These scenarios or models appear in a mix. Mintzberg advises us to de-emphasise the government-as-machine, to be more appreciative of the network model because of the complex unpredictable environment, and to shift to the normative model, as there is no substitute for human dedication. McDaniel develops an alternative mix of mechanisms of governance:

• competition for students, staff, funds and reputation;
• diversified funding;
• consumer power;
• serving different constituencies, being accountable for that;
• publication of external independent quality assessment.

Whereas McDaniel understates rules, regulations and legal norms as a steering scenario, Mintzberg seems to underestimate the need for rules in the network- and normative-control model, without neglecting the importance of regulations for the other models. It is evident that the development of juridical instruments lags behind the rapid development of refined policy instruments and steering models. This is alarming for societal interests which the law usually defends, as well as for institutions and people that are protected by legislation
Developments in the area of tuition fees in central and eastem Europe which lead to two classes of students, paying and not-paying, without an advanced complementary scheme of loans and grants that is adjusted to the social, economie and fiscal needs of the country, is an example of the gap between models and justice-in-practice. The new act on university governance of the Netherlands (Stb 1997, 117) implements decentralisation, business-like management and enhancement of institutional autonomy, hut also shows a tendency toward autocratic leadership which could be problematic if a society is less open. The corrective mechanism of the State is 'people' the Dutch minister appoints at the supervisory board of public universities, who in their turn appoint the members of the executive board, who again appoint the deans of super-faculties. It is not ‘structure’. A daring experiment.

2.3 Toward a shadow-model for lawyers I

The combination of teaching and research in probably the fastest moving environment in modern life, needs a highly complex mix of operational and human resourees governance. Universities and polytechnics are hybrid organisations, operating in the public as well as the private domain, with a multitude of goals and a wide scope of operations. The implementation of ideas about loosely coupled systems, self-steering organisations and cybernetic governance can easily backfire. An important explanation for the limited success of government policies seems hidden in McDaniels' ten reasons for it. It is elear without further evidenee that higher education cannot be well governed topdown. Governance requires a give and take between equals, be it organisational units or persons. Public law is often considered as too rigid to steer education and research. This is not necessarily so, and also for other good reasons, the concepts of public law are indispensable. Nonetheless, the freedom to make contracts, the frame of reference of contract law, provides a necessary complementary starting point. Complementary, beeause agreements do not suffice to ensure open access, affordable higher education for every social group, democratic governance and other elements that make education a merit-good. In higher education and research, we need to find a balance between democracy through law, and the democracy of the market. We need a balanced approach of regulating by process-design, directives and incentives, complemented by steering on products through agreements. Here, descriptive and comparative literature are scarce. Literature that develops models, abundant in policy studies, is practically non-existent in the field of law, where the word 'model' even has another meaning as we have seen in 1.2.3. Mitnicks' distinction (1980) between regulating by directives and regulating by incentives is helpful. However, it lacks two, perhaps even more important, lateral categories:

• regulating by process-design;
• regulating by product-design.

Process-regulation is effective, especially when interdependencies are complex and limited conflicts of interests frequent. Process-design, the creation of checks and balances, and principles of contract law, shouldform the backbone of legal drafting flor higher education and research. This ensures democracy, and leaves the responsibilityfor the contents ofthe primary processes where it belongs.

Direct product-regulation (output-control) is not really adequate for steering education and research. Directives are needed for a limited number of issues: definition of the system and the task of institutions; to ensure academic freedom; open access. Institutional autonomy is ensured by a clear division of responsibility, and regulation of processes to prove accountability.
Democratic governance can be strengthened through regulating by process-design: the buildmg of checks and balances (elections, interaction during decision making). Efficient governance can be enhanced through regulating by process-design; not having to search for decision makers.
Regulation by a limited number of incentives is sufficient to ensure output. This requires patterns of interaction. E.g: if quality assurance is meant to improve quality, if that is the product of university, it requires an extremely well-designed process, hut not much control.
The key-concepts of policymakers also need translation. For example, 'self-management' may be clear for a manager, but 'responsibility for' and 'self-maintaining checks and balances' are more readily linked to the literature and toolkit of the jurist. Further steps cannot use traditional legal language either. 'Shall' or 'shall not' are fine for the Penal Code and business contracts, but applying modern steering methods requires that processes are captured by regulations. Jurists must be brought into a position:

• to write out the processes of cooperation and decision-making, including interdependencies;
• to work out the specifications of the chosen steering model;
• with a view to the chosen typology of management;
• with a view to the specific constituency;
• in such a way that regulations do not need to be changed when a manager resigns (the rules of the game do not necessarily change when the players do), and the own professional input of jurists is respected.