Big Data: What does it mean for International Relations?
Book review: Big Data: A Revolution That Will Transform How We Live, Work, and Think. By Viktor Mayer-Schönfelder and Kenneth Cukier. Eamon Dolan/Houghton Mifflin Harcourt, 257 pages. R$54,29 (kindle, amazon.com.br)
There is a growing consensus around the world that "big data" -- the capacity to process unprecedented amounts of information relatively cheaply -- will fundamentally change many different fields. These range from public health (where google searches provide hints of where the next flu outbreak will take place, or how many women opt for self-induced abortion), political campaigns (knowing what voters care about), business (allowing customers to purchase a plane ticket when it is cheapest) and education (knowing exactly in which chapter readers decide to stop reading a book on kindle). In their excellent introduction to the topic, Viktor Mayer-Schönberger, a professor at Oxford University, and Kenneth Cukier, senior editor of The Economist's digital products describe a "shift in mindset how data could be used": Once thought as static or stale, today data can, they say, "become a fountain of innovation and new services."
The authors argue that big data analytics are revolutionizing the way we see and process the world, actually comparing its consequences to those of the Gutenberg printing press. Yet while the impact of ever cheaper and accessible data-crunching technologies on business is now studied by many, still very few wonder how it will change international affairs. There is little doubt that it will matter greatly: Even though their book does not focus on international relations (the largest part is dedicated to business practices), Mayer-Schönberger and Mr. Cukier promise big data will become “part of the solution to pressing global problems like addressing climate change, eradicating disease and fostering good governance and economic development.”
Interestingly enough, Big Data suggests that greater access to large swaths of information will lead us away from our fixation with causality. Rather, we'll be able to understand a lot more about correlation -- i.e., what is happening, but not why. In the same way, big data will reduce an urge to get it exactly right, and make scientists more comfortable with messiness and educated estimates -- as is already the case when dealing with large amounts of data, like the U.S. census. "Census", the authors remind us, comes from the Latin word "censere", which means "to estimate". All that will sound rather counterintuitive to traditional scholars, but the authors make a strong case when they argue that "big data, with its emphasis on comprehensive datasets and messiness, helps us get closer to reality than did our dependence on small data and accuracy."
The social sciences are likely to be among the most affected, as sampling -- a practice born out of the incapacity to reach everyone -- becomes increasingly obsolete. Analyzing millions of calls made provides insights into social structures that would have been impossible before. For example, scholars can now relatively quickly analyze all the statements uttered by U.S. presidents over the past two centuries -- and discover correlations between, say, the use of the word "crisis" and unemployment, GDP growth and inflation figures at the time (something the authors call "recombinant data"). International affairs scholars can discover which countries have most often appeared in U.S. presidents' speeches, and in which context. Legal scholars can map which Supreme Court rulings over the past centuries cited which previous court decisions.
Big data is already influencing the way some government act internationally. Drones -- the central element in 21st century warfare -- already generate millions of bits of information about the location of other nations' submarines, as a new report for Britain's Parliament points out. US drones kill individuals based on "circumstantial evidence" -- i.e., because big data-driven analyses suggests that there is a strong correlation between a series of specific behavioral patterns and membership in a terrorist organization -- even though there is no actual proof of the latter, or any knowledge of why a certain individual behaves this way.
This generates a lot of ethical concerns, as the authors readily recognize. They stress the risk of "the dictatorship of data", pointing out that big data, in theory, could allow police departments to arrest people before they commit a crime, largely because circumstances point into that direction -- Yet this also suggests that big data may allow us to predict important political events -- such as a coup d'état, sectarian violence or genocide -- with greater accuracy, based on people's search requests on google, posts on social media, and purchases online. In an ideal world, the international community could use big data to help create a better early warning system and identify countries at risk. Conversely, genocidal regimes may have an easier time locating members of a certain minority, and repressive dictatorships may use big data to predict where and when subversive movements emerge. The book strikes a noticeably cautious tone -- and criticizes Robert McNamara for overly relying on data when taking key decisions as President of Ford, Secretary of Defense or President of the World Bank.
Of course, for many broader trends in international relations, bid data will be of limited use, as the number of occurrences (such as wars, secessions and diplomatic spats etc.) is simply too small. Yet innovative scholars are still likely to benefit greatly, for example by mining the large amounts of data made available through wikileaks, articles on international affairs, or trade data.
We are at an early stage of the datification revolution, Mayer-Schönfelder and Cukier point out, and it is too soon to tell how exactly it will play out. It seems clear, however, that few parts of life, including the field of International Relations, will remain untouched. Both policy makers and scholars won't be able to escape the profound implications of the rise of big data.