Statistics and International Relations: Why Data Collection Matters

With “Looking toward the Future in the Taiwan Strait: Generational Politics in Taiwan,” Shelley Rigger demonstrates the crucial role of data collection in international relations. Had there been no provisions to run surveys and collect relevant data over the decades, there would have been no means of discovering the changing patterns of the Taiwanese identity. This, in turn, can help align foreign policy with domestic opinion. Over the years, there has been a tremendous shift in the Taiwanese people’s sense of ethnic identity: evolving from “Chinese” to “both Taiwanese and Chinese” and, finally, the most common response today, “Taiwanese.” If pre-1970s public opinion might justify the concept of “recovering the [Chinese] mainland,” the same concept can no longer be a guiding factor in cross-Strait relations given this new understanding of shifting the Taiwanese identity.

Another example of how data helps understand phenomena that can ultimately guide both internal and external policy making is the work of the Levada Center in Russia. Previously called VCIOM, the Levada Center is an independent, non-governmental polling and sociological research organization in Russia.¹ Since 1987, the Center has been carrying out surveys on various topics such as electoral preferences, Western values and democracy, and the opinion of the Russian middle class.²

In his recent book, The Return: Russia’s Journey from Gorbachev to Medvedev,³ Daniel Treisman relies heavily on Levada Center data. In analyzing this data he identifies a strong relationship between the index of economic sentiment—a measure of how people perceive current and future economic conditions—and the popularity of the ruling leader.⁴ This helps explain why Boris Yeltsin’s popularity ratings slipped with falling oil prices, and moreover, how Vladimir Putin derives most of his invincible aura from the fortunate timing of his presidential term, which coincided with a soaring international price for oil and Russia’s steady recovery after the 1998 default.

Another survey from the Levada Center gives a very interesting insight into Russia’s present political sentiment. When asked, “What is more important: order, even if it leads to the violation of democratic principles; or democracy, even if it gives freedom to destructive elements?” an overwhelming 81 percent percent respondents voted in favor of order in April 2000, with the remainder being split between democracy (9 percent) and difficult to answer (10 percent). The same numbers in December 2010 showed an increase in pro-democracy sentiment: order (56 percent) still won the majority vote, but democracy (23 percent) and difficult to answer (21 percent) made gains.⁵ This should serve as a warning to the pseudo-democratic Medvedev-Putin tandem: while a large majority of Russians still prioritize order over democracy, there has been a significant shift in favor of democracy over the past decade.

Such surveys can be particularly useful in conflict management as well. As a citizen of India, I have been exposed to a range of views on India-Pakistan relations, from (a) my grandparents’ generation’s considerable animosity against Pakistan because of the partition; (b) my parents’ generation’s relative neutrality; and (c) my own generation that looks for ways to build people-to-people contact, and is more open to bridging the gap with Pakistan. But without data from surveys, like the ones Rigger uses, it is impossible to determine whether this is my personal experience or the broader consensus. If there were a way to prove that the bitterness between the countries is now restricted to the cricket field, at least among the younger generation, perhaps politicians would not so readily blame India’s neighbor for what goes wrong within domestic borders.

Data collection and interpretation methods are not bulletproof. Errors like selection biases (when the sample surveyed is not reflective of the overall population), or observational errors (when a response is incorrectly recorded, or when a respondent gives an incorrect response) are bound to affect findings. Treisman himself cautions that the index of economic sentiment in Russia could have been affected by state propaganda, and was not necessarily outside the realm of influence of the head of state.⁴

Still...