Just off the top of my head, I don't think it should be difficult to go through case studies dealing with the processes of policy change (which is what most case studies describe) and track the role of bureaucratic leadership.

In the several cases I know of, or have read of, the leadership of the *Kyokucho* [Bureau Chief] is a strong indicator for major policy changes. And in most cases, parallel ad-hoc *shingikais* are created, along with the permanent and formal ones" Such would be the case when the bureau (bureau chief) wants to initate a major policy chage, and has to pursuade related interest groups (often closely related to each section--*ka*). For less conflicting issues, the *somu ka* [General Affairs Section] of each bureau functions as the "first among equals."

Where cross-bureau coordination is necessary within ministries, then one usually sees "working groups" within the ministry to hammer out differences with or without the *private shingikais* of the Kyokucho. Even in such cases, I think the role of the *kyokucho*s are critical because the only person that outranks them, the vice-minister, should prefer to be a more-or-less nutural arbitrator.
And often in such cases the *Kanbo-cho* [Chief Secretary] is expected to function as an inter-bureau moderator before matters get out of hand. Similarly, in inter-ministerial disputes, it is usually played out as an inter-bureau negotiation (eq. recurring disputes between MOF-Banking Bureau and MOPT-Savings Bureau), [MOPT: Ministry of Posts and Telecommunications] although of course the ground work is done by his/her subordinates.

I have mentioned above (which might well be mistaken) is because my impression is that (a) the leadership of the *kyokucho* is required to initiate a major policy change (even if it fails) and (b) the Japanese bureaucracy is set up and has created several devices (as the ad-hoc shingikais and working groups, or the somu-ka, daijin kanbo [ministerial secretariat] bureau) to transform policy differences into inter-jurisdictional differences, instead of cross-jurisdictional factional disputes.

Even if there are some serious personal disputes reinforced by policy disputes (or vice versa) they tend to be acted out as inter-bureau disputes. Otherwise, if inter-bureau disputes get out of hand, it could invite the intervention of professional politicians which could lead to a politically divided agency at the serious cost of losing ministerial autonomy. Furthermore, inter-jurisdictional disputes will reduce bad blood from accumulating within the ministry, since subordinates are expected to move to different positions within three years.
Whether intentionally or not, rotating positions also seems to me to make policy disputes jurisdictional ones as opposed to personal or factional ones. Thus the rigid procedures and standand operational devices seem to discourage the creation of personal policy factions.

On the other hand, I am not sure how "bureaucratic factions" can be related to shigikais. Also, I think one has to be aware that the information/ideas career bureaucrats obtain that shape their policy orientations derive from their positions (by interacting with client groups, specialists, etc.) and not from other sources of training or academic work.

But I think whoever is interested in the Japanese buraucracy should also consult Barry Keehn (at Cambridge Univ. ) who has done public administration work on bureaucratic coordination and should know a lot about the inner workings of the bureaucracy in Japan. (more than myself. Then why the long post?...beats me!)

Nobuhiro Hiwatari