Her report has details on all presentations, but I only excerpt her description of my session below. (If you want to see just how academics went from theory to bribery in the session, you can watch the video.)

[begin write-up:] Theories are all well and good, but real life doesn’t always behave. Since David Zetland, an economist at UC Berkeley, does most of his work in the muck and hubbub of the real world, he wanted to ground the discussion in reality by performing an experiment. The conference participants negotiated Kyoto II on the spot.

Zetland assigned the participants to four groups, A, B, C and D. The treaty was bilateral, to be negotiated between America (the A group) and China (the C group). The bureaucrats formed the B group, and they gathered information and administered agreements. The D group was “da rest of the world,” who were mere observers but could lobby freely.

China presented the first proposal: a global cap-and-trade system, with emissions permits allocated according to population. In addition, China wanted to receive permits for projects done in other parts of the world with clean technology it develops.

America responded with its own proposal, a global carbon tax. Ideally, the US said, revenues from the tax would go entirely to the US, but since it recognized that not everyone would see the manifest virtues of such an arrangement, dividing the revenues according to gross domestic product would be acceptable. In addition, a fraction of the global revenue should go into research and development (which would take place primarily in the United States). Countries who participate in the tax would receive discounts on the technology produced, while those who didn’t would be subject to trade sanctions. As a further stick, the US pointed out that it had hostages, in the form of China’s money (in treasury bonds) and people (its graduate students at American universities).

The rest of the world said it was willing to negotiate. It wanted to get money to do nothing. They held a non-binding vote and supported the China proposal, six-to-two.

America came back with a counter-proposal: Cap-and-trade was acceptable, as long as the permits were allocated according to GDP. China objected that America’s history of carbon production needed to be taken into account. Then the two sides retreated to consider their strategies.

In the interim, the bureaucrats (whom Zetland himself led) decided on a change of rules. The bureaucrats and the rest of the world would vote on the proposals, and the members of the winning delegation would be rewarded with being first in line to get wine during the reception afterward.

China proposed cap-and-trade with permits allocated by both population and GDP: 80 percent by population, 20 percent by GDP. The bureaucrats jumped in and said they’d support any proposal in which they managed the funds (with appropriate fees, of course). America responded by accepting cap-and-trade, but having the proportions switched: 80 percent by GDP, 20 percent by population. Over 50 years, though, the proportions could gradually move to China’s proposal of 80 percent by population and 20 percent by GDP. Administration would be by the bureaucrats. In addition, the Americans offered a bribe to five of the eight countries constitutingthe rest of the world for the purposes of the experiment: annual international aid of 30 percent of their GDP. The bureaucrats calculated that this would cost the US 10 trillion dollars annually, or 2/3 of US GDP.

China said that the US proposal was fine, with one change: the transition in the allocation scheme of the permits should take place over 20 years rather than 50. The US counter-proposed 40 years, and insisted that China take the burden of half the cost of the bribe. The Chinese didn’t agree, so it went to a vote.

Bottom Line: Bribery worked. The US won.

While the game was played with tongue in cheek, Zetland had a serious purpose, which turned out to be beautifully illustrated by the game’s unexpected conclusion. Human beings behave in unpredictable ways. To be useful guides, mathematical models have to allow for human uncertainty.