30 April 2007


From the Economist (subscription):

How can the outside world help Africa? There is no easy answer. ... The latest aid-givers' consensus is to identify “good” countries, still quite a small bunch, and let them spend the cash as they see fit. Yet time and again, good guys—most recently, Ethiopia's Meles Zenawi and Uganda's Yoweri Museveni—slip back into old despotic ways, putting aid-givers into a quandary. By punishing governments, are they not hurting the innocent poor?
It seems to me that Africa is crying out for outcome-based policy. A Social Policy Bond regime would inextricably tie rewards to outcomes. In Africa that is an urgent need. Also important is that Africa's problems are so desperate that the continent's wellbeing can be accurately targeted by quantifiable indicators, such as literacy, infant mortality, longevity etc. (That's not always true in the rich world.) We see in Africa (and North Korea ) governments' total cynicism: they are quite happy to bargain away their citizens' lives for a few more years in power. We need ways of mobilizing people to remove government corruption, or to bypass or, failing that, to undermine recalcitrant governments, in order to achieve basic human needs for their people. Social Policy Bonds, whether backed by concerned in the west, or by wealthy individuals or NGOs, would do generate incentives for such a mobilization. The current political system is failing. True, 'there's no easy answer', but motivating people to channel their ingenuity into finding answers has hardly yet been tried.


Anonymous said...

It is hardly surpassing when African leaders act corruptly when there is no effective legal system to hold them to account for their actions.

One solution might be to create a corruption resistant legal system, one that was not funded and staffed by the corrupt government. Once leaders knew that they faced an independent legal system then they may be less likely to engage in corrupt activities. We (ie the donor countries) could tie any aid given to the establishment of such a legal system.

An Australian Reader

Ronnie Horesh said...

Thank you for your comment and interesting an idea. I fear though that to impose an independent legal system on corrupt governments (and it would have to be imposed) could play into the hands of those governments; much like Robert Mugabe or most of the Arab leaders could cynically claim that colonialists or Israel caused all their country's problems. Rallying support against outsiders in this way, though not justified, can be quite successful.

Harald Korneliussen said...

Hmm, I see a potential problem. Imagine that there are bonds targeted to some good indicators of human well-being, and that they are quite ambitious, requiring perhaps a 25% improvement for payoff. The country in question is peaceful, but hopelessly corrupt. Then figure that the bondholders see two options:

1. They can work for gradual change. This will almost certainly work, but it will be very hard to reach the 25% target. Perhaps only a 5-15% increase is feasible in the best case.

2. They can get some trusted people in, stage a coup/revolution, and roll out vast plans of improvement if they succeed. If it succeeds, the 25% goal will be met, but it only has a 10% chance of succeeding. If it fails, the country will be thrown into bloody civil war and horrible misery.

To the bondholders, the second case may seem more desirable, and that's not nice.

It seems to me that all bonds based on threshold levels (such as I believe your suggestion for Afghan literacy bonds were?) are vulnerable to this, but it could perhaps be avoided if payoff was linked to the indicators in such a way that the holders would be guaranteed some payoff, unless they completely botched it.

I do not think this is an academic question. It seems foreign aid specialists face similar questions all the time, in deciding if sacrificing the best for the good is worth it.

Ronnie Horesh said...

Thanks Harald, I think you are correct in that this is not just an academic question. Social Policy Bonds would be radically different from the current system, and such potential pitfalls would need to be anticipated as much as possible. Issuers I think, at least for the first bond issues, would need to have provisos that stipulate that certain other indicators (such as one for violent political conflict, your example) shall not rise or fall by a certain specified level if the bonds are to be redeemed. Another possibility is that at first relatively unambitious targets would be chosen, so that bond issuers could see how progress is going before targeting bigger goals.

As you say, foreign aid organizations have similar problems in today's regime; as do foreign governments that genuinely want to improve conditions for the world's poor or to overthrow oppressive regimes. Being explicit about goals is one way in which minds would then be concentrated on the tradeoffs that may or may not be involved.

Harald Korneliussen said...

"Issuers I think, at least for the first bond issues, would need to have provisos that stipulate that certain other indicators (such as one for violent political conflict, your example) shall not rise or fall by a certain specified level if the bonds are to be redeemed."

No, that wouldn't work. Violent conflict would only come if option 2 went bad, and then the bondholders would have lost anyway. My point is that bondholders could still want to take the chance at option 2, if there was a treshold level for payoff, rather than a propotional one.

Ronnie Horesh said...

(Revised) Thanks again for making me think about this. Actually the provisos would be embodied in the bonds, so that at the time of issue, potential investors would know that if violent political conflict were to occur (for any reason) the bond issue would be invalid.

The problems you forsesee though could arise with negative-but-unanticipated actions. This reminds me of the incentives to fund managers to take big risks with their clients' capital in order to have a higher chance of coming top of the tables in terms of returns. In both cases, the costs of the downside do not fall on the risk-takers.

This concern does seem to point to the need for either smaller incremental increases in targets, or to proportional payouts (rather than thresholds), at least until the Social Policy Bond principle has been tried and refined, and at least for those goals where it cannot be assumed that bondholders would follow the spirit of the bond issuers, rather than the letter.

But perhaps Social Policy Bonds with ambitious goals in practice would turn out to generate proportional payouts. Very ambitious goals would mean a very low market value for the bonds. People buying the bonds therefore could multiply their investment simply by, for instance, raising literacy in Pakistan by 5 per cent, instead of (say) 50 per cent. I think this initial group of investors would probably be unwilling to jeopardise the chance of such a large gain for the much smaller chance of a spectacularly large gain if they were to organize a coup along the lines you imagine. The same reasoning would apply all along the path to redemption, especially if bondholders are a loose and shifting coalition of interests.