Academic journal article Journal of East Asian Studies

The Nationalization of Japanese Elections

Academic journal article Journal of East Asian Studies

The Nationalization of Japanese Elections

Article excerpt

The postwar electoral dominance of the Liberal Democratic Party (LDP) was founded on (1) strong incumbency advantage, which insulated its legislators from declining party popularity, and (2) the malapportionment of districts, which overvalued the electoral clout of the party's rural base. The LDP's demise in 2009 was due to the reversal of both factors, each of which was related to electoral reforms in the 1990s. First, I demonstrate that elections are becoming more "nationalized," due to the growing weight that voters attach to the attractiveness of party leaders. Past performance has become a poorer predictor of incumbent reelection, giving way to large partisan swings that are increasingly correlated across districts. Second, malapportionment was reduced by almost half in 1994, meaning that rural votes are now worth fewer seats. As a result, parties that can attract swing voters nationally are better positioned for victory than those with a narrow regional base.

KEYWORDS: Japan, elections, malapportionment, party, electoral reform, public opinion, partisan swing


JAPANESE POLITICS HAS TRADITIONALLY HINGED ON PERSONALISTIC LINKAGES between legislators and their constituents. Politicians attracted support through their individual accomplishments, ties to local organizations, and the promise of particularistic benefits. These linkages were especially strong in rural areas, where social networks were denser and local businesses relied heavily on government contracts and subsidies. This type of personalism produced high levels of incumbency advantage and low electoral turnover, especially for the long-ruling Liberal Democratic Party (LDP), which leveraged its access to fiscal resources and higher quality candidates to dominate the Japanese Diet from 1955 to 2009.

Recent election outcomes, however, suggest a fundamental transformation in this political nexus. Steven Reed, Ethan Scheiner, and Michael Thies (2012) find that in 2005 and 2009, party affiliation supplanted candidate characteristics as the strongest predictor of electoral victory. As voters' preferences became unmoored from personalistic ties to specific candidates, a large number of incumbents lost their seats. This instability in voter preferences manifested as a landslide victory for the LDP in 2005 and, more consequentially, as a massive swing in favor of the Democratic Party of Japan (DPJ) in 2009. This last reversal of fortunes was particularly striking in rural regions such as Kyushu and Shikoku, where the LDP had long enjoyed an electoral monopoly.

In this article, I investigate one overarching question regarding current and future trends in Japan: Do parties have a better shot at winning a majority if they diversify their support base nationally to attract swing voters than if they cultivate a narrower, stable clientele in particular regions? This is of particular relevance to the LDP, whose supporters have historically been concentrated in rural Japan. Although the DPJ won in 2009 by increasing their vote share nationally, should the LDP mimic this approach or, instead, double down on its rural base? Is the urban-rural cleavage in voter preferences, which has long influenced patterns of fiscal redistribution, still salient today?

To answer this question, I analyze two separate factors before and after electoral reform in 1994: (1) the magnitude of incumbency advantage, and (2) the relative weight of rural votes. First, I demonstrate that elections are becoming more "nationalized," meaning that personalistic support no longer insulates incumbents from national shifts in voter sentiment. Nationalization is a result of greater voter attention to the attractiveness of party leaders than to individual candidate qualities. Its extent can be estimated through the magnitude of electoral volatility--fluctuations in vote share over time--and cohesiveness--covariance of vote shares across districts. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.