KM on a dollar a day

Musing on knowledge management, aid and development with limited resources

Poker face: betting on development

with 2 comments

Duncan Green recently wrote  Honduras is building a charter city? This is never going to work which as you can tell by the title is rather skeptical about the attempt in Honduras to build a charter city along the lines proposed by Paul Romer. In the opening paragraph we offers readers to make a wager that it doesn’t work, such is his confidence.

Roving Bandit offers (kind of) to take him up on his challenge by correctly pointing out that just like in betting on poker sometimes it’s worth while betting on a long shot, if the potential reward is big enough and you have enough resources to  bet on a long shot without losing everything.

“So what odds do you say Duncan? I’ll give you £10 if it fails and you give me a £1000 if it works?”

This is a very important observation because if you never bet on something untested and risky, then you would rarely be willing to invest in something new, untested and potentially risky but which could potentially yield large benefits –   exactly the kind of challenge we face in development aid.

That said betting on a long shot in poker is different from doing it in the real world in two important respects:

  1. In poker you know whether you have won the hand – the rules of success are clear.
  2. In poker you might not know what cards the other player has, but you can calculate the odds and use them to guide your bids (at least according to those poker tournament shows they have on the teevee).

So going back to the Guatemala Charter City idea, although different people will have different opinions about the likely success of this enterprise, we don’t have a very clear picture neither of the real odds or the potential payoff. In addition if I were to take Duncan or Lee’s bet, how would I know if I had won or lost – what is the definition of success?

All this is a very roundabout way for me to come to the under-exploited potential of “prediction markets” for development. What if you could run a book on the potential outcome of a project investment? If you could get enough people to take bets on whether this (or any other project or initiative) is successful then you might get a much more accurate sense of the likely risk and potential benefit of this project. In order to so this you would also have to come up with a specific definition of success in terms of what happens by when.

As well as being fun for people to place a little wager, this type of approach could also tell us a lot about where to place our investments in development. Interestingly enough, while we often take our pilots projects to be make or break, in reality a failure or success of a pilot, including the current one on charter cities, doesn’t tell us whether an approach is right or wrong, it just helps us reevaluate the odds.

(for the record I agree with Duncan – but what do I know?)


Written by Ian Thorpe

December 23, 2011 at 8:50 am

2 Responses

Subscribe to comments with RSS.

  1. Hi Duncan. Three concepts that I think are relevant here.

    1 • Prior and Informed Consent: Do the Honduran people know what they’re getting into? I doubt it, because I think the promoters have little idea of what they’re getting into.

    2 • Sophisticated Investor: High risk investments are for sophisticated investors who understand the risks, and can afford the losses. Typically, people in poverty are not sophisticated investors.

    3 • Though mathematical gambling suggests that to win or lose on a 50/50 bet is the same, real people act “irrationally”, by preferring to avoid losses. But mathematics of gambling was designed by mathematicians, whereas human biases are the product of evolution. Double your crop, you just have bunch of rotting extra food; lose your crop and your dead. In real life, the “irrational” risk-averse response is much smarter.

    Why do Westerners always think that if they have an untested idea, the place to try it out is on some poor people. Why don’t they try it out in the Hamptons? (I’m sure we could figure out a suitable version.)

    Where’s the ethics of experimenting on the poor? Didn’t we just apologise for using them for medical research in the 1950s? Are we going to learn some day?

    At minimum, we should also expect that all the proponents, researchers, consultants and financial backers put their own life savings on the same bet. If the charter city doesn’t work, they give up all their assets which are transfered to the Hondurans as compensation; if the charter city works, they get double their money for all the great work they’ve done.

    Let’s see if they take the bet.

    David Week

    December 25, 2011 at 5:48 am

  2. David – good points. Lee’s post actually does mention the issue of opportunity costs in which you need to consider the size of the bet in terms of whether or not you can afford to take the risk. The comments do also point out that in this specific case no-one is forced to move to the charter city.

    While I agree with you about westerners not trying their untested ideas on “poor people” as a general point, especially given the very real opportunity costs of failure, many kinds of anti-poverty interventions wouldn’t make sense in the Hamptons and thus you wouldn’t learn anything useful about improving the lives of the poor by testing them there. For some things the only way to test them would be to test them on “poor people” – however this should be done with consent, and taking into account the opportunity costs of failure. If you couldn’t test anti-poverty ideas on the poor then you would be faced with either i) not taking any new initiatives i.e just stick with whatever you are already doing however well it’s working or ii) don’t test things at all if you can get them approved through local political approval – just go ahead and do them at full scale. I’m not sure I think either of these options is preferable.

    I introduced the idea of prediction markets here because I think that some sort of side-betting would give us a helpful measure of how likely something is to work, and whether or not the real life risk is worth taking.

    (P.S. I’m personally not that persuaded by the idea of charter cities but I probably know much less about them than other commentators – that’s why I’d like to see what the wisdom of crowds might tell us about the odds of success)

    Ian Thorpe

    December 27, 2011 at 9:41 am

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: