The uprising index, explained

February 25, 2011

The Uprising Index Chrystia refers to in this week’s column ranks 80 countries on the likelihood of a domestic uprising based on the average of four equally-weighted factors: corruption; vulnerability to rising food prices; political freedom; and internet penetration.  Our thesis is that an uprising is more likely in a country if corruption is high, if rising food prices have a big effect on a country’s economy, if political freedom is low, and if internet penetration is high.  After crunching the data, here are the 25 countries that scored highest by our measure (out of a maximum score of 1):

Uprising Index

As Chrystia noted in her column, this is a back-of-the-envelope calculation that’s meant to be suggestive and provocative, not definitive.  We limited our sample to the 80 countries for which we had data on vulnerability to rising food prices, and this excluded a few places that seem like they ought to have a high latent potential for rebellion, such as Iran, Jordan, and Cuba.

There are plenty of quants out there creating models that will predict the next uprising—the Political Instability Task Force has a model that predicts instability with over 80% accuracy over the period from 1955 to 2003.  One analyst I talked to compared this kind of approach to the search for “El Dorado:” attractive and desirable, yet elusive.

The raw data used to construct the index comes from the following sources:

  • Corruption data came from Transparency International’s 2010 Corruption Perception Index, available here.  Transprency International uses business opinion surveys and other assessments of bribery, kickbacks, embezzlement, and anti-corruption efforts to compute its index.  We divided the countries into deciles ranking from 0.1 to 1.0, with 1.0 being the decile of countries that are perceived to be the most corrupt.
  • Data on the vulnerability to rising food prices came from Nomura’s September 2010 Food Vulnerability index, which you can view here.  Nomura constructed their index using three components: nominal per capita GDP in U.S. dollars at market exchange rates; the share of food in total household consumption; and net food exports as a percentage of GDP.  Once again we divided the countries into deciles based on their score and gave them a ranking from 0.1 to 1.0, with 1.0 being the decile of countries most vulnerable to a food-price shock.
  • Political freedom data came from Freedom House’s 2010 Freedom in the World survey, available here.  Freedom House uses two surveys, one for political rights and one for civil liberties, to rank countries on a scale of how much political freedom their people enjoy.  Once again we divided the countries into deciles based on their score and gave them a ranking from 0.1 to 1.0, with 1.0 being the decile of countries that are the least free.
  • Internet penetration data came from 2009 estimates from the International Telecommunication Union on the number of internet users per 100 inhabitants in a country, available here.  Once again we divided the countries into deciles based on their score and gave them a ranking from 0.1 to 1.0, with 1.0 being the decile of countries that have the highest estimated number of internet users per 200 inhabitants.

We encourage you to download our spreadsheet, play around with the data, and offer your suggestions for how we could improve it.  One way to refine this index would be to get the data for the past decade and backtest to see if it would have predicted any of the color revolutions of the 2000s. Be sure to leave your ideas in the comments here.

Special thanks go to Steve Davis, Jeff Friedman, and Paul Swartz for their advice and assistance with creating the index.

Posted by Peter Rudegeair.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

You might consider adding another factor: availability of safe drinking water.

Posted by JohnF500 | Report as abusive

I actually did some under-graduate work in this area a few years ago. I’d be interested in looking back at this in a a few years to see how accurate the vulnerability to food prices is, my understanding is that it is difficult to get an accurate account of GDPPC and other economic indicators from regions where political instability is a concern.

I’d also be interested to break out GDPPC and see how it compares to the vulnerability to food price vulnerability in the past, and maybe working in JohnF500’s suggestion of looking at water availability/prices since there have been a number of studies I am aware of that have looked at GDPPC, but relatively few looking at food and water price/availability.

Looking at internet penetration in this way is novel, I’ll be very interested to see how well it works out as a predictor especially since it is at least somewhat correlated to higher GDPPC which would tend to indicate greater stability. It would however be difficult to look back at the importance of internet penetration since even just two or three years ago it was likely significantly lower, so we may see it become an increasingly important predictor in the near future, but not much earlier.

One factor to look at which may be easier to judge than internet penetration would be cell phone penetration, which will be much higher and less strongly correlated to economic indicators in the recent past (last ten to fifteen years or so) and provide a similar level of connectivity to the outside world and amongst citizens.
(Remember the lesson of former Philippine president Estrada in January of 2001)

For anyone interested, I would also suggest checking out the work of Brandt et. al. on “Real Time, Time Series Forecasting of Political Conflict”, they use TABARI/CAMEO to trawl news articles for relevant events and construct predictive models of various types of political instability, I think the actually study is still on-going, but the proposal contains a good look at their reasoning/methodology as well as some excellent cited works and other resources.

Posted by Zephyr256k | Report as abusive

I have personal experience with Vietnam. It seems quite stable. I don’t see much chance at all of there being trouble there. Would love to be proven wrong.

Posted by BobDobbs | Report as abusive

It looks like Chrystia left out the USA

Or did “freedom house” leave it off their list?

They didnt predict OccupyWallStreet?

Posted by b.welldone | Report as abusive