This post continues my series on the elections of September, 2005.

As I reported in the earlier post, this election was called early after Prime Minister Junichiro Koizumi, head of the Liberal Democratic Party (LDP), lost a key vote (in parliament’s upper house) in early August on a bill to privatize Japan Post. More than a post office, this agency functions as the world’s largest financial institution, and has long been a massive source of off-budget funding for pork-barrel projects that help LDP politicians maintain their personal popularity.

The most recent polls contain good news for Koizumi, as the LDP leads its main opposition, the Democratic Party of Japan (DPJ), in the party preference, 27% to 18% (fully a third are undecided, however).

I want to focus here on the larger issues surrounding this election. The LDP has long had a well deserved image as being corrupt and wedded to pork-barrel politics. It has long contained factions that compete for control of the party. In that context, most prime ministers have had short-lived tenure, which makes Koizumi’s reign—dating to April, 2001—remarkable. Koizumi has tended throughout most of his premiership to be more popular than his party, presumably because he has cultivated an image as a reformer, although some insist it is his hair.

Japan’s pork-barrel politics and the factionalization of the LDP cannot be understood without reference to its former electoral system, the single nontransferable vote (SNTV).

Used in Japan from the late 1940s until 1993, SNTV means that legislators are elected from relatively small but multi-seat districts. The voter has a single vote, and that vote can help elect only the candidate for whom it was cast (this is the non-transferable part). Without giving away everything that I will want my students to figure out in class this fall (some might be reading this!), I will just say here that this system means that parties have to be careful to ensure that their voters are “distributed” to one candidate rather than another in the same district.

Here is where the pork barrel comes in: Pork is something that can be attributed to individual legislators, unlike the party label and program, which is something that competing candidates of the same party share (and hence is useless at differentiating them).

For decades LDP leaders tried to change this system, but the legislators elected under it always resisted. In 1993—as a result of the party split that resulted in the LDP’s brief period in opposition—the electoral system was changed.

The new systems is about 60% single-seat districts (one candidate per party, plurality wins) and the rest is party-list proportional representation (PR). Voters have two votes: one for a candidate, the other for a party.

Unlike in Germany and New Zealand—which have superficially similar mixes of plurality and PR, and also have elections this month—a party in Japan wins its proportional share of the PR seats (based on the party vote) PLUS however many seats it wins in single-seat districts. In Germany and New Zealand, on the other hand, a party is entitled to a share of seats in parliament based only on its party vote; these seats are comprised of the single-seat districts won plus whatever number of PR seats are needed to get it up to its total entitlement. The difference sounds terribly wonky, I know, but it is critical, and boils down to the following:

In Japan, it still matters how many candidates win their own district races, whereas in Germany and New Zealand, it hardly matters at all.

And here is what is really worth watching. Who will win, the traitors or the assassins? In the vote on postal privatization, 37 LDP members voted against the government. These are the traitors, and some of them are quite senior within the party, and even include some former cabinet members. Enter the assassins. This is the term being given by the media to the new candidates that the LDP has recruited to run against the traitors.

In the past, under SNTV, it was common for dissidents to run as independents and then get invited back into the party if they managed to win. Some traitors are still counting on that. But Koizumi has vowed that he will not bring any of them back in, even if they are needed to give his party a majority in parliament.

If many of the assassins—recruited for their personal style, rather than their personal connections—were to beat the traitors, it would be a powerful signal that the new electoral system is working as intended. It was supposed to generate a more “party-centered, policy-centerered” politics, in the words of a prime minister who tried, and failed, to reform the electoral system over 40 years ago.

By eliminating intra-party competition, the new system was supposed to generate campaigns based on the party and not the individual. In fact, much was made in the last election of the party manifesto presented by the opposition DPJ. This was something of a novelty in Japan: the idea that parties campaign as units, instead of collections of inividual candidates, each promising something to a small swath of voters.

In fact, in the party-list vote in 2003, the DPJ actually outpolled the LDP (37.4 to 35%), suggesting that its “reform” message was resonating with voters. Nonetheless, the LDP won 237 seats and the DPJ only 177, on account of the LDP doing so much better in the races for single-seat districts (43.9% of the vote, to DPJ’s 36.7). Part of this better performance in districts than in the party-vote contest was due to the LDP’s governing coalition partner, Komeito, not competing against it in most districts, and part of it was due to the LDP’s ability to field experienced and well-connected candidates who are more popular than the party itself. Both factors matter under Japan’s new electoral system, because a party that can win a lot of districts can win a disproportional share of seats in parliament.

That the personal appeal of candidates is still critical in winning single-seat district races in Japan is revealed strikingly by the types of candidates Koizumi recruited to be his ‘assassins’: Many of them women with little or no experience in politics. In a country like Japan, where political and business connections have always been paramount in elections and where the rate of representation of women as legislators is one of the lowest in the world, these women and male newcomers have created a stir. As one of the candidates, Yuriko Koike, a former news anchorwoman, said:

The problem is that in the past, it was difficult for newcomers and women to be fielded as LDP candidates. But Koizumi is turning the LDP into a new party.

In this election, the DPJ has been caught completely off guard by Koizumi’s gambit of making the election a ‘referendum’ on his postal privatization plan and of drawing media attention to the assassin-candidates. The DPJ (called Minshuto in Japanese) claims to be a reformist party, in contrast to the LDP, but it objects to specific provisions of Koizumi’s plan for Japan Post, and voted against it in parliament. It even tried to avoid taking a stand on the question at all in this campaign. One DPJ candidate expressed his fear that

the LDP would steal the show and paint Minshuto as an “old guard of vested interests” because few in the party were willing to discuss the postal privatization issue.

If the DPJ looks more like the party of vested interests than the LDP, Koizumi has pulled off, on behalf of his party, one of the great image changes of all time.