Carbon Markets in the Laboratory

Markets are a powerful policy tool. The market mechanism allocates scarce resources to their highest value users, so enabling a policy objective such as a cap on overall water use or carbon emissions to be met at the lowest possible cost to society. By their very nature, markets will follow their own path. A
well functioning market will find the genuinely efficient price for a scarce good – and will allocate resources optimally in the process. The attraction to the policy maker is that there is less need to know the internal (and hence basically unknowable) circumstances of the regulated parties in order to create
efficient policy. However, this also means that market outcomes (prices achieved, innovations brought forward) cannot be predicted with certainty, and are likely to differ from the outcomes that would be reached by non-market regulatory policies.

There is currently great interest in the process of designing and implementing markets for public policy due to the forthcoming Australian Emissions Trading Scheme (AETS) for greenhouse gas emissions. There is much debate and lobbying around designing a set of market rules which will work ‘best’. As the latest discussion paper from the Garnaut Review points out, these rules should be simple and consistent. A growing body of evidence from existing environmental markets, as well as laboratory market simulations, shows that the details of market design can have a very significant impact on market outcomes.
Not all markets work equally well. The more complex the rules, the harder it is to predict exactly how a market will work, and the greater the chances of it failing in its task of efficiently allocating scarce resources. Therefore for any market-based policy, and particularly one of such scope and impact as the AETS, great attention must be paid to the details.

Economic theory provides a starting point for market design. However people frequently do not conform to the narrow assumptions of economic models, which can lead to unanticipated consequences.
Sound market design therefore requires real human behaviour to be incorporated into economics. One approach to doing this is experimental economics (pioneered by Nobel laureate Vernon Smith), which observes economic decision-making under controlled laboratory conditions. In a typical economic experiment, participants trade in a simulated market through computer terminals. To increase the realism there is real money at stake – participants are paid based on their success in the market.

While experiments are by no means the final word on economic behaviour, they provide many insights which can guide policy design. The first and most crucial is confirmation that, when exposed to the vagaries of human behaviour, the details of market design really matter.
For example, the ‘irrational exuberance’ which leads to asset price bubbles and subsequent crashes is readily observed even in simple laboratory markets in which people are trading artificial assets with clearly defined values. Modifying the rules of the market to improve availability of information about supply and demand, for instance by incorporating a simple forward (futures) market, can dramatically reduce this volatility (1). In a forward market participants can agree on prices for future trades, which effectively provides advance warning of future market conditions.

In terms of an AETS, the Garnaut Review recognises the importance of forward markets in providing sound price signals and hence guiding medium and long term investments in alternative technologies. To some extent forward markets emerge spontaneously – in fact, some market participants (particularly
companies involved in electricity generation and retailing) are already trading forward against the anticipated compliance units, even though they have yet to be defined (or even named). For a forward market to function well, information about future permit availability must be made available as openly
and accurately as possible. Dysfunctional forward markets will distort spot (current) markets and lead to poor investment decisions.

A similar issue arises with the creation of offsets and the banking and borrowing of permits proposed by
Garnaut. The number of permits in the market, needs to be publicly known to guide future investment. Experimental simulations of the Mandatory Renewable Energy Target (MRET) market suggest that a lack of transparency in the number of permits created is likely to have impaired market performance (2). When market participants were able to secretly hoard permits, many did so, resulting in low supply and rising prices early on, followed by a price collapse as it became evident there was an oversupply. Similar
patterns have been seen in the EU ETS and the NSW Greenhouse Gas Abatement Scheme (GGAS) markets, in part due to uncertainties around the number of permits actually in the market.

Another key market design question is the initial allocation of permits. Garnaut indicates that there is no case for freely allocating permits, so they should therefore be auctioned (with compensation for trade-exposed industries and adversely affected regions and households). In addition to its public finance merits, this proposal also works from a behavioural economics perspective. Experiments show that people tend to place a higher value on items they hold than on those they do not (3) – in behavioural economics, it appears a bird in the hand really is worth two in the bush. As a result people prove poor at trading items which they have been given for free. Running an initial auction forces all market participants to consider their optimal position and place a reasoned value on permits, and is likely to lead to more ‘rational’ and efficient subsequent trading activity.

Psychologically people relate very differently to losses and gains (4). The prospect of incurring a loss has a far greater mental impact than the prospect of an equivalent profit. Whether an outcome is considered as a loss or a gain depends on the way the decision is framed. In laboratory experiments on simulated emissions trading schemes, people will go to far greater lengths to avoid paying a penalty for being non-compliant than to make a profit from being compliant, even though the financial consequences are identical. Paying a ‘penalty’ is not only seen as a loss, it also implies bad behaviour, which has reputational consequences. This may explain the apparently irrational behaviour seen in the NSW GGAS market in which some market participants were actually prepared to pay slightly above the penalty rate in order to avoid being non-compliant (which also created a significant problem in the forward market – see 5). Therefore the way in which any sanctioning mechanism is labelled (e.g. penalty, levy, make-good provision) is likely to impact on market behaviour.

Great care should be taken around the design of any initial permit auction within an AETS. There are parallels here with the auctions run by governments around the world for 3G mobile phone licences in 2000-2001. Each government opted for a slightly different auction design, which contributed to the vast differences in revenue raised. Once again the details really matter, and the dividends from thorough design and testing can be huge. The economics laboratory provides a ‘policy wind tunnel’ in which
proposed market mechanisms can be tested and refined, reducing the risk of costly policy failures. There will certainly be surprises in store in terms of how people and firms, individually and collectively respond to the new market signals arising from an AETS.

A final thought concerns the future of voluntary abatement activities and carbon markets as we move into a statutory AETS. Simple economic theory would suggest that introducing additional incentives through an AETS can only increase people’s abatement activities. However quite the opposite may be the case, and there is evidence to suggest that introducing regulations or financial incentives can lead people to stop doing things voluntarily (6). If voluntary carbon markets are to have a future they will need to be very clearly delineated from an AETS, perhaps by focussing on sectors or activities which fall beyond the scope of the statutory market. While many people donate to medical charities, no one donates to Medicare. Why should carbon be any different?

1)
See Porter, D & Smith, VL (1995) Futures contracting and dividend uncertainty in experimental
asset markets, Journal of Business 68, 509-541 and Kluger, BD & Wyatt, SB (1995) Options and efficiency: Some experimental evidence, Review of Quantitative Finance and Accounting 5, 179-201.

2)
Nolles, K (2006) The impact of the delay between energy generation and REC creation in MRET: A
laboratory investigation on the performance of the Australian Renewable Energy Market. Aton
Experimental Economics Laboratory.

3)
Kahneman, D, Knetsch, JL & Thaler, RH (1990) Experimental tests of the endowment effect and the
Coase theorem. Journal of Political Economy 98, 1325-1348

4)
Kahneman, D & Tversky, A (1979) Prospect theory: An analysis of decision under risk. Econometrica
47, 263-291

5)
Nolles, K (2006) How the forward market has raised the penalty rate for MRET, GGAS and GES by 20%. Aton Experimental Economics Laboratory.

6)
Frey, BS (1997) Not Just For the Money: An Economic Theory of Personal Motivation and Reeson,
AF & Tisdell, J (2006) When good incentives go bad: An experimental study of institutions, motivations and crowding out.