Thanks, Kentaro Fukumoto, for pointing out the paper that Masataka Harada and I are working on about Japan's election deposit.

For those who are interested, we use regression discontinuity design to evaluate the deterrent effect of the election deposit (on frivolous repeat
candidacies) during the 1947-1993 SNTV period. Japan has historically had one of the highest (if not THE
highest) election deposits in the world. At the moment, it is 3 million yen for SMD races, but throughout the 1970s it was 1 million yen, and throughout the 1980s and until the electoral reform it was 2 million yen.
Because candidates who either marginally lose or retain their deposit can be treated as observationally equivalent in all other respects (quality, effort, etc.), a difference between the two groups allows us to estimate the causal effect of losing the deposit on a losing candidate's decision to run again or quit.

Our findings indicate that the election deposit did not serve it's supposed purpose, as Ellis mentioned, of deterring frivolous or "freak" candidacies (at least with regard to those candidates who tried to run at least once). While losing candidates nominated by serious parties (the LDP, JSP, Komeito, etc.) did indeed tend to quit (or lose the nomination) after the losing the deposit, independents, JCP candidates, and candidates from minor fringe parties were not so deterred.

These results call into question the raison d'être of the election deposit (if it was indeed to deter such "freak" candidacies). Since the introduction of SMDs, the overall number of candidates (e.g., from the JCP) has decreased as expected given the new mechanics of the electoral system, but new fringe parties like the Happiness Realization Party still run candidates in SMDs where they have no chance of winning (and nearly always lose the deposit). And "freak" candidates like the eccentric inventor Dr. Nakamatsu and "smile therapist" Mac Akasaka (of the "Smile Party") are notorious, especially in the House of Councillors elections. Perhaps the national exposure and free NHK air time are worth the cost of the deposit to these candidates/parties.

Dan

--
Daniel M. Smith
Postdoctoral Fellow
Walter H. Shorenstein Asia-Pacific Research Center Encina Hall, Stanford University