Application: the constitutional reform in Japan
When (i) voters have a convex utility function and (ii) the distribution of their most preferred policies is polarized, candidates choose policy divergence, political ambiguity, or any combination of the two. As discussed in Chapter 1, voters may have convex utility functions on non-economic issues. Although policy divergence is observed for some non-economic issues with polarized voters, such as the debate around same-sex marriage in the United States (Kamada and Ko- jima, 2014), candidates also prefer choosing an ambiguous position. Another example of political ambiguity is the constitutional reform in Japan.
The Constitution of Japan was enacted in 1947 as the new constitution for post-war Japan. In 1947, Japan was occupied by the Allies, mainly the United States. Thus, the Constitution was written by non-Japanese, although the opinions of many Japanese were considered. Therefore, constitutional reform has been a topic of frequent discussion since Japan gained independence. Article 9 is the most controversial, as it prohibits Japan from maintaining a military, air force, navy, or settling any international dispute using force. Nevertheless, Japan has had a defense force that has held military power since 1954. Public opinion on constitutional reform is divided. According to a 2017 poll conducted by NHK (the public broadcaster in Japan), 43% of the responses were in favor of the reform, while 34% of the responses were against the reform. These findings have been consistent over time.7 This issue is not related to the economy; therefore, it may be a convex issue, and the distribution of voters’ opinions is polarized. Thus, the conditions for political ambiguity are satisfied.
Since 1955, the Liberal Democratic Party of Japan (LDP) has run the government, except during the periods, 1993-1994 and 2008-2012. In the early period of the LDP administration (e.g., the Hatoyama administration, 1954-1956), many LDP members claimed that the Constitution should be written by the Japanese people. However, since the 1960s, LDP administrations have avoided discussing (and almost given up on) this issue because public opinion was so divided and an intra-LDP faction hesitated to implement reforms (Machidori, 2016, p. 4). Consequently, the Japanese Constitution has not yet been revised.
Recently, Prime Minister Shinzo Abe explicitly promised to reform the Constitution in the 2017 general election. The 2017 LDP manifesto (a booklet containing campaign promises) devoted two pages (out of a total of 38 pages) to this promise. In contrast, in the 2012 and 2014 elections, in which Abe was also the party leader, the LDP manifestos devoted only one-sixth to half a page (out of 26 pages) to this issue. Moreover, even in the 2017 manifesto, details on the reform remained vague since most parts only show some images of the constitutional reform. Voters believed that the LDP was more likely to revise the Constitution than its opponents(в? > dt); however, it was ambiguous[в^ Ф l).
In an election, parties and candidates usually announce and promote their economic policies. Indeed, most LDP manifestos in 2012, 2014, and 2017 laid significant emphasis (large weights) on explaining economic policies (popularly known as Abenomics in Japan). On the other hand, candidates prefer maintaining a degree of ambiguity on social and national-security issues. Possibly, they prefer specifying an economic policy because voters have concave utility functions for such economic policies. However, they prefer ambiguity for non-economic and polarized issues, where voters may have convex utility functions.
Summary
Prior studies usually interpret political ambiguity as a lottery. This chapter also supposes that voters can choose between lotteries, rather than a single policy. Further, it identifies the conditions under which political ambiguity occurs in equilibrium, given the convex utility functions of voters. In the deterministic model, if voters have concave or linear utility functions, the median policy is still the Condorcet winner. However, if voters have convex utility functions, the existence of the Condorcet winning lottery is not ensured because the space of campaign promises has multiple dimensions. On the other hand, in the probabilistic voting model, candidates choose an ambiguous promise in equilibrium when (i) voters have convex utility functions and (ii) the distribution of voters’ most preferred policies is polarized. Therefore, to have political ambiguity as an equilibrium phenomenon with convex utility functions of voters, voters need to be polarized, and candidates must be uncertain about voters’ preferences.
