Japan held two by-elections yesterday – one in Hokkaido 5th District, which has been without an MP since the death of veteran LDP lawmaker Machimura Nobutaka last summer, and one in Kyoto 3rd District, whose scandal-hit LDP MP Miyazaki Kensuke resigned in February. The LDP held the Hokkaido seat and lost the Kyoto seat – a net gain for the opposition, but not one from which the nascent Democratic Party can take much comfort, because the nature of the results raises tough questions about the much-vaunted “opposition coordination” approach.
First, let’s look at the seat the Democratic Party won – Kyoto 3. The DP candidate, Izumi Kenta, won with over 65% of the vote, gaining over 10,000 votes compared to his performance in the 2014 election. Turnout, however, was the lowest ever in the post-war era, at around 30%, not least because the LDP didn’t actually contest the seat. Miyazaki Kensuke’s scandal (he was caught in an adulterous affair only days after making a big deal of taking paternity leave to support his wife following the birth of his first child) was headline news for days and provoked a huge backlash; the LDP wrote off the seat and chose not to run a candidate. Second place in the race, then, went to the newly minted Initiatives From Osaka, whose candidate managed less than a third of the votes of the DP candidate. Incidentally, Izumi is already a member of the House of Representatives – he lost Kyoto 3 in the last general election but was elected on the DPJ’s proportional list for the Kinki region. The new DP lawmaker joining the house, then, will be Kitagami Keiro, who takes over Izumi’s proportional list seat.
Kyoto 3 doesn’t really tell us much useful about the shape of Japanese electoral politics, then. “DP candidates win seats which the LDP has pulled out of after hugely embarrassing scandals” isn’t headline news, and the low turnout makes it impossible to measure any possible influence which the tentative detente between the DP and the Japan Communist Party has had; the JCP didn’t run a candidate in the race, but whether that contributed to Izumi’s vote total and in what degree is impossible to calculate.
Hokkaido 5, then. Turnout here was a lot healthier than in Kyoto, at 57.6% (down less than a single percentage point from the 2014 general election), and the election presented a perfect laboratory for checking on the potential of opposition coordination to tackle narrow LDP leads. In 2014, the LDP candidate faced off against a DPJ candidate and a JCP candidate, winning 50.9% of the vote to the 49% won by the opposition parties (36.8% for the DPJ, 12.2% for the JCP). In this by-election, the DP and the JCP backed a single candidate (along with the People’s Life Party, the Greens and the Social Democratic Party), going up against a non-incumbent LDP candidate (with the backing, of course, of the LDP’s coalition partners Komeito, along with a couple of fringe conservative groups).
To my mind, there are three types of seats which opposition coordination can target. The first are “marginal seats” – places where the DPJ was within a few percentage points of the LDP in 2014, and where the support of just a small fraction of JCP supporters would swing the election. There are nine of these seats, and honestly, the DP should be aiming to win them in the next election without help from other parties – if it can’t reverse a few percentage points in marginal seats when competing against a government whose core policies are all disliked by voters, then the whole purpose of its existence as a party is in question. The second type of seat is “opposition majority seats” – places where a simple mathematical combination of votes for DP and JCP candidates in 2014 would have yielded a majority. There are 70 of those seats (67 in which the DP candidate could have won with JCP support, and 3 where the JCP candidate could have won with DP support) – enough to deliver a powerful blow to the LDP’s majority and probably end Shinzo Abe’s premiership, but not enough to reverse the LDP’s lower house majority.
The third type of seat is the “combined opposition marginal” – a seat where a combination of opposition votes in 2014 would have put them within a couple of percentage points of victory over the LDP (or Komeito) candidate. Hokkaido 5 is a perfect example of this kind of seat, requiring not just good coordination between opposition parties but also a few percentage points of support swing (or a boost in turn-out, breaking strongly for the opposition candidate) to shift control of the seat. If the DP (and other opposition parties) can start to make breakthroughs in this kind of seat in the next election, it blows Japan’s political landscape wide open for the first time in many years – perhaps not giving the DP another shot at government, but at least forcing the LDP to work with other parties to pass key legislation, and putting Abe’s more ambitious goals, like constitutional amendments, out of commission entirely.
Getting there, though, is going to be an uphill struggle. In Hokkaido 5 yesterday, the LDP won by over 12,000 votes. The combined DP and JCP candidate didn’t make up the gap between opposition and LDP at all; in fact, her vote total of 123,517 was around 3000 votes short of the combined vote totals of the two parties in 2014. The LDP’s Wada Yoshiaki, meanwhile, picked up 4,500 more votes than his veteran predecessor had commanded, despite the lower turnout.
What can we conclude from this? Well, the opposition coordination idea works, in nuts and bolts terms; with the DP and JCP supporting the same candidate, that candidate was able to pick up almost all of the votes that had previously gone to the two parties separately. This undermines the narrative from the DP’s centre-right figures, who claim that working with the JCP will cost the party scores of votes from centre-right voters; a claim which has always seemed dubious to me, since I’m not sure the DP really has many centre-right voters to begin with. It also assuages concerns that JCP voters, having stuck with the party through thick and (mostly) thin, would balk at casting a vote for a DP-backed candidate. Most voters dutifully turned out to cast for the coordinated candidate, which bodes well for the 60 to 70 seats that could be turned to the opposition simply by effective coordination strategies.
That’s the positive. The negative is that if the opposition can’t win Hokkaido 5, it’s dubious whether any of the “combined opposition marginals” are within its grasp at the moment. Essentially, the opposition has not increased its popularity since 2014; if anything, it may have slid backwards. In order for the opposition to win power, or even to seriously threaten the LDP’s majority, it needs to boost turnout, convincing disaffected voters to go to the polls and vote – many of them for the first time since 2009. Hokkaido 5’s result suggests that the DP is even further from that outcome today than it was two years ago.
There are of course local factors in play, and it’s unwise to project the political fortunes of a nation from a single by-election in a peripheral constituency; but this was the first real-world test of opposition coordination, and its results suggest a low ceiling on what the DP and its allies can achieve through this strategy alone. On a good day, opposition coordination might cost the LDP enough seats to put Abe’s future in doubt; but even on a very good day indeed for the opposition, it would take far more than coordination between parties to kick the LDP out of power. For that, the opposition needs to offer what it has failed to offer since 2012; an attractive, clear, direct and credible alternative to the LDP’s policy platform.
On a related note – if coordination within Japan’s political opposition interests you, I did an interview with Michael Penn of the Shingetsu News Agency a couple of weeks ago on this topic – I’ve embedded the video from their YouTube channel below.