Big data allows us to quantify political risk and we must do so, in order to avoid pointless attempted coups and excessive political repression. It may also be useful for investors.
There’s some really interesting data on Africa in this report – including the 96% annual chance that Nigeria retains the democracy label it won after the important 2015 elections (and that “failed state” risk is very low indeed), and the 97% chance that SA remains a democracy despite fears of many that it may end up following Zimbabwe’s path to greater poverty and less democracy. Both look more solid countries than many think. Political risks are three times higher in other African countries than Nigeria – as we highlight below, from Ivory Coast to Mozambique. Meanwhile annual democratisation risk is much higher than we expected in Egypt (13% a year) and Morocco (9%).
The best dancer at a wedding I attended last month was an open-minded member of China’s Communist Party who I tested this report out on. He thinks it’s wrong. China’s communist party will be in power for another 100 years he reckoned, as we have yet to see the third generation of family rule – and Chinese dynasties don’t only last two generations. Data tells me that China is changing because of rising income levels, and that he is making the same mistake as those Turkish military officers who in July said that military coups have generally succeeded in Turkey, so will succeed in 2016. They are trapped by their understanding of history, and their lack of understanding of the power of the middle class.
In yesterday’s media, we saw three stories about democracy and autocracy, from the pro-democracy groups winning seats in Hong Kong’s elections, to Gideon Rachman’s piece asking a set of fascinating questions about politics and geo-politics to the news from Russia, ahead of the parliamentary elections on Sunday 18th September. Russia has stopped the Levada Centre, the only independent pollster (who we met at our 20th annual conference in June) from producing polls – as it is now a foreign agent.
Our new report has something to say on all these issues, and many others.
1) We reiterate our 2011 argument that all countries become democracies as they get rich, except those countries that export a lot of oil per capita (see fig 2 in the report – because Russia does not) and the city state of Singapore. Rachman thinks the “end of history” thesis only lasted 20 years. I think he’s wrong. Look at fig 3 – which shows all the countries in the world that are not democracies. Above $14k of per capita GDP, only large oil exporters, Belarus and Singapore, and above $21k, only large oil exporters and Singapore, are not democracies. Where then is the successful challenge to the “end of history thesis” that all countries eventually become rich and democratic? This table also answers Rachman’s question about why poor China in 1989 stayed autocratic while wealthier central Europe democratised.
2) We show that no countries ever lose a positive Polity IV rating once they have a high per capita GDP. The Turkish coup attempt could not succeed – or to be 100% accurate, we have never (yet) seen a country lose democracy at Turkey’s income level which is over $15k.
The percentage chance of a country losing a positive Polity IV score (full democracy, democracy or open anocracy) and getting a negative score (autocracy, closed anocracy) at certain per capita GDP thresholds (2011 PPP dollars).
Source: Polity IV, Penn table, Renaissance Capital
3) We examine the countries that have seen a deterioration in their democratic rating –Turkey, Russia and Malaysiawere all more democratic a decade ago than they are today. We highlight that as all rich countries are democracies or oil exporting autocracies, and none occupy the middle ground (Singapore is an exception), this tells us that
a) Russia, Turkey and Malaysiawill remain “open anocracies” but stagnate economicallyand remain middle income or
b) They become “democracies” – a surprisingly high 10% chance each year (but thedata sample is too small for us to be confident about this) or
c) They do grow fast, while remaining “open anocracies” – which has never happenedbefore.
4) We highlight how the oil price crashed in 1985 and triggered a wave of democratisation among countries that exported only a little oil per capita (from the USSR to Mexico). We may see a new wave of democratisation since the oil price crash of 2014-15, and suggest democratisation may occur in 2018-20 inRussia, Iran and Venezuela. See figure 16.
5) Political regimes are actually very stable. There is an 86-100% chance each year that the current regime will remain in place next year. The greatest risk of significant change is in the poorer countries. See fig 25 for EM, fig 26 for Frontier and fig 27 for Beyond Frontier – to see the chances of political change in the coming year. African countries fromIvory Coast, Mozambique, Tanzania and Zimbabwehave a 1 in 7 chance of a big change – with more risk of a shift towards autocracy than democracy. Nigeria and Kenya are far more stable, with a 96% and 95% chance respectively of remaining a “democracy”.
6) Democratisation is coming toChina, Vietnam, Morocco, Egypt and Jordan. China is already at the income level where there is a 4.5% chance each year of a shift towards democracy. This risk will grow to 8% in the 2020s, when per capita GDP is higher. Vietnam is likely to lag China. Morocco already carries a 9% chance annually of a further shift towards democracy, while Jordan and Egypt have a surprisingly high 13% chance of such a shift in any given year. Democratisation at high income levels tends to be relatively peaceful.
7) The Pinochet argument– that autocracy is better at low levels of GDP per capita – is undermined by our data analysis. See figure 19 and fig 20. Some autocracies do grow faster than democracies, but a number from Cuba to Tajikistan do not.
8) However, we do find that there is some correlation between ease of doing business, corruption, legal scores and political system – the extremes of democracy and autocracy tend to favour better results than those with very mixed systems. See figures 21-23.
9)The demographic argument– that lots of young people are destabilising for any political regime – is supported by our data. However, we found that too many young people (more than 30% of the population are 15-29), as well as too few (25% or less), suggests no change. It is 25-30% which is the key level that encourages change. China and Russia might be at less risk of political change if they get old before they get rich. See figures 29 and 30.
We have not looked at all the market implications of political change. This was a report which in 37 pages aims to show what will happen. We need to produce another report to look at the market implications of this. One challenge will be the fact that we have more political history than financial history for many emerging and frontier markets.
But one conclusion we would highlight is that EM and Frontier markets are more politically stable than we might think – and the risks some of us remember from the 1990s when Russia’s military was shelling its parliament and Turkish tanks provoked the fall of a government – are now risks that are concentrated in beyond frontier markets like Zimbabwe, rather than larger markets.