Otto van Bismarck, who knew a bit about political risk, said the thing about revolutions was to intuit God’s movements in history and seize the hem of his garment as he sweeps by.
Easier said than done, of course.
Back in December we posted Eurasia Group’s top ten risks for 2011. Here’s a quick reminder:
1. G-Zero world
2. Europe
3. Cybersecurity and geopolitics
4. China
5. North Korea
6. Capital Controls
7. US gridlock
8. Pakistan
9. Mexico
10. Emerging markets
A week or so later we asked a senior executive at the firm why there was no direct mention of MENA in the ranking. We were told it was because there was only a small amount of perceived contagion risk. Lebanon might fracture but that wouldn’t lead to further instability, for example.
But as we’ve all noticed, God’s garment has been swishing by rather fast of late.
Political forecasting — like its economic sibling — is hard. Eurasia’s analysis is almost always insightful, and unlike some other firms it attaches probability estimates to its predictions rather than hedging its language with “may”s, “could”s or “should”s. Further, forecasting per se is only part of what customers pay for. Analysis of ongoing political debates and regulatory changes are also important.
But we wonder whether we’re any better at this type of forecasting than in 1989, when events in Germany, Poland and Czechoslovakia took most analysts by surprise?
Missing nascent signs of Arab revolution — should such things exist — was far from a unique error, as confirmed in a presentation released Monday by Citigroup’s Senior Global Political Analyst Tina Fordham. She shows ratings from the Economist Intelligence Unit’s Risk Assessment Model, as of January 2011.
But there’s little reflection about why this happened and what political risk analysts should be doing to change their assumptions and models. Instead, the presentation looks at MENA’s current state and what might happen next.
In February 1790 Edmund Burke used a speech in Parliament to predict doom for the French revolutionaries. A couple of centuries later we’ll have to make do with this powerpoint slide:
That’s a little cruel, we know. But it’s odd to begin with the non-linearity of political transition and use an analytical tool that is inherently linear. (For more on this, see Edward Tufte.)
Have we really not improved since 1989? This is not a political science blog, but it seems that despite advances in understanding the causes of conflict and the rise of political prediction markets, we’re still not very good at appreciating the mechanics of political change.
However, one possibility is that in an effort to quantify, we neglect important evidence that is harder to measure. In other words, we’ve stretched too far to apply economic models to politics, and quantitative bias is the result. This is the point made in an otherwise complementary review by Boston University’s Jalal Alamgir of the Economist’s “shoe-thrower’s index“:
There are two main weaknesses in this index. First, as The Economist itself admits, it discards factors that are “hard to quantify,” including unemployment information because they’re not comparable across the countries in the region. This quantitative bias is typical of many risk approaches. More qualitative factors such as ideological motivation or support (such as between Islamism or secularism), leanings of leaders (such as between non-violence or violence), leanings of the armed forces, control over governmental employees, and ideas of justice/injustice are important predictors of not just the occurrence of instability but the duration and extent of it.
The second important factor, which can be both quantitiative and qualitative, not included in the index is “resources.” Political scientists have shown that revolts, and specifically democracy movements, are critically dependent on organizational, technological, and infrastructural resources available to protesters. Simply put, without access to technology, such as Twitter, Facebook, or satellite TV channels, all the other “factors” may not have produced the type of instability that is sweeping through the region. Resources allow isolated show-throwing to snowball into concerted political upheaval.
We don’t know, but this seems right and it’s clear that some of the models need changing — much like economic models following the financial crisis.
Here’s one project that’s trying to do so — let us know if you find any more.
Related links:
Geopolitics is final piece in risk jigsaw – Gillian Tett
The Arab World’s 1989? Who’s next? – Power & Policy
Measuring political risk in the Middle East – Globalization, Risk and Forecasting
The shoe-thrower’s index – The Economist