Though thousands of man hours are invested each year into studying international relations, scholars still failed to predict seismic global events. International relations theories are sophisticated yet still prone to inaccuracy.

---

International relations theories are sophisticated yet still prone to inaccuracy. Whether one is speaking of the Cold War or the financial crisis of 2008, scholars failed to predict big events. Demand for accurate political forecasts has existed for as long as politics itself. In recent years, there have been significant changes in the methods used for these forecasts.

 

Not a utopian dream
One high-profile reason for failures in international relations forecasts is that, in contrast to the macro physical world, international politics is highly unstable. Does this mean that accurate predictions of international relations are just a utopian dream?

“The regularity of international relations is the basis for prediction research,” said Chen Dingding, a professor from the School of International Relations at Jinan University in Guangdong Province. Chen argued that although international relations covers sophisticated issues in various fields, such as politics, economics and society, its development and evolution are not random or disordered. From a macro perspective, changes in the international pattern and order are regular. To be specific, as the main actors, nations participate in international affairs based on certain laws, Chen said.

Lu Lingyu, a professor from the School of Politics and Public Management at East China University of Political Science and Law, cited Mark Blyth, a researcher on international political economy at Brown University, as saying the political world was divided into three levels: normal distribution, Poisson distribution and Pareto-Levy Mandle-brot distribution.

At the first level, information about the predicted events is easily collected, so records of the historical events tend to be in the normal distribution range. In international relations, the foreign policy of mature countries and the behavior of intergovernmental international organizations fall broadly into this category.

At the second level, events are random. Although detailed historical data can be collected, real decisive variables of the event are hard to find. Wars and terrorist attacks belong to this category.

At the third level, the data is highly random and variable. Researchers are unable to predict these even if they collect as much data as possible because the sample distribution does not provide traceable clues.

In this light, the more we depend on knowledge about the past, the more flaws the predictions will have. Predictions of global economic trends fall within the third level. Lu said the possibility of successfully predicting trends in international relations depends on which level the event is categorized into. Small events at the first and second levels are predictable, while events at the third level are often unpredictable.

 

Theoretical prediction required
When assessing the potential accuracy of international relations predictions, understanding of the necessity of forecasting needs to be further clarified. For a long time, few theories of international relations gave enough attention to predictions, and specific predictions backed with academic support have been rare.

An important reason is that many scholars believe that the core function of international relations theory is to explain rather than predict. Lu said theory is a simplified tool in terms of information. Theory that has strong explanatory power must be as simple as possible. A successful prediction must incorporate as many variables as possible in the theoretical model, which runs counter to the principle of theoretical simplicity.

Although Lu does not believe that predictions are a necessary function of international relations theory, he agreed that international relations theories are a necessity for predictions. He said that in the establishment of a forecast model, it is necessary to learn from the previous more commonly accepted theoretical research results in order to determine which variables should be included and which should be excluded.

Chen’s attitude towards theoretical prediction is more positive. From the perspective of theoretical construction and subject development, forecasts are a step forward in perfecting the theory and discipline system,” said Chen. He added that it is not proper to hastily deny the applicability of a theory because it failed to predict the occurrence of an empirical phenomenon.

Lin Minwang, deputy-director of the South Asian Research Center at Fudan University, said forecasts are an important measure to examine the quality of policy research. One common worrisome phenomenon in policy research is the pursuit of the “perfect.” Scholars either attempt to conclude all possibilities of the development of events, offering no falsifiable prejudgments, or trace back, analyze and list reasons accounting for the occurrence of events. It is difficult to really improve the level of policy research in this manner.

 

MAO LI is a reporter at the Chinese Social Sciences Today.