I am grateful to Prof. Okazaki for engaging with my book and welcome the opportunity to briefly respond to his review.
My book explores a particular constitutional design of democracy called ‘semi-parliamentary government’. It analyzes cases that fulfill minimal conditions of a semi-parliamentary system (the Australian Commonwealth, the bicameral Australian states and Japan) and discusses the potential of various new semi-parliamentary designs. Prof. Okazaki’s review focuses on the latter and on one specific design in particular. This focus may lead to misunderstandings about the book’s main arguments.
First, he states that only if there is a randomly selected second chamber, can there be a separation of powers. This is incorrect. The bicameral separation of powers and the method of selecting the second chamber are two separable features of any constitutional design. What is true is that one particular design option (pp. 135-8), which does not require two separate chambers, might in some sense reduce the degree of the separation of powers. It does so on purpose. According to this proposal, parties and voters could reduce the degree of powers-separation to the extent that they are able to coordinate on two competing pre-electoral coalitions before the election. Other designs of semi-parliamentary government, including all existing forms of semi-parliamentary bicameralism, do not share this feature.
Second, Prof. Okazaki makes two related claims: (a) that a randomly selected second chamber should be able to dismiss the government in a no-confidence vote and (b) that this requires a revision of the definition of semi-parliamentarism. The first claim is questionable, given what we know about how parliaments work. A randomly selected second chamber would probably be ill-suited to select a government and to provide continuing confidence to it. The second claim misunderstands the concept of semi-parliamentary government as well as the conception of the separation of powers used in the book. The second chamber’s lack of a no-confidence vote is what constitutes the separation of powers in the first place. The very point of the concept of semi-parliamentary government is to define a system in which the second chamber is at least as democratically legitimate as the first chamber (and has robust legislative veto power) but does not have a no-confidence vote. The kind of system that Prof. Okazaki envisions would be more akin to the bicameral parliamentary system we find in Italy.
Third, given his own preference for a randomly selected second chamber, Prof. Okazaki correctly notes that the competing goals in designing the elected chamber (identifiability/accountability versus proportionality) would have to be reconciled differently – to the extent that this is possible. To do this, he plausibly suggests proportional representation with a majority bonus. However, he also claims that the semi-parliamentary system “is weak in reconciling the ideal of identifiability and accountability and the ideal of proportionality.” This claim is incorrect and neglects the empirical evidence presented in chapters 5 and 6. In fact, there is an obvious way in which semi-parliamentary systems are better at reconciling the respective goals: they do not require any disproportionality in the second chamber to achieve the identifiability of competing governments in the first chamber. However, they pay the price of the separation of powers: the government is likely to lack a majority in proportional second chamber. This also reduces accountability. Parliamentarism with a majority bonus has a clear advantage in this respect, but it pays the price of significant disproportionality. Moreover, it pursues the goals of identifiability and accountability by means of an arbitrary seat bonus rather than by relying on the actual votes of citizens. This may be an important disadvantage in terms of democratic legitimacy.
My book’s main argument is that semi-parliamentary government is superior to presidential government. It does not claim that semi-parliamentarism or any particular version of it is the best system overall. Careful decisions about constitutional design require further theoretical discussion and empirical research. I thank Prof. Okazaki for engaging in this important debate.
GANGHOF, Steffen (University of Potsdam, Germany)