The dissertation leads research about government accountability in a new direction. Previous models inquire "how much does the government account separately to the legislature and to the public?" The current study posits a much more comprehensive and dynamic question. It asks: "how much does the government account simultaneously not only to the legislature and the public, but also to the judiciary, international organizations and internal governmental bodies?"

The dissertation refines the conventional institutionalist explanation, which posits that "regime type" defines government accountability. The study suggests that the level of political development is also an important causal factor. I propose that three characteristics of unconsolidated democracies affect accountability patterns. The fragmented party landscape weakens the power of the legislature to sanction the incumbents. The weak civil society makes public opinion less critical of the office-holders. The disregard for institutional rules strengthens the sanctioning capacity of the president.

I devise a new measure of government accountability, which gauges how often the government is sanctioned for public allegations of malfeasance or incompetence. Media allegations are used as units of analysis because they overcome informational asymmetries between various principals. The proposed method improves existing measures, which gauge accountability with proxies, such as democratic values and civil rights. I compiled an original database of more than 1, 634 media charges and 5, 980 articles critical of the governments in Bulgaria and Germany and Russia.

My findings largely support the hypothesis that the "regime type" and the "level of democratic consolidation" determine the relative effectiveness of sanctioning bodies. I find that in Germany, a consolidated parliamentary democracy, the prosecutor imposes most sanctions on the government, followed by governmental investigations, international inquiries and parliamentary questions.

In Bulgaria, an unconsolidated parliamentary democracy, the government is mostly sanctioned by the Supreme Administrative Court, followed by lower courts, parliamentary committees, internal governmental investigations, the government's approval ratings and the audit chamber. In Russia, which is a presidential unconsolidated democracy, the most influential punisher is the president, followed by presidential approval ratings, the Constitutional Court and the prosecutor.

The dissertation is the first to apply empirically two normative criteria for assessing accountability: "who sanctions the government" and "how is the government sanctioned?" It concludes that executive accountability in Russia is largely flawed.