Prediction Markets for the Poor

Look, Obama Won!

In the game show, Who wants to be a Millionaire, contestants have the opportunity to ask the audience, or an expert, if they want to get an answer to a question they can’t answer. Proponents of “crowd theory” are always delighted that the crowd is more accurate than the expert.

However, since the average person can be expected to guess randomly, that means that – in a large group of people – the few who do know the answer will make the crowd appear to know the correct answer by biasing the result.

The same is true of questions about happiness, where it turns out that Cubans are happier than Americans. Yet, if this were true, we would be more likely to see boatloads of Americans washing up on Cuban beaches, than the other way round.

The best way of figuring out what people really want still comes down to what they are prepared to risk their own money or time on.

This has led to the creation of what are called Prediction Markets.

Say a company wants to know whether or not it will be able to ship a new product on time. A virtual stock exchange is set up to allow trades in the opportunity. Those employees who believe that it is possible, buy; those who don’t, sell short. Siemens, using such a system on its own employees, discovered that a new software product wouldn’t ship on time despite assurances from their more mainstream consultants. Google and Microsoft also use such tools to allow their employees to voice their opinions.

However, the approach is much more powerful than that.

The University of Iowa, which is at the forefront of this research having started their Iowa Electronic Market (IEM) back in 1988, has set up epidemic prediction markets. They recently set up an Avian Flu Market to provide an early-warning system to track that disease. This is an extension of an existing market to track influenza. Their Flu Activity Market provides information on when a flu outbreak will start, peak and end in any state. Doctors, who see the majority of sniffles, can buy or sell shares based on what they believe is happening.

IEM also allows people to take positions on the US electoral process and individual candidates. Their accuracy in predicting the last American elections was better than most opinion polls. Their US election market is currently predicting 51.5% of the votes to Barack Obama, and 47% to John McCain.

The systems don’t use “real” money, but produce accurate behaviour none-the-less. It also provides a mechanism which can answer a difficult development question.

What is it that poor people want and believe will do most to support their economic development?

The government believes that land and houses are what the poor are after. The question is, if a poor person was given a quantity of cash, what would they spend it on? The creation of a prediction market to allow sub-economic groups to trade for the goods that they would most like, and which they believe will create more wealth for them, would be a better mechanism for revealing “truth” than the prognostications of those in higher office.

The prediction market also forces the market-creator to think very clearly about what is possible. It is impossible, for instance, to “buy” a job (much as governments try through direct subsidies). You can buy a house, a car, land, a factory, better roads, public transport, infrastructure, security, courts, and, well, a lot of things. But there is no such thing on the market as “economic growth” or “sustainable development”.

So, bright minds at universities around the world, how about it? How about a prediction market for the poor? Let them have the opportunity to influence state spending on their behalf by buying the things they most believe will create wealth for them and their families.