Worldwide, the remarkable diffusion of transparency and access to information laws poses a monumental challenge to the state’s most enduringly undemocratic feature— excessive secrecy. Will recent laws lead to an effective surrender of secrecy? The incipient literature on transparency reform says little about the strength of current legislation or how strong laws emerge. This dissertation addresses these theoretical and empirical gaps. First, it articulates a theory on the political determinants of strong access to information laws. Second, employing an original evaluation, it scores the strength of twelve access to information laws advanced throughout Latin America between 2002 and 2010. Two extreme outcomes are examined in detail: a failed comprehensive reform in Argentina (1999-2005), which resulted in a limited presidential decree (2003), and the adoption of a seminal law in Mexico (2002). These cases are then compared with others across Latin America with special attention placed on Brazil, Chile, Guatemala, and Uruguay. I find considerable variance in the strength of the region’s laws: the average score is “moderately strong,” while the median and mode scores are “moderately weak.” Evidence shows that while civic coalitions and external pressure often help drive reform, they cannot explain observed variation in legal strength. Rather, I find that laws emerge more robust and earlier-on within the electoral cycle (within the first half of a president’s term of office), in countries where 1) presidents lack control over the legislature and 2) news media coverage of access to information laws is strong. By contrast, where news media coverage is weak and presidents possess strong negative agenda setting powers (partisan majorities or constitutional means of denying a vote), I find that laws tend to emerge later-on during the electoral cycle (within the last third), and are considerably weaker. I also find that press advocacy for access to information laws tended to be greater in countries where presidents were weaker and news media ownership concentration was low. The dissertation addresses key institutional preconditions for good governance and transparency reform. More specifically, it speaks to the determinants and power of the news media as an agent of democratic advancement (and stagnation), and the importance of weak leaders and partisan competition in promoting good governance reform. / text

半總統制的運作特色是行政權力分別由總統和總理掌握，當半總統制的總統權力較小，就會由總理主導行政權力，此時半總統制的運作模式趨近於內閣制。但過去較少有文獻比較總統權力較小的半總統制國家，與內閣制國家的實際政治運作，是否有顯著的差異。本文先參照Wu(2012)的分類，選擇準內閣制做為總統權力較小的半總統制次類型，再與內閣制國家做比較。 本文的研究結果，發現準內閣制與內閣制之間，在國會的有效政黨數目、無黨籍總理出現的機率、總統總理同黨的機率沒有顯著的差異。顯示準內閣制的總統如同Wu分類的定義，不會介入任命過程，另外，總統選舉的衣尾效應也無法使大黨的席次增加。在總理任期的比較方面，雖然兩種制度的總理任期沒有顯著的差異，不過準內閣制的總理遭遇提前選舉下台的風險，低於內閣制，且準內閣制的總理連任機率也低於內閣制。這些結果皆表示，準內閣制的總統雖然不會介入組閣過程，但有可能對於國會的選舉時程產生影響，使總理無法在最佳時機發動國會提前選舉，進一步降低總理的連任機率。 綜合本文的研究發現，再對照過去的相關理論。半總統制與內閣制在實際運作上，仍然有一些差異。近年來，學界認為總統權力不是半總統制的重要條件，實際上，即使在總統權力較小的國家，總統權力可能還是有重要的影響。 / The executive power in semi-presidentialism is shared by the president and prime minister where the prime minister will control executive power with a weak president, and the political practice will be like those found in parlimentarism. There are very few articles examining the difference between parlimentarism and semi-presidentialism with a weak president. This paper compares the political practice of semi-presidentialism with a weak president and parlimentarism in different countries. The case selection criteria of a weak president is based on Wu(2012). In his research, quasi-parlimentarism is a subtype of semi-presidentialism, that the president has weak power. The result of this research demonstrates that effective number of parties in parliamentary, probability of non-party prime minister, president and prime minister in the same party have no significant difference between two institutions. This means that in the quasi-parlimentarism the coattail effect of a presidential election does not ensure more seats for a large party, and the president has a weak power to appoint the cabinet. Although the duration of the term of a prime minister has no significant difference, the hazard of early election, and probability of prime minister renewal in quasi-parlimentarism is less than parlimentarism. This means that a president in quasi-parlimentarism has blocked dissolution of parliamentary, and the prime minister cannot call for an early election at his or her favorable time. In summary, the political practice of semi-presidentialism is different than parlimentarism. Presidential power is an important feature of semi-presidentialism, even in small presidential power, a semi-presidentialism country, and its president have great influence on politics.