[latexpage]

Introduction

Terrorism has always been a difficult subject to define and explore; multiple viewpoints around the world about the subject tend to oppose each other strongly due to the political and emotional nature of terrorism, and various variables attributed to terrorist attacks, such as the motive, the target, and the type of attack, have allowed for a wide array of data on terrorism that can be troublesome to consolidate. The Global Terrorism Database GTD  recorded at the University of Maryland, however, offers a way to organize terrorism based on such variables. With over 100,000 attacks recorded over a timespan from 1970 to 2011 and around 120 variables to describe them, the GTD provides an excellent resource to better define terrorism through exploration. One variable that could have a significant effect on terrorism is the quality of the government whose country experienced the attack. We attempt to define a relationship between the democracy of a country and its experience with terrorism.

In order to explore the effect on democracy, We utilizes the QoG Standard Dataset from the University of Gothenberg. The variable extracted from the dataset, Institutionalized Democracy, defines democracy as three elements: the presence of institutions and procedures through with citizens can freely choose policies and leaders, the existence of constraints on the leaders of the government, and the guarantee of civil and political liberties. The variable is distributed on an eleven-point scale from 0-10, with 10 being the most democratic and 0 being the least. Democracy scores were recorded for 179 countries over a timespan from 1946-2011. In order to better define the relationship between democracy and terrorism, it is imperative that as many other variables as possible are controlled. Therefore, we focuses on a specific region of the world, Central America , which contains countries with similar cultures and beliefs.

Central America Stands out Globally

The plot above shows the number of terrorist attacks in the world from 1970 to 2011. The thickness of the line represents the total number of people killed and wounded in the attacks, and the color filling indicates the number of perpetrators in the attacks. The black horizontal line explains the average number of attacks and the vertical line indicates 911 event. Other red dotted lines : 1st the year number of attacks reached highest in Center America; 2nd the year when the attacks attained maximum globally; 3rd the latest condition in GTD.

The above plots provide an interesting comparison between Central America and the world in general. The first thing to note is the lack of a rise in the number of attacks in Central America from 1992 onwards, while the world witnessed an overall increase in attacks after 2001, likely a result of conflicts in the Middle East. One possible reason for the low number of attacks in Central America could be the increased political stability in the region, as nearly every country started to have free and fair elections as time passed. The number of perpetrators and the number of people killed and wounded in Central America also decreased drastically after 1992. This observation occurred because, by the year 1992, nearly all civil wars in the region had ended, and stable democracies began to form.

Relationship between Democracy and Terrorism

In addition to a general comparison between the world and Central America, we did an analysis by combining the value of democracy and attack numbers to better understand the hidden reasons behind this unique pattern of Central America and check the potential correlation between them.

The figure above plots the average democracy score of all the Central American countries over time from 1970 to 2010. The color of the points is a scale representing the total number of perpetrators for each year. The size of the points represents the total number of attacks in each year.

The above figure shows a general answer to the main investigation of our project. According to the figure, as the overall democracy score increased in Central America over time, the number of attacks and perpetrators noticeably decreased. However, one limitation in this graph is that the full range of democracy scores (vary with the year) for Central America is not shown in this graph since it only plots the average democracy score. To further explore the negative correlation between democracy and the number of attacks, a map of Central America showing the change in the number of attacks is required.

The figure is a map of Central America’s terrorist attacks before and after 1992. The area of the colored squares indicate the relative percentage of each attacktype which could be easily interpreted by the corresponding Histogram on the right.

The above provides an overview of the locations of where most of the attacks in Central America. From the map, it can be observed that most of the attacks before 1992 occurred in El Salvador, Nicaragua, and Guatemala, while Nicaragua had the attacks that resulted in the most number of people injured or killed. After 1992, the number of attacks that caused a large number of deaths or injuries drastically decreased, and there was a significant overall decrease in attacks throughout the entire region. These findings further prove a benefit in having a stable democratic government.

The table shows the top 5 most attacked countries in Central America. The success rate is defined as $succeess\; rate=\dfrac{{\displaystyle \sum_{i}}success_{i}}{length\left(success\right)}$.

Conclusion

The findings presented offer valuable insight into whether or not a democracy provides a beneficial society for its citizens. Although the number of terrorist attacks are miniscule compared to the amount of other types of crime, the emotional impact of terrorism is too great to overlook. The data in this report showed a strong negative correlation between a country’s level of democracy and the terrorist attacks it experiences. Central America provided a great example of a region whose governments increased their stability significantly over time as they became more democratic. The number of terrorist attacks, the number of people killed or wounded, the number of perpetrators, and the success rate of the attacks all decreased throughout the region. These correlations strongly suggest that democracies in general lead to a much safer environment.

There are some limitations, however, with the data we presented. Although we selected a region with a similar culture, religion, and beliefs, not every variable was controlled in the analysis. Variables such as the state of the economy, motives other than political reasons, and foreign attacks are all unrelated to a country’s level of democracy and could all have a significant effect on terrorism. Additionally, the definition of democracy given by the variable we chose from the QoG dataset is not a fixed globally-accepted definition; there were multiple variables that measured democracy in that same dataset that could have provided different results from the ones obtained in this paper. A possible further avenue of exploration would be to look at the entire world at one point in time and examine the changes in terrorism based on the level of democracy. If enough variables are controlled in this exploration, then we could be one step closer to proving the relationship between democracy and terrorism.

About

This report is a joint project by Yinsen Miao, Yuan Chang (Department of Statistics, Rice University), and Rohit Reddy (Hanszen College, Rice University). Full pdf copy of the report and related R script can be acquired based on request.Email Me