Diversionary War Theory, specifically between Turkey and Greece, is true – studies and calculations prove.
Marius Mehrl and Ioannis Choulis, 3/17/21, Marius Mehrl Postdoctoral Researcher at the Geschwister Scholl Institute of Political Science, Ioannis Choulis Assistant Lecturer Department of Government, University of Munich, “Diversionary Politics and Territorial Disputes: Evidence from Turkish Airspace Incursions,” Foreign Policy Analysis, Volume 17, Issue 3, July 2021, orab007, https://doi.org/10.1093/fpa/orab007,
We present our empirical results in table 2. Model 1 includes only Approvalm−1, the independent variable of main theoretical interest, and term-fixed effects in addition to the AR(1) and May terms, while model 2 adds the dummy variables indicating pre-election months and the gas dispute. Model 3 then adds the variable measuring Turkish citizens’ perceptions of the economy, while model 4 instead includes GDP per capita as well as Greek and Turkish defense spending. Econ. perceptionsm−1 is only available from 2015 onward, leading to reduced temporal coverage of the time series and hence our sample size. We thus omit this variable from the most complete specification, model 4. The results in models 1–4 are in line with our hypothesis as Approvalm−1 exhibits a coefficient that is consistently negative and statistically distinguishable from zero at the 99 percent level. Since the dependent variable in these models is the transformed number of airspace violations while the unit of Approvalm−1 is percent, we can interpret its coefficient as the percent change in airspace incursions for a 1 percent increase in approval. Based on model 4, a 1 percent increase in public approval for Recep Tayyip Erdogan leads to a 0.0319 percent decrease in airspace ˘ violations. In contrast, the control variables do not assert a statically significant effect on airspace incursions. To examine the substantive effect of approval ratings on airspace violations in more detail, we present a plot of the predicted number of Turkish incursions into Greek-claimed airspace over the range of observed approval values in figure 2. The figure underlines that approval rate has a substantive, if moderate, effect on airspace violations as it shows that model 4 predicts 141 such incursions if Erdogan’s approval was 37.5 percent in the previous month, whereas ˘ this number decreases to almost 100 if 50 percent of respondents approved of his leadership. Next, we test the robustness of our results to several alternative specifications. Most importantly, we follow earlier studies (e.g., DeRouen and Peake 2002; Pickering and Kisangani 2010) by using VAR to model the effect of approval ratings on airspace violations and vice versa as an endogenous relationship. This allows us to examine whether approval drives airspace violations and/or the other way around. For identification purposes, it is necessary to restrict the contemporaneous effect of one of the endogenous variables to zero (Enders 2015). Approval data take time to collect with the Metropoll survey data only being collected toward the end Downloaded from https://academic.oup.com/fpa/article/17/3/orab007/6204770 by University of Michigan user on 22 July 2022 8 Diversionary Politics and Territorial Disputes Table 2. Approval ratings and airspace violations Dependent variable: ln Airspace violations. The Effect of Approval Ratings on Airspace Incursions Note: Predicted Value of ln Airspace Violations across different values of Approval Rate, based on model four. Italicized numbers indicate corresponding non-logged number of violations. Dotted lines indicate 90%-Confidence Intervals. The Effect of Approval Ratings on Airspace Incursions - VAR results Note: Impulse Reaction Functions. Grey areas indicate 90%-Confidence Intervals. of the month, meaning that they only become available after some time has passed. In contrast, reporting on airspace violations is more imminent as they are mentioned in the news and on social media. We thus allow airspace violations to have a contemporaneous effect on approval ratings, while only past approval values can affect airspace violations.7 Following standard practice, we present impulse response function plots in figure 3, showing how one endogenous variable is affected by a shock in the other endogenous variable, as well as results from Granger causality tests to interpret the VAR results instead of relying on regression coefficients. In line with our expectations, figure 3 indicates that positive shocks to approval ratings result in less airspace violations. In contrast, positive shocks to the number of airspace incursions are found to have no statistically significant effect on leader approval. Accordingly, Granger causality tests suggest that approval ratings drive airspace violations but not the other way around. These results provide additional support to our expectation that low approval ratings lead to diversionary action. However, they also raise the question why this is the case if such diversionary acts do not result in higher approval. One possible answer is that these diversionary actions do not target a wider domestic audience but instead key players within the governmental coalition (Morgan and Bickers 1992; Nicholls, Huth, and Appel 2010). Turkish territorial incursions may thus be designed to garner support from ultranationalist elements within the AKP and MHP by showing that the government is not afraid to confront Greece over the contested territory, even if such a policy may not find support, and hence boost approval, among more centrist or opposition voters. We present further specifications examining the robustness of our results in the online appendix, but the result that lower approval ratings increase airspace incursions persists. Conclusion Diversionary foreign policies have been the topic of a large literature that, however, reports mixed findings on whether domestic problems result in interstate conflict. 7 The VAR models include the control variables from model 4 and, based on information criteria, a lag order of 1. This paper adds to this debate by arguing that while leaders may be hesitant to start an outright war to boost their approval, they will pursue smaller scale aggressive actions within existing disputes to gain domestic support. Results from tests using monthly data on leader approval and airspace violations in the territorial dispute between Greece and Turkey offer support for this argument: low approval ratings for Turkey’s leader result in more Turkish incursions into contested airspace. This result most directly contributes to the literature on diversionary conflict. First, we introduce Turkey as a new case for quantitatively testing the theory, whereas most applications focus on the United States. Second, we add to a nascent literature studying diversionary acts below the crisis or militarized dispute level. Third and most importantly, we show that when facing domestic disapproval, leaders resort to material acts of aggressive foreign policy that are less costly than a new conflict but also more than cheap talk to increase support. This is an important result as such aggressive acts, the intensification of an existing territorial dispute in the case of Turkey, can lead to actual armed conflict further down the line. Our research thus also provides insights into the escalatory processes leading up to armed confrontations between neighbors such as those on the Indian–Chinese border in early summer 2020. This study suggests several avenues of further research. First, future research should investigate whether territorial incursions or provocations are used in a similar, diversionary manner in cases other than Turkey. Second, they should examine whether domestic problems generally give rise to aggressive foreign policy behaviors below armed conflict or whether this requires the previous existence of a dispute whose salience leaders can tap into. Third, our research produced the surprising finding that airspace incursions do not actually succeed in boosting support from the Turkish public. As such, future research should further examine to what extent such diversionary policies target the general public or more specific, strategically important audiences such as coalition partners