Socio-economic Development and Anti-government Protests in Light of a New Quantitative Analysis of Global Databases

Previous studies have revealed a somewhat paradoxical strong positive correlation between per capita GDP and the intensity of anti-government demonstrations observed for the vast majority of countries (indeed, it turns out that the better people live, the more likely they are to join anti-government protests). The goal of this article is to identify possible causes of this unusual correlation. Our tests show that the processes of democratization and urbanization, as well as the expansion of formal education, are likely to be the main factors determining the positive relationship between per capita GDP and the intensity of antigovernment demonstrations, as urbanization, democratization, and expansion of education lead to an increase in the intensity of protests. Moreover, when controlling for these factors, the relationship between per capita GDP and anti-government protests becomes negative. Thus, high per capita GDP turns out to be a direct (proximate) significant negative factor affecting the intensity of anti-government demonstrations, but at thesame time it is an ultimate, even more significant positive factor in the intensity of protests. The growth of per capita GDP is quite naturally accompanied by an increase in the level of urbanization, democratization and education, which more than compensates for the direct inhibiting effect on the protests on the part of the growing per capita GDP (at least for low- and middle-income countries). In addition, the negative binomial regression model that we propose can explain not only the strong positive correlation between per capita GDP and the intensity of protests, which can be traced for a range of GDP per capita values of less than $20,000, but also the weaker negative correlation recorded for the range exceeding $20,000. The fact is that in rich countries urbanization, democratization and education indicators reach saturation levels and the vast majority of high-income countries have more or less similar levels for all three indicators. As a result, for a zone of per capita GDP values of more than $20,000, we are essentially dealing with automatic control of the correlation between GDP per capita and the intensity of protests for factors of democratization, education and urbanization, and, as our model predicts, the final effect of GDP per capita on the intensity of protests for high-income countries becomes negative, not positive.