2.2 Models and the limits of government

Governmental steering models and typologies of university governance require a complementary set of regulations at five or six levels: government, intermediary, institution, faculty, department, individual. The choice of a model and typology seems less relevant for the choice of a juridical approach. Six reasons are worth mentioning; this essay concentrates on the first four.

• the translation of instructive, but abstract, models into manageable juridical instruments is underdeveloped;
• models seldom indicate how to link the administrative layers;
• to draft a complementary set of regulations at six different levels is difficult enough when it is well co-ordinated; when each (semi-)autonomous body directs its own jurists without a solid frame of reference, it becomes a labyrinth;
• the limitations of traditional legal drafting, the juridical tool-kit, are underestimated; it is mainly Weberian 'order by rule' that guides the lawyers;
• conflicts of interests during implementation are solved by compromises that appear in the regulations. This not only perverts the model, but also common sense in daily practice.
• the strength of the unwritten rules of the game (Scott-Morgan, 1994) is underestimated.

Business administration is also in search of The Grail. Even excellent and successful companies that combine virtues as close to customer, hands-on value-driven, autonomy and entrepreneurship, productivity through people, simple form lean staff, etc, are not safe for ever (Peters & Waterman, 1982; Waterman, 1994). The short life-cycle of policy systems (In 't Veld, Füssel, Neave, 1996; p. 46-50), and the dynamics of the internal and external environment, obviously apply to business as well as government. McKinsey’s 7S framework: management through a balance of structure, system, style, staff, skills, strategy, shared values, is still popular. In this model, rules appear only hidden in 'system' and 'structure'. Translation into juridical concepts is hardly needed. In business, regulating is an exception; the company-structure is leading and there is no government intervention in strategy, skills and staff. The 7S-framework is also influentual in the public sector, but here regulations are the finishing touch of political decisions. However, the legislative process frequently starts after, instead of parallel to, policy development, so embedding juridical elements is left to the last stage: when the jurists make an appearance to draft regulations. Subsequently, the policy has often to be revised during the legislative process, or implementation is considerbly delayed, or both.
An attempt to transform the 7S into a parallel model that would give guidance to drafting regulations, has never been made. The same holds true for higher education steering models that have been developed, prescribed or analysed. A guide to transformation into viable juridical concepts -'a juridical shadow model' is also missing in traditional classifications of steering models like Napoleonic, Humboltian and Anglo-Saxon or State, market, institution-oriented (Clark, 1983), as well as typologies of governance as for example state regulated, state-agency, state-controlled, state-aided, state-supervised (Dill & Sporn, 1995), corporate, selfregulated, loosely coupled (Weick, 1976) or cybernetic (van Vught, 1995), and most certainly also in managerial approaches like: total quality management, 'walking around', 'thinking aloud', entrepreneurship, teamwork, customer focus, business re-engineering, learning organisation, etc.
In what is for practitioners in governmental organisations an interesting dissertation, McDaniel (1997) proves ambiguity when 600 experts outside government, give their views on government steering and supranational action. It should not be a surprise that experts, even when including those in ministries, have no interest in claiming success for government intervention or strongly defending it. It is evident that modern governance requires a modest government. However, a lame or absent government would be disastrous. McDaniel deduces ten explanations of the limited success of government policies, giving us also an insight into the ambitions of steering models and their makers:

• intrinsic limitations of controllability;
• use of classical steering methods (rules do not reach the shop-floor);
• wide-ranging and complex control signals;
• negative interference between policy instruments;
• no sanctions available or applied;
• policy development takes place in an ivory tower;
• budget-cuts as reasons for change lead to policy games;
• political compromise without consultation of those who implement decisions;
• slow and sluggish decision making;
• short implementation-periods, without feed back by evaluation.

McDaniel concludes that the 'how' of government intervention, has become more important than the 'why'. His contrasting steering scenarios are: planning; intermediaries; incentives; autonomy; competition, whereas Mintzberg (1996) distinguishes five models lor managing government:

• government-as-machine (rules and standards);
• government-as-network (Weicks' loosely coupled systems);
• performance control (governance as businesses: as to Mintzberg, a new label for the machine-model);
• virtual government (government by privatisation and negotiating contracts only; the best government is no government);
• normative control (key-elements: socialisation, guidance, responsibility, judgment and selection).