Staatszerfall in Kolumbien und Angola? Über die Erklärung und Prognostizierbarkeit von Staatszerfall anhand des Global Model

DSpace Repositorium (Manakin basiert)


Dateien:

Zitierfähiger Link (URI): http://nbn-resolving.de/urn:nbn:de:bsz:21-opus-31797
http://hdl.handle.net/10900/47572
Dokumentart: Abschlussarbeit (Master)
Erscheinungsdatum: 2007
Sprache: Deutsch
Fakultät: 6 Wirtschafts- und Sozialwissenschaftliche Fakultät
Fachbereich: Politikwissenschaft
DDC-Klassifikation: 320 - Politik
Schlagworte: Staat , Kolumbien , Angola , Frühwarnsystem
Freie Schlagwörter: Staatszerfall , Frühwarnsysteme , Angola, Kolumbien , Global Model
State Failure , early warning systems , Angola , Colombia , Global Model
Lizenz: http://tobias-lib.uni-tuebingen.de/doku/lic_mit_pod.php?la=de http://tobias-lib.uni-tuebingen.de/doku/lic_mit_pod.php?la=en
Gedruckte Kopie bestellen: Print-on-Demand
Zur Langanzeige

Inhaltszusammenfassung:

Das bis heute umfangreichste Projekt, Staatszerfall wissenschaftlich aufzuarbeiten, stellt das im Frühjahr 1994 durch den damalige Vizepräsidenten der Vereinigten Staaten von Amerika, Al Gore, zusammen mit der Central Intelligence Agency (CIA) initiierte State Failure Project (SFP) dar. Kernstück der Forschungsbemühungen war die Entwicklung eines quanitativen Frühwarnsystems für Staatszerfall, des sogenannten Global Forecasting Model of Political Instability, kurz Global Model (GM). Basierend auf der im Mai 2006 erstmals erfolgten Veröffentlichung eines Großteils der Daten und Ergebnisse des Projekts, unternimmt die vorliegende Arbeit eine kritische Analyse des GM. Zunächst wird, nach einer Einführung in die State Failure Debatte sowie die Frühwarnforschung, das GM auf methodologische, theoretische und analytische Schwächen im Bereich der Makroebene des Modells untersucht. Dies beinhaltet eine detaillierte Exploration der vorgenommenen Definitionen, der Datenauswahl sowie Modellbildung mit Fokus auf der abhängigen Variable. In einem zweiten Schritt werden die durch das GM selektionierten fünf unabhängigen Schlüsselvariablen auf ihre Konsistenz mit anderen Forschungsergebnissen, ihre statistische Signifikanz und ihre Verortung im kausalen Gefüge geprüft. Im Anschluss daran erfolgt eine Evaluierung der gezogenen kausalen Schlüsse sowie der generellen Vorhersage- und Erklärungsperformanz des Modells. Um das Vorhersage- und Erklärungspotenzial des Modells im konkreten Einzelfall zu testen, werden anschließend die fünf unabhängigen Schlüsselvariablen auf zwei Fallbeispiele, Kolumbien und Angola, angewandt. Die vorliegende Arbeit deckt in Folge der Analyse schwerwiegende methodologische, theoretische und systemische Schwächen auf. Als Quintessenz der Makro- sowie Fallanalyse lässt sich das Urteil fällen, dass das SFP nicht in der Lage war, ein überzeugendes Frühwarnmodell für Staatszerfall zu entwickeln. Im Bemühen das GM auf eine rein empirische Basis zu stellen, wurde die Theorie als Grundlage der Modellbildung vernachlässigt. Dies hat, wie in der Arbeit nachgewiesen werden kann, direkte Konsequenzen für die Prognose- und Erklärungsperformanz des GM in Form von unzureichenden Definitionen, ignorierter Pfadabhängigkeit sowie unklaren kausalen Verortungen und Schlüssen. Darüber hinaus zeigt sich der Charakter der unabhängigen Variablen als inadäquat für zeitnahe Frühwarnung. Vier von fünf Variablen sind Strukturvariablen wie z.B. Säuglingssterblichkeit und Regimetyp, welcher per se nicht zeitsensitiv genug sind, um im Model vorgegebenen Vorhersagehorizont von zwei Jahren Staatszerfall vorhersagen zu können. Die Arbeit endet mit einem Ausblick auf mögliche pragmatische Verbesserungsvorschläge für das GM sowie allgemeinen Lektionen für die aktuelle Staatszerfalls- und Frühwarnforschungsdebatte.

Abstract:

Based on the first time disclosure of data and results of the project in May 2006, the present paper undertakes a critical review and analysis of the Global Model. In the first instance a general introduction to the state failure debate and the early warning research is given. Following that the Global Model is examined for methodological, theoretical and analytical deficits on the macro level of the model. This implies a detailed exploration of the definitions used, the data selection process and the model building, focusing on the outcome variable. In a second step, the preselected five key independent variables are examined for their consistency with current and established research results, their localisation in the causal chain and the resulting forecast and explanatory power. In order to test the forecast and explanatory power of the model in concrete and particular cases, the five independent key variables are applied to two case studies, Columbia and Angola. In the course of the analysis it is shown, that the model suffers from grave methodical, theoretical and systemic shortcomings. As quintessence of the macro and case study analysis this paper comes to the conclusion that the SFP failed to develop a convincing and coherent early warning forecasting model for state failure. Due to the attempt to develop the model on a purely empirical basis, theory as the fundamental tool for model building has been ignored. This has resulted in direct, limitating consequences for the explanatory and forecasting power of the model; for example insufficient definitions, ignored path dependency as well as unclear causal localisations and inferences. Furthermore the nature of the five key independent variables proves to be inadequate for early warning. Four out of five variables are structure variables (for example regime type or child mortality) which do not provide the necessary time sensitivity for the forecasting of state failure within the given time frame. The paper concludes with an outlook on pragmatic proposals for how to readjust the Global Model in order to improve the forecasting and explanatory performance as well as general conclusions for the scientific state failure and early warning debate.

Das Dokument erscheint in: