2012

Prices, or Prizes?

When it comes to global warming and pollution, the conventional wisdom among economists is to say that we've developed a set of public policy solutions -- among them, cap-and-trade, carbon taxation, and research-and-development subsidies -- but note their implementation is restrained by unfortunate political realities.

A cap-and-trade program would set a ceiling on emissions of greenhouse gases -- most importantly, carbon dioxide -- creating emission credits for polluters, and establishing secondary spot and futures markets for credits which establish individual rights to pollute. A carbon tax follows a similar approach: polluting activity would be subject to tax. There would be no solid cap of emissions, but a tax could align more exactly with the marginal external costs of pollution and thus be more economically efficient. And research-and-development subsidies in energy technology are a no-brainer, as they should bring clean tech to market more quickly and open up new channels for pollution reduction.

The economists' viewpoint on these questions can be distilled to three goals: shift demand for resources to less pollutive ends, make polluters pay for the full social cost of their actions, and support the expansion of supply of clean energy.

The economists' proposals, I contend, have also failed -- both politically and in light of new economic arguments. It is time for a new conventional wisdom on global warming, a new playbook of policy proposals.

Cap-and-trade programs worked well in the reduction of chlorofluorocarbons (CFCs), sulfur dioxides, and nitrogen oxides. That has not proven quite as true in the extensions of cap-and-trade programs onto carbon dioxide. Europe's Emissions Trading System has struggled to gain traction, with rising coal use and an unstable and illiquid market for carbon credits. The UN's system has resulted in the diversion of pollution from one pollutant to another, according to a recent story in The New York Times. Legislation to establish federal cap-and-trade program, known as Waxman-Markey, died in the Senate in 2009. Carbon taxation has shown some promise through a tax-shifting plan enacted in British Columbia which used the revenues from a carbon tax to reduce its personal and corporate income taxes. And yet it will go nowhere politically.

More importantly, schemes which increase the price of polluting are more likely to shift pollution to where those activities go untaxed rather than shift production into less pollutive practices. Noah Smith put it well in a recent post, writing that "[c]arbon taxes are undermined by free trade...[polluter location-shifting] will not completely cancel out the effect of a U.S. carbon tax, but they will work against it substantially. The only way to stop this would be to tax both carbon exports and the implied carbon content of imports. This would lead to big rises in tariffs...And you know what the U.S. can't tax? Cheap-carbon Chinese-made products replacing expensive-carbon U.S.-made products in global markets." Smith also writes that "[g]lobal carbon taxes present a basically insoluble coordination problem," given that lack of an international governing body to enforce pollution property rights leaves the problem in "tradegy-of-the-commons" mode.

Noah writes that the hope of environmentalists should go with natural gas for the medium run and subsidies for energy research-and-development in the long run. I agree with the first. But I think plain-old research subsidies are unlikely to work in speeding the introduction of clean energy technologies.

The problem is that both the short- and long-run supply of research are highly inelastic. This means that an increase in the demand for research, as subsidies would achieve, will go almost exclusively to increasing the price of research rather than the quantity thereof. More to the point, subsidies will increase the pay of the researchers, but better pay alone is unlikely to increase the research produced. This is the unfortunate conclusion of a 1998 paper by Austan Goolsbee, who was more recently the Chairman of the Council of Economic Advisers. Goolsbee found that conventional estimates of the effectiveness of public R&D subsidies may be 30 to 50 percent too high, and that there are strong crowding-out effects of public subsidies on private research.

Another study from Paul Romer in 2001 finds that the institutional arrangements of higher education restrict long-run elasticity of supply for research -- basically, if we made the education system more responsive to price signals, then we could produce significantly more scientists. This is one of the reasons why petroleum engineering has to pay six-figure entry-level salaries to convince students that the field will be lucrative -- it takes extreme price signals to shift any significant number of students among potential courses of study.

All three major policy proposals -- cap-and-trade, carbon taxation, research subsidies -- are either politically dead, economically ineffective, or some unfortunate combination of the two.

Therefore we need a new, sound policy to tackle pollution and global warming. I suggest a tournament approach for energy science research, paired with education reform to establish more science and technology charter schools and infrastructural development to facilitate a switchover to natural gas.

Tournaments revise how government finances innovative research and development. Instead of giving a team of researchers a grant to pursue research in a given field, government sets a goal -- say, a car which runs on conventional gasoline, costs less than $40,000 when produced at scale, and gets over 110 miles to the gallon on the highway and 67 in the city. (The Prius and Volt both get somewhere in the 90s and 50s in each respective setting.) Teams of researchers from around the world compete for a large cash prize. Sometimes they receive patent rights also; in other times the body of research has gone into the public domain.

Why are they superior to vanilla subsidies? For one, they create externalities inherent to the tournament set up -- more research is produced, and the reward is greater due to the effects of recognition. Experimental tests of research tournaments find powerful effects; it turns out that researchers are competitive enough that they will organize themselves better and generate an effective increase in supply under the incentive structure established by tournaments. Other historical studies find that several major innovative advances, including sea, train, and plane travel, were achieved under tournament structures. Related work in medical research also suggests that, given the extreme risk-reward profile in disruptive research, conventional subsidies are unsuited to the task, and tournaments with large prizes are better.

Prizes, rather than subsidies and patents, seems to be the right way forward for alternative energy research and development. "A modern R&D prize system could focus innovative efforts on important societal problems where the answers are not obvious, and which does not rely on a proprietary approach to knowledge," another study reports. (See also here for the particulars on climate technology.)

Addendum: Yichuan Wang, who turns out to know a lot about the state of space research, discusses here public/private involvement and the prize paradigm for technological advances.


  1. You sold this solid liberal on your proposal. The prizes would have to be over a $100M - but probably a pittance compared to what the government spends now. The $1M robot challenge by DARPA - and the amazing advances over a few short years - validate your idea.

    1. Thanks. One of the best signs that the public sector should try prizes, in my opinion, is that the private sector is turning to them. Consider Netflix's algorithm competition, for one.

  2. How good a job do you think a government is likely to do in setting the goals for which it will award prizes, and in setting the magnitudes of the prizes? Your example—“a car which runs on conventional gasoline, costs less than $40,000 when produced at scale, and gets over 110 miles to the gallon on the highway and 67 in the city”—is not very compelling. How would such numbers be chosen in the political process (by the way, how did you choose them in making up the example?), and how would another important number, the magnitude of the reward, be determined?

    I'm skeptical.

    1. That was just an example, and I merely moved the bar slightly higher than what we could already do with tech. As to selecting prize goals and prize amounts, that seems to be a technical question which can be settled by the governmental bodies we use for granting out research money. I would imagine private-sector input would also be informative.

  3. How would research toward the prize be funded? By the expected value of the research? This is very low considering an actors small probability of winning.

    1. The competition might provide funds for qualified entrants, and/or it would come from already existing research budgets for private companies.

  4. Evan, I was a bit inspired by your previous post about the housing market and the macroeconomy to post this:

    1. Thanks, Julian. Just read your post. Given that your view comes down to over-indebtedness and overoptimism shifting suddenly to pessimism and deleveraging, I should recommend that you look at Fisher's work on debt deflation. Fisher concluded that over-indebtedness and price bubbles can resolve themselves without macroeconomic disturbance if there is an expectation of stability in nominal income or the price level.

  5. Evan - your basic point - prizes are a great way of generating research break throughs - is an excellent one.

    But looking for a research-based solution - a super-efficient car, for example - risks perpetuating the myth that we can do something about global warming without making major lifestyle adjustments. (Super-efficient cars exist right now b.t.w.- they're called electric/hybrid bicycles)

    So, sure, fuel-efficient planes are better than non-fuel-efficient planes. But better by far is just not flying at all.

    Here's my break-through save-the-environment idea: stay home and take a nap.

    1. Frances,

      Thanks for stopping by. You should know that Worthwhile Canadian Initiative is one of my favorite blogs. I don't know if I agree that it is a "myth" that consumption shifting into pollutive goods and services will be enough, and that saving the environment requires inevitably a reduction in total consumption. I think it could be enough; it would depend upon the shape of consumers' indifference curves when it came to pollutive and green consumption.

  6. Evan,

    You are the only 19 year old I am listening to these days :)

    I have a feeling that the death of Friedman has left a big void on Planet Earth. Everyone seems to be in central planning mode. In fact my hunch is that this entire prolonged financial crisis is simply because Friedman is no longer alive :D and the left liberals, as they usually do, with their power over language and zero experience with "building anything" have captured too much of the mind space of our time. That alone must be sub consciously frightening business investment and causing all that "uncertainty" all over :)

    Anyways I think you got this one wrong!

    It is never the prize that is the problem with any risky investment or project or research. Every one knows the "Prize" for say being able to replace "Oil" is in the Multi Billion dollar range. There is a lot of private money invested in this area and I would bet if IT can be done IT will be done by these guys.

    In fact from the world of University type Research I would bet on a Black Swan in the fundamental sciences like Physics, Chemistry etc. Something which a Government or central planner cannot pre-define.

    The government should save its money for the time someone actually comes up with something dramatic to make the infrastructural investments they will require.

    1. The obvious response to your argument that there is a lot of private money in this is that government should only be subsidizing R&D prizes in areas where technological advances will produce large positive externalities. In such fields, no matter how great the private incentive, the optimal incentive at the level of the society should be greater.

    2. The eternal problem is who decides which technological advances produce large positive externalities including the caveat that the private sector misses this opportunity. What History tells us is that especially these things cannot be centrally planned. The "internet" could never have been the "imagined goal" of the government before beginning to fund it. The best you can do is give a lot of smart people lot of money and have patience. The private sector (VCs ) are already doing this in a big way and have already seen some green shoots. The government tried it and failed big time. Though it works better with fundamental research.

      Unintended consequences and black swans are the realities of our world. But I would probably agree with you that Prizes might cause less harm than other government interventions that are already pervasive in this field. I am not for zero government intervention but I think it needs to wait till the private sector is unanimous in their view of where they need government funding. Which is generally after a major invention/discovery has been proven to work.

  7. Hi Evan,

    Great post; I'd like to add a couple things form a political and (forgive me) anectdotal perspective.

    First, there is a major political issue with cap and trade and other taxing schemes, namely, that increasing the scope of the tax code and/or creating a regulatory regime leads industry to lobby for exceptions and other favors. The Waxman-Markey bill may have enacted cap and trade, but it basically amounted to a massive corporate subsidy (at least by some readings on the right). I won't claim to have read the bill, but that is inevitably the result of a new regulatory regime.

    From someone closer to the right, natural gas is obviously good, but unfortunately opposed by the left (I think in the spirit that the good (gas) is the enemy of the perfect (solar, wind, tidal, etc.)).

    Regarding research supply and demand, my (admittedly anectodal) experience at MIT is that the willigness to spend federal money leads to creation of initiatives like the MIT Energy Initiative (MITEI). Without the backing of hard data here, it seems like MIT combined a sincere interest in energy with the reality that an 'initiative' at a research institution just means an opportunity to obtain (1) federal money, and (2) donor money. (So there is a chance for a multiplier effect).

    More important is the effect such initiatives have on students. By raising awareness and providing programs that e.g. give degrees in energy (MIT has only a minor) you encourage students who, as Romer says, have no way of knowing their employment prospects and may indeed make decisions based on more noble motives e.g. environmentalism.

    Regarding the research you provided on R&D supply response, relating to the above
    Goolsbee: From the abstract, I gather he's talking about wages, not structural changes at institutions as discussed in Romer, so I don't think he contradicts the above
    Romer: I've just skimmed this so far, but it appears that he is talking about economic incentives. Important but not the be-all, end-all. (I could have missed discussion of non-economic motivations though)

    Finally, returning to politics, it's important that prizes are an acceptable tool on both the left and right. McCain had prizes in his platform, ironically, he was made fun of for it. As a reponse to Philo, above, I'd like to point out that there are many fewer incentives for corruption (basically, the worst is that a private company might try to lower the goal to get the money if they think can outcompete on a short time-scale) and many fewer opportunties for bad market distortions, since the government only gets to set the criteria and the prize (i.e. they don't dictate that people must compete for it, that the product be bought etc.) Empirically, these prizes have worked well in the past, as was pointed out

  8. Speaking of success and failure, Evan Soltas, have you ever read the book Adapt by the British economist Tim Harford? Success often starts with failure. He discusses in the book that in order to encourage research, you need a certain model...I forget what exactly however. But I know that there's a certain model that is more conservative to get returns on success, and there's another that tolerates a lot of failure for an important success.