«by Dyron Keith Dabney A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Political Science) ...»
It is evident that Usami, in the most urban district, and Matsuzawa, in what we have termed the suburban district, are substantially similar, while Genba, our rural district case study is different. The key is perhaps that Usami’s district is largely Ohta ward, which is more oriented to small business than most of Tokyo, while Matsuzawa’s district is all within Kawasaki City, which is the eighth largest city in Japan and right in the middle of the greater Tokyo metropolitan area. The heterogeneity of its sub-districts does make it suburban in character, but it is on the urban side of suburban.
While the tactical choices made by Usami and Matsuzawa cannot be differentiated in any meaningful way purely on the basis of their survey responses, the contrast between both of them and Genba is quite clear. Genba’s top factor was
associational support was also a big difference, 12 for Genba compared with 8 and 7.
Party images, on the other hand, were seen as much more crucial in the more urban districts, 17 or 18 points compared with 6 for Genba, as was media attention, at 15 or 16 in the urban districts versus 8 in the more rural area. These differences are exactly what is meant by the distinction between Type A and Type B campaign strategies. The graphic presentation of the survey results in Figure 5.1 visually illustrates the behavioral modifications of each candidate between elections.
Lastly, the tactics that do not differ between the urban and rural campaigns are worth a bit of consideration. First, personal image: differences between the Type A and Type B approaches might possibly lie in what sort of image the candidate wants to project—perhaps paternalistic and soothing for Type A, and an energetic change-agent for Type B. However, the descriptive evidence does not show any such distinction—all three candidates tried to look fresh and lively and standing for reform and change.
Similarly, the high position of policy issues among all three could be covering a difference in what sorts of issues are emphasized. As much of the literature on electoral systems described in Chapter 2 has pointed out, we would expect Type A candidates to stress particularistic programs for a locality, traditional pork, and Type B candidates to emphasize more general policy issues. Again, that was not the case in the campaigns we observed. All three candidates talked more about “reform” and “change” than anything else. Indeed, it may not be too much to say that for them the term “policy issues” was more about conveying an image of being interested in policy issues than about appealing
While no one would claim that this “survey,” with only nine “cases,” is conclusive, it is still quite interesting. By and large, the results of this objective, quantitative exercise bear out what was discovered in the subjective, qualitative research reported in Chapters 3 and 4. For example, the Type A election strategy certainly has not disappeared. Beyond that, the survey results help us rethink some assumptions that underlay the research, such as the ambiguous meaning of “policy and issues.” Accounting for Variations in the Campaign Behavior of the Case Studies We turn now to trying to account for the differences we have seen among these campaigns, drawing on the earlier two chapters supplemented by our survey, in order to throw light on the dynamics of election campaigning in Japan.
We have seen variation over time and variation across districts. What factors account for the variations we have observed? Answers to this question throw light on broader issues, including explanations for why campaign behavior in Japan seems to be so distinctive. In particular, we focus on the most distinctive aspect of Japanese campaigning, the widespread use of the Type A election strategy. Way back in 1971, Curtis opined that the Type A strategy he described so well would, over time, give way to the Type B strategy. As the literature from that time through the electoral reform in 1994 to today even indicate, the continued importance of the Type A strategy and its failure to disappear remains a central preoccupation of scholars in this field, so we should take it up
obvious, for explaining the prevalence of the Type A strategy in election campaigns.
1. The Type A strategy is the product of values and norms among voters, which are rooted in Japanese social and cultural patterns.
2. The Type A strategy is the product of candidates’ expectations, based on their own experiences or what they are taught about what works to win election campaigns.
3. A Type A strategy prevails because an effective Type B strategy is precluded by tight election laws and regulations.
4. The Type A strategy is the product of the multimember electoral district system.
These hypotheses are not mutually exclusive—in fact, all of them could easily be true.
The question is therefore whether they seem to be born out by evidence, and how much explanatory power each has. They are more or less susceptible to testing with our qualitative and quantitative data. The fourth hypothesis is of most interest in the literature and we will examine it the most, including a couple of sub-hypothesis, but the others are significant enough to discuss briefly.
Hypothesis 1: The Type A strategy is the product of values and norms among voters, which are rooted in Japanese social and cultural patterns.
As noted in the introduction, the conventional wisdom as well as a substantial literature on Japanese electoral behavior link the expectations of Japanese voters, and therefore the campaign behavior of Japanese candidates, to norms and values in broader society. These values and norms are best captured in the “vertical” social organization in Japan structured around social relationships, exchanges and obligation (e.g. Richardson, 1974; Flanagan 1991; Richardson and Flanagan 1984; and Ike, 1972). This hypothesis would seem to be supported by the fact that the Type A strategy is still quite common,
other hand, Japanese values and norms appear to have changed a good deal as well.
Within our study, the time span is too short for a direct test of this hypothesis, but an indirect test is provided by the urban-rural dimension. Although no doubt a little too simple, the proposition that values and norms in urban areas are more “modern” and in rural areas more “traditional” is overwhelmingly accepted in Japan as elsewhere. The fact that our campaigns span both types of districts thus offers some insight.
To summarize, rural voters are seen as socially integrated, cooperative, and personally connected to political leaders, while urban voters are less socially integrated, have higher levels of political participation and express greater knowledge about issuepolitics (particularly national politics). Thus, election campaigning in rural districts is seen as constructed around the personal vote and organized political machines (networks of local politicians, organizations and associations). Conversely, in an urban election district, election campaigning is seen as constructed around party labels and policy issues.
Our quantitative analysis above reflected these contrasts (the ambiguous nature of policy issues aside), and the contrasts between Genba’s rural approach and Usami’s urban approach were clear in Chapter 4. It is Matsuzawa’s “middling” or suburban case in Chapter 3 that is most interesting. The quantitative responses above make him nearly indistinguishable from Usami, but the qualitative evidence gives a different impression.
The most telling point is that Usami almost ignored organizational considerations in two of his three campaigns, while Matsuzawa paid a great deal of attention to organizing and maintaining his kouenkai, building up relationships with various associations, and
that regard Matsuzawa looked considerably more rural.76 The evidence of Matsuzawa’s rural-like behavior needs a somewhat broader look.
One could argue that Usami was not a very skillful candidate to start with—his initial victory could be called a fluke, and he lost in his next two tries.77 Compared to the other two candidates he was deficient in planning, and in mindfulness about how campaigning works. And it is worth noting that his opponents in 1996 and 2000 did make more efforts at more traditional styles of organization despite the urban district, to good effect. In short, the evidence from Usami’s own campaigns notwithstanding, it appears clear that Type A campaigning persisted at least to the end of the 20th century throughout Japan.
As to whether it is rooted in voters’ values and norms and in Japanese culture, we will present other possible alternatives below, but it is worth noting that the candidates themselves certainly thought so. Chapters 3 and 4 are full of examples of Matsuzawa and Genba trying to meet the expectations of their voters.
Hypothesis 2: The Type A strategy is the product of candidates’ expectations, based on their own experiences or what they have learned.
Curtis (1971) speculated that Japanese voters would change fairly rapidly but politicians would take a long time to catch up. There is a great body of practical precepts about how to win Japanese elections, largely based on assumptions that might be right or wrong about Japanese voters. Most of the precepts are associated with the Type A strategy. The precepts are reinforced in people’s minds by the cases of candidates who 76 Matsuzawa’s survey responses, curiously, say otherwise, rather that he is more urban. The data collected through interviews and participant observations suggest that candidates’ self-reflections of campaign behavior do not always coincide with observed behavior.
77 Although our study only accounts for campaign behavior up to the 2000 general election, it is worth noting that on his third attempt in 2003 Usami did finally reclaim a Lower House in the Tokyo 4th district
candidates who rejected the precepts and succeeded. Tsurutani (1977) found enough of these to proclaim that “new style campaigning” or, in effect, the Type B strategy was already coming to the fore and would be the wave of the future, as candidates caught up to “post-industrial,” post-materialist shifts in Japanese society. However, whatever trends he detected did not seem to take hold, or in any case did not really shake the conventional wisdom.
Our study provides a good test of whether the conventional wisdom does lead politicians to opt for Type A behavior, not by comparing among our three candidates but by seeing if all three were different from other candidates running for the Lower House, particularly in 1993 general election. That is because this election, their first at the national level, came fairly shortly after they had graduated from the Matsushita Institute of Government and Management (MIGM). That school was explicitly established to train a new kind of politician geared towards Type B rather than Type A campaigning. It was seen as an American approach as opposed to a Japanese approach to campaigning, and most MIGM students including our three candidates spent time in the US as participant observers of American politics and electioneering.
In interviews after the 1993 election, all three spoke of how valuable the MIGM experience had been to them. Matsuzawa even chronicled the electoral lessons he learned while in the US supported by MIGM in a 1986 Chuo Kouron publication. One obvious contribution was that they saw their attendance there as central to their images as fresh, young, innovative, policy-oriented candidates. They also thought that what they had learned was helpful in their initial efforts to structure a national-level election
three candidates in 1993 appeared considerably more Type B than the conventional precepts, more Type B than what we know in a general way about the usual strategies and tactics of Japanese candidates, and more Type B than the strategies of some of their opponents.
Several caveats are warranted, however, with respect to the candidates’ Type B orientation in 1993. First, 1993 was a year of unusual turmoil, with widespread criticism of established ideas and institutions in Japanese politics, not to mention the awakening of a new party boom. Our candidates were all more or less “progressive conservatives” who were running primarily against the LDP. That stance leads naturally to the Type B strategy, and if our reference group was composed of candidates who fell into that category, rather than all candidates (including e.g. LDP incumbents) any impact of MIGM training per se would be considerably attenuated. Second, similarly, the fact that this was their first campaign was certainly important. A first-time candidate lacks experience, and he also typically lacks the resources for an effective Type A campaign (although that is much less true when a candidate directly inherits a jiban, or support base of relationships from a politician who has died or retired—particularly a relative).
Kouenkai organizations take time to build, local politicians are unlikely to be impressed by some neophyte the first time he comes around, and so forth.
To the extent that the above is true, we would expect to see a shift in emphasis from the Type B to Type A strategy from the first to the second campaigns. A candidate learns areas where he fell short (and perhaps where his opponents did well). Moreover, as an incumbent, he has the time and resources to work on building up relationships
All three candidates (even Usami to some extent) pay more attention to the organizationally-driven Type A strategy in 1996 than they did in 1993. We will qualify this claim a bit below, but we consider the contrast between 1993 and 1996 as a powerful indication that there is a “political life cycle” for election strategies.
If logic dictates a movement from Type B to Type A in a second campaign, what about a third campaign? Probably there is no strong prediction across candidates.