Networks cannot replace the States' task of social engineering (in other words: connecting various networks) for the benefit of society, as is the European tradition. Networking easily neglects process design. Thus, regulating and financing will remain intervention instruments of the State. The question is about typologies of regulation.
An example of what 1 would call 'network-process-design': the Dutch government imposed a quality assessment system by trading it for more autonomy, programming freedom and accountability ex post. Legislative instrumentalisation took place according to the following instructive principles:

• the institution is responsible for developing new study programmes and guaranteeing their quality;
• universities are consulted collectively to ensure co-operation and efficiency at the national level;
• the advice of a mixed commission (introduced by an amendment in Parliament) has to be sought before a new programme is registered;
• the new programme of an institution that is listed in the appendix of the law, is automatically recognised, unless the commission advises that the new programme should not be registered and the minister decides accordingly. The grounds for rejection can only be expected quality and national efficiency. If the programme is not registered, the diploma is not recognised and the programme is not State financed;
• the institution organises self-evaluation; the institutions collectively organise external quality assessment; the semi-independent State Inspectorate evaluates the evaluations and oversees their follow-up.
• the programming freedom is founded in some simple provisions in the framework law, whereas the system of quality assurance is only mentioned in general terms. However, it is extensively regulated by agreements, manuals and international networking.

This is a form of crypto-regulation that redefines the relationships, hut does not change the intensity and has nothing to do with deregulation. The example shows that the interconnection of two complex networks, of the State and of higher education, is an important, hut potentially weak link of the chain. Solving the problems on the interface of public and private requires strengthening by transparent co-ordination that effectively connects the State and the institutions. Regulation by network-process-design may improve transparency and generate mutual faith. Precondition is that the participating networks play according the rules of the game.

5 Preparing new regulations

5.1 Fewer checks, more balance

Could we detect features of regulations that increase the chances of success: a checklist to improve the governance of complex hybrid higher education? Is it possible to design 'intelligent regulations', that stimulate an autonomous drive of institutions to seek a democratic form and flexibility, improve core business, and help to survive budget instability? Is it possible to design a juridical model in which strong State control is banned, while the State and its taxpayers as well as the academic community feel comfortable? Structures of governance become essential when trends change. An organisation is often too late to change its structure and style; it cannot move in time. Moreover, much goes wrong at the beginning of the development of juridical structures that govern (new) institutions or units.
Jurists 'quest for The Grail should have a low level of ambition: creating juridical structures that facilitate anticipatory governance; matching efficient, effective and democratic governance. This requires continuous attention to the reduction of checks and improvement of balances at every level of the system, as well as between the levels: Parliament, government, intermediaries, higher education institutions, faculties, departments, individuals.

5.2 State intervention

Legitimate interests of the State are reasons to limit institutional autonomy, for example:

• public welfare (economy, culture, intrinsic value of educated citizens);
• ideological and political motives (social and international mobility, societal change, supply to the labour market);
• accountability (public investment, efficiency, budget cuts);
• mitigation of undesired market mechanisms;
• corrective policies (regarding enrolments, supply, intrinsic conservatism, academic games);
• protection of vital interests (students, staff).

Each intervention of the State in higher education affairs may be acceptable as such, but the sum of them might not be. Allocation models, directives, incentives, performance indicators, quality assessment, end-terms, are restrictions on autonomy. This seems a bit much; a reason to look for minimum steering methods. As said in the above: fragmented discussion on each and every element of autonomy is not desirable. On the one hand, public authorities should show transparency and predictable behaviour. Regulations should filter out irrational behaviour. On the other hand: the institutional leadership cannot simply request flexible legislation and 'carte blanche'. Checks and balances require an elaborate ratio of participatory and professional management, and legitimate State intervention.

5.3 Objects and objectives from a juridical perspective

To fine-tune regulations, the objects and objectives of higher education governance should be specified. Three elements:

• An institution should, maybe even be legally obliged to, carry out regularly a long-term market analysis of the regional, national as well as international dimensions of supply and demand. The 'market' of an higher education institution is multidimensional and highly dynamic.
• The analysis should be followed up by a strategic self-assessment of what the institution is now; what it wants to become; and what the constraints and feasibilities, strengths and weaknesses are (Figure 3).
• To timely draw the consequences from such analyses, a solid operational structure that facilitates changes is a prerequisite. Dynamic markets and complex production processes do not comply with traditional rigid regulations on governance that are based on annual routines.