Agenda Setting and Policy Design
In every modern society, there are hundreds of issues that are potentially a matter of government concern, but a limited number actually appear on the policy agenda. Ideas and issues that do not reach the policy agenda will not cause any policy action. Research on agenda setting tells us that this initial stage is a highly competitive game, and because it frames the next stages of the policy process, it is a very important stage (Bachrach and Baratz, 1962, 1970; Cobb et al., (1976). Quite often, reforms are prompted by signals in society, for example, from salient stakeholders or policy entrepreneurs (Kingdon, 1984), that there is a ‘problem’ with the current system. Thus, the first policy questions addressed are: What is the problem? Who addresses the problem? The description and analysis of this policy arena includes an exploration ofthe rationale for the structural reform and the extent to which the rationale and problem analysis are supported by different stakeholders.
Once an issue is on the agenda, policymakers need to decide on a course ofaction. Various alternatives that might solve the issue need to be explored and finally a decision on a set of actions (a policy) has to be determined. Governing means using policy instruments; without them public policies would be no more than abstract ideals or fantasies (Hood and Margetts, 2007). Instruments concern action to shape or change behaviour to pursue the set policy objectives. As regards structural reforms in higher education, in choosing how to achieve its goals, governments have many instruments to select from. Mergers for example can be imposed on the higher education sector by legislation, but the government can also decide to encourage mergers through financial incentives. Or it may try to settle agreements with some parties, or use a dialogue-based approach to convince the higher education sector that mergers are desirable.
Typologies of policy instruments abound (e.g. Hood, 2007; Hood and Margetts, 2007; Howlett, 2004, 2009; van Vught and de Boer, 2015; Voss, 2007). In this book, we take the point of view that characteristics ofinstruments matter and therefore opt for using a generic approach.

Fig. 1 Analytical framework of structural reforms
A well-known typology within the generic approach is Vedung’s distinction between ‘carrots’, ‘sticks’ and ‘sermons’ (Bemelmans-Videc et al., 1998). Elmore (1987) sees government instruments as variants of four intervention strategies - namely, comprising mandates, inducements, capacity-building and system-changing. Schneider and Ingram (1990, 1997) distinguish comprising authority tools, incentive tools, capacity tools, symbolic tools and learning tools.
In this book, we will stay close to (an adapted version of) the probably most well-known typology: Hood’s NATO scheme, based on the resources a government has: ‘nodality’, ‘authority’, ‘treasure’ and ‘organisation’ (see also Hood, 1983; Hood and Margetts, 2007; Howlett, 2000, 2009; van Vught and de Boer, 2015). In general terms, nodality refers to the use of information and communication. It concerns advice and training to get messages across, trying to affect the cognitive base of the recipients and as the result of that changing their behaviours. Authority tools are intended to command and to forbid, to commend and to permit. Certificates, licences, contracts, quotas, permits and code of conducts are examples of authority instruments. Treasure enables governments to buy favours. It can exchange money for a good or service, or it can transfer payments without requiring any quid pro quo. Grants, loans, bounties and tax expenditures are examples of treasury instruments. Finally, organisation tools, sometimes referred as ‘architecture’, refer to the government’s capacity to establish (institutional) structures such as bureaucracies, agencies, networks or partnerships and the like.
The selection of tools is a delicate process, since tools are not neutral. Aspects that deserve attention in this stage are political feasibility (it is as much a political as a technical process), resource availability and the behavioural assumptions about the targeted populations. ‘The choice of the policy tool is a function of the assumed behaviour of the policy target’ (Birkland, 2001, p. 176).
In the agenda-setting and policy design arena, the power of the views and ideas of the involved actors, their resources and their position in policy subsystems determine the outcomes. Who are these key actors, how do they interact and what is the effect of this interaction on policy design, policy instruments and formulation? (Jordan and Schubert, 1992). With respect to these questions, five dimensions must be taken into account (Forester, 1984): the number of actors in the decision-making (single versus multiple), the organisational setting (closed versus open), the definition of the problem (well-defined versus vague), type of available information (perfect versus contested) and the time available (infinite versus manipulated).