Collateral Damage Autocracy? - Tobias Lechner Development Economics and Policy

Скачать книгу

nor for his coalition, and may rely on repression.

      In contrast, a high level of development (variable 6) helps the population to decrease the information asymmetry and to overcome the collective action problem, which may lead to a revolutionary attempt. When the population, however, agrees on the leader’s offer (more or less public goods), it can lead to stability and an equilibrium. This happens especially when the level of legitimacy of the leader is very high (variable 4), and when the leader succeeds in stimulating nationalist feelings. In such cases, sanctions have a so-called rally around the flag effect. Another threat to the leadership comes from his winning coalition: If the coalition does not accept the goods it got (or didn’t get), it can stage a coup d’état (vertical threat). A coup is more likely if the sanctions target specific groups within the system as most democratic sanctions do (variable 2).

      After facing the mentioned vertical and horizontal threats, the leader still has two tools: The first one is increased (or also decreased) repression of the population or a purge of his coalition. If repression is too expensive or not effective, he can rely on his second tool: (de-)centralization of his political power. He can ←21 | 22→centralize political power, or he can establish political institutions and co-opt oppositional forces or enfranchise larger parts of the population. His actions rely on the mix of variables.

      Methodology

      The unit of analysis is the country-year of countries that were targets of sanctions (n =1440). Due to data issues, only sanctions with the UN, the EU or the U.S. as sender are considered. Only sanctions that were imposed between 1990 and 2015 and were in place for more than one year are included (60 target countries). In addition to the known sanctions datasets HSEO, GIGA, TIES, and TSC, primary sources such as the UNSC resolutions and other legal texts are used. The dependent variable is the change of the level of democracy. In order to compensate for the flaws of both scores, a combined score of Freedom in the World and Polity IV is used.

      Sanctions variables: Variable 1, economic costs of sanctions to the target state, proved to be most difficult to operationalize. If the measurement is only based on the design, the results will be misleading. The most comprehensive sanctions will have no impact, if the pre-sanctions trade level is close to zero (as North Korea post-1993 shows). If the measurement is based on absolute numbers such as the decline of GDP, it is impossible to distinguish the sanctions effects from other effects such as economic stagnation (cf. post-2014 sanctions against Russia). Therefore, most scholars simply estimate the economic damage done by sanctions. This book uses a combination of ex-ante threat (using GIGA) and ex-post damage (using TIES or HSEO). Variable 2, democratic goal of sanctions, is based on the GIGA sanctions dataset. Each sanction episode that explicitly focuses on democracy is coded as 1; otherwise, it is coded as 0 (dummy variable).

      Regime variables: Variable 3, personalist regime, is a dummy variable based on Geddes et al.’s Autocratic Regimes dataset (GWF). In addition, liberal democracies are coded as non-personalist. Variable 4, regime legitimacy, is taken from the State Fragility Index and Matrix (SFIM).

      Economic variables: Variable 5, economic vulnerability, measures the export dependency of the target state. The World Bank provides comprehensive data on the exports of goods and services in percent of the GDP. Variable 6, development, uses data from the Human Development Index (HDI), with the sub-indicators life expectancy at birth, expected years of schooling, mean years of schooling, and GNI per capita.

      Control variables: One variable for natural resources controls for the resource curse hypothesis, using World Bank data. Another variable controls for civil war, using the dataset Major Episodes of Political Violence (MEPV).

      ←22 | 23→

      The econometric model for the variables is:

      democracy=α1costs+α2goal+α3personalism+α4legitimacy+α5vulnerability +α6development+β1resources+β2war+μit

      Data type: panel data; regression analysis with two-step system GMM (SYSGMM); statistical software: Stata 14.

      Findings

      There is no correlation between any of the sanctions variables and an increased level of autocracy. Instead, some of the variables show a correlation between sanctions and an increased level of democracy, some are not significant at all.

Variable Hypothesis Result
IV 1: Economic costs The higher the economic costs of the sanctions to the target state are, the more likely is a negative impact on the level of democracy in the target state. Not confirmed, opposite correlation: The higher the economic costs of the sanctions to the target state are, the higher is the level of democracy.
IV 2: Democratic goal The less sanctions focus on democracy, the more likely is a negative impact on the level of democracy. Confirmed. Furthermore, sanctions with a democratic goal correlate with a higher level of democracy.
IV 3: Personalist regime The more personalized rule in the target country is, the more likely is a negative impact on the level of democracy. Not confirmed, no correlation.
IV 4: Regime legitimacy The higher the level of legitimacy of the targeted regime is, the more likely is a negative impact on the level of democracy. Not confirmed, no correlation.
IV 5: Economic vulnerability The higher the level of economic vulnerability of the target economy is, the more likely is a negative impact on the level of democracy. Confirmed, but weak and negligible impact.
IV 6: Development The less developed the target country is, the more likely is a negative impact on the level of democracy. Confirmed. Furthermore, the more developed the target country is, the more likely is a positive impact on the level of democracy.

      ←23 | 24→

      Variable 1: Previous literature argued that sanctions have a negative impact on the level of democracy, especially when they are comprehensive. Economic theory suggests an even stronger political impact with higher economic costs. The present analysis, however, did not confirm that, and found a correlation with a higher degree of democracy instead. Economic sanctions do not cause autocratization, both variables do not even correlate. This does not necessarily mean that sanctions are a cause of democratization, but it shows that anecdotal evidence of cases of sanctions with a negative impact on the level of democracy cannot be generalized. The finding is in line with the most basic sanctions mechanism that was described as naïve: Sanctions with high economic costs disrupt the targeted political system and have transitional effects, perhaps in the context of a global, regional or temporal trend. External punishment legitimizes oppositional forces and may help the citizens to overcome their collective action problem.

      Variable 3: A personalist regime does not correlate with a higher or lower level of democracy. Many autocratic regimes are perhaps not consolidated enough to distinguish different mechanisms.

      Variable 4: A high degree of legitimacy of the regime in the target state does not correlate with a lower or higher level of democracy.

      Variable 5: In the main model, a high level of vulnerability (economic interdependency or interconnectedness) correlates with a lower level of democracy, ←24 | 25→but the coefficient is negligible and varies substantially in different

Скачать книгу