Page 32

The Hinckley Journal of Politics

random selection. This method allows for some direct testing of the aforementioned hypotheses. While the results of an analysis like this are useful, they are not confirming. This analysis can help pinpoint which factors have a statistically observable effect. The regression analysis will be analyzing Hypotheses 1,3,4, and 5. This regression includes all but one of the world’s 35 members of the OECD, and will take into account migration from all active conflict zones and terrorism activities creating refugee-like populations. The choice in limiting this specific data analysis to a relatively small selection of countries is due to several reasons. The primary reason being that migrants tend to focus on moving toward wealthier countries. Second, as Turkey is an OECD state, its inclusion in this part of the data analysis will complement and help to bridge the more qualitative aspects of this research paper to a detailed quantitative analysis. This data analysis takes into account refugee-like situations. This includes officially considered refugees and those believed to be in similar circumstances but who do not have the same legal classification from all active conflict zones in the world as recorded by the UNHCR. While the Syrian refugee crisis may consume the majority of media and public attention, the majority of migrants, refugees, and asylum seekers are not Syrian. It is also important to mention that the measure of democracy in this statistical test is taken from the Polity 4 data set (Marshall, 2015). Many similar studies use this dataset to incorporate into panel data as well. Log gross domestic product per capita roughly helps capture state capacities to operationalize that portion of my theory. I use the log version of the data to scale the coefficients to a manageable figure. Clearly, this is not the ideal measure. A number of variables can capture the nuances of economic development, and perhaps in some ways, log GDP/per capita does not capture all aspects. Data regarding involvement in armed conflict is taken from the PRIO dataset and is used as a dummy variable (Gleditsch, Wallensteen, Sollenberg, & Strand, 2002). In addition, the variable called “terrorist history” will control for clustering of terrorist incidents as well as help test for specific targeting of certain states. The following table outlines more information about the data used in this panel setup.

Dependent variable Independent variables

Whether a state is involved in at least one armed conflict in that year

The Peace Research Institute Oslo (PRIO)

Democracy

Yearly measure of democratic quality

Polity 4 Project

Terrorist history

Dummy variable of whether a terrorist incident had occurred in the past year

START Global Terrorism Database

VARI-

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

ABLES

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

Model 7

Model 8

Model 9

Refu-

0.000113

0.000110

0.000111

0.000110

0.000109

9.86e-

9.62e-

9.95e-

0.000103

***

***

***

***

***

05***

05***

05***

***

(1.02e-05)

(1.02e-

(1.04e-

(1.04e-

(1.04e-

(1.01e-

(1.27e-

(1.04e-

(1.03e-

05)

05)

05)

05)

05)

05)

05)

05)

2.354**

2.350**

2.295

2.293**

2.516**

2.618**

gees

Year

(1.083) Coun-

(1.084)

(1.092)

(1.094)

(1.090)

(1.121)

-2.37e-08

-4.03e-08

-4.29e-08

-1.70e-

-1.65e-

07***

07***

(6.33e-

(6.15e-

(6.08e-

(5.40e-

(5.92e-

try Population

08)

08)

08)

08)

08)

7.092

7.282

7.864*

8.218*

(5.079)

(5.069)

(4.679)

(4.882)

GDP/

13.77

-9.705

-7.883

-5.959

capita

(12.53)

(9.923)

(10.93)

(12.78)

Armed

55.56***

53.52***

41.18***

Con-

(11.53)

(13.09)

(11.52)

Terrorist History Log-

flict Polity 4 5.868* -1.207

Unit

Source

Terrorism incidents

Per country per year

START Global Terrorism Database

Number of refugees recorded in that country per year

United Nations High Commissioner for Refugees

All persons of concern

Refugees, like situations, asylum seekers, returned individuals

United Nations High Commissioner for Refugees

Log GDP/capita

Current USD logged GDP/ capita per year per country

World Bank Data

(3.311)

(3.179) 4.617

-4,732**

-4,724**

-4,616**

-4,549**

stant

Variable Name

Refugees

Armed Conflict

Table 5: Terrorism incidents and refugee/migrants from conflict zones 2010-2015

Con-

Table 4: Regression Analysis Variable Table

2017

Obser-

-

-

5,015**

5,218**

88.42

1.990

(3.673)

(2,179)

(2,182)

(2,198)

(2,203)

(2,194)

(2,256)

(56.15)

(3.375)

204

204

204

204

204

204

198

198

204

35

35

35

35

35

35

34

34

25

vations

Number of groups

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

Table 5 contains some interesting results. First, the refugee variable is significant and positive for all models. However, the variable is not practically significant for all cases. For example, model 7 predicts that roughly 11,000 refugees would need to increase in a state to expect one more terrorist attack. While there are several countries that admit far more refugees than even this number each year, the majority of host states admit far less. Therefore, I argue that this correlation is not practically significant due to the lack of predictability for most states. Still, the results suggest some hint of a correlation for states that admit more refugees than average. However, this does not hold under closer

19

Profile for mlusty

Hinckley Journal 2017  

The Hinckley Journal of Politics is the only undergraduate-run journal of politics in the nation and strives to publish scholarly papers of...

Hinckley Journal 2017  

The Hinckley Journal of Politics is the only undergraduate-run journal of politics in the nation and strives to publish scholarly papers of...

Advertisement