Reward the frugal and charge the profligate

Emission permits trading is said to offer a solution to greenhouse gas emissions. Under such a scheme an organisation will require a special permit before releasing greenhouse gases. Permits are allocated, sold to, or traded between organisations whose activities cause greenhouse gases. The effect is to increase the price of goods whose production causes the emissions of greenhouse gases. Other ways of producing the same goods with less gas emissions will naturally arise because they will become price competitive.

At first sight this line of reasoning appears to offer a sound solution to the problem of reducing greenhouse gas emissions.

I recently asked a number of experts why it is going to take until 2010 to set up an Australian emissions trading system. This seems a long time to wait given the immediacy of the problem.

I was told the reason for the delay is that setting up an emissions permits system is difficult, with the major problems including defining the property rights associated with permits; how to issue permits; what to do with the money raised from permits; enforcing the property rights; and understanding the effect of the system on the economy.

This is going to take at least two years and we are still unsure whether, at the end of that time, it will achieve a reduction in greenhouse gases.

As someone who builds information systems for a living I know that any information system (and an emission permits system is an information system) that takes two years to define is not going to work as expected. An emissions trading system is complex. We know that it is best to build complex information systems incrementally. We start with the simplest possible system that will achieve the minimum usable objective, we build it and see what happens. We learn from our mistakes, make incremental changes and allow the system to evolve as we add increasing complexity.

I then asked how the system was going to be tested. I was told that this was done by modelling the system on a computer. Economic computer modelling is carried out by building a theoretical model on how an economy works.

Unfortunately the record of computer models of economic systems is problematic when they are asked to include human behaviour. It means we are testing how emissions trading will work with simple computer models not with humans. This is almost guaranteed to give results that will be difficult to trust because human systems are not static computer models. People invent new rules and continually change the systems they use.

This then set me thinking about the objective of emissions trading and to think of a better way to achieve the objective of reducing greenhouse gas emissions through an incremental adjustment of existing economic systems rather than creating a whole new regime of property rights.

The underlying purpose of emissions trading is to direct investment to ways of producing goods without greenhouse gas emissions (like energy).

A solution to this starts by putting a price (any price) on carbon and to use the money raised to invest in ways to produce goods (like energy) without creating greenhouse gases.

This is a well known and standard approach to the problem of reducing greenhouse gases. The argument against this is that it is not market-driven. This is a problem because markets are acknowledged as the best way to efficiently allocate resources. The market for the goods produced already exists, as does a market in greenhouse-reducing infrastructure, so there is no need to invent any new market to distribute resources on infrastructure to reduce greenhouse gases.