In recent years, studies on policy advice have highlighted a process of change in advisory systems showing trends towards externalisation and pluralisation. These processes are particularly relevant for Napoleonic politico - administrative systems where the role of central government cabinets represented a centre of gravity based on political appointment and monopoly of legal and procedural know how. This arrangement is now incrementally slipping towards a partially different adjustment where politicisation and personalisation of advisory relationships are leading towards more contingent arrangements that include pluralisation of locations, sources of knowledge, types of advisors and mechanisms of their recruitment. This means that after a long phase of static dominance of a solely cabinet-centred approach, we are now observing a dynamic adaptation towards more contingent, fluid, promiscuous and multi-location system.

This paper presents evidence on this process of policy advisory system change in Italy, where the process of pluralisation according to contingent arrangements is taking on particular relevance in terms of intensity and variety of organisational forms. The purpose of the paper is providing the reader a set of assorted evidence to display the different ways in which advice is practised in the case study in order to isolate emerging trends.

This dynamism is illustrated in the paper through three clusters of evidence: (i) a specific focus on policy advice arrangements and practices in two Italian governmental arenas (the second Conte government 2019-2021 and the Draghi government 2021-2022); (ii) a focus on advice in the Italian parliament; (iii) a sub-national focus on advice in some Italian regions. The empirical evidence comes from the second phase of a research promoted by the Italian National School of Government (SNA) of the Prime Minister's Office through a pool of academic and institutional researchers.

In recent years, governments worldwide have sought to further the systematization and institutionalization of expert knowledge in processes of policymaking to increase the use of such knowledge by policymakers. One example of the institutionalization of knowledge is the establishment of policy evaluation systems. Evaluation systems can be understood as institutionalized evaluation cultures, which come with certain epistemic standards, regulations and processual guidelines for the evaluation of planned, ongoing or completed policies. Research on knowledge utilization have thus increasingly turned toward the question of how to develop and maintain such evaluation systems for them to increase evaluation use. However, we still know little of the actual effects of these evaluation systems and whether they further some forms of evaluation use rather than others. This paper addresses this gap by analyzing how the organization of evaluation systems within the central administration of the Netherlands and Denmark influences the function and use of evaluations in policymaking processes. We specifically focus on the role of the Ministry of Finance, as this ministry provide the institutional anchorage of the evaluation system in the two countries, as well as in many other OECD countries. We apply the concept of the constitutive effects of evaluations to elucidate how the ministry of finance not only influence the function and use of the evaluation system through formal (regulatory) means, but also through informal means (e.g. norms and discourses) which create unintended consequences. The two cases are analyzed through a combination of document analysis (evaluation guidelines and evaluation reports), and a series of interviews with centrally placed civil servants in the Ministries of Finance, Court of Audit, and the Ministries of Labor and Public Health in both countries. While it is often assumed that the systematization and institutionalization of procedures for knowledge production, -dissemination and -use within policymaking organizations and processes will naturally increase the influence of such knowledge on policymaking, our study clearly nuances this assumption. We highlight how the anchorage of the evaluation function in the Ministry of Finance constitutes a specific economic outlook on evaluation, which 1) narrows down the applied evaluation methods and criteria; 2) changes the primary evaluation user from the parliament to the ministry of finance; and 3) inserts accountability rather than learning as the main form of use. Both evaluation systems thus mainly supports "defensive" functions of keeping checks and balances on cost spending – which ultimately hinders the systems in supporting more significant policy changes and forms of learning.

Policy evaluation is part of the democratic life of many countries, and provides governments and the public with an external perspective on public action. This democratic function of policy evaluation lies in tension with the expert background of this type of knowledge production, that sometimes triggers accusations of technocracy. Recently, evaluation authors have called for a renewal in evaluation criteria used to assess public action, arguing for instance that sustainability should become a key criterion (Patton 2019) next to the more traditional ones (e.g., effectiveness, efficiency). The argument raised is that of the urgent threats faced by humanity, among other, cast a shadow on the democratic governance of common goods like for instance health or environmental resources. This contribution raises the question of evaluation criteria in evaluation research, i.e., the dimension against which a policy program is assessed. Based on a literature review of policy evaluation articles in the international scientific literature (N=179 analyzed articles), it compares evaluation criteria used in five different policy area: social services, land-use planning, teaching in higher education, vocational education, and environment. First, it identifies the commonalities and difference in evaluation criteria across policy areas; Second, it examines how the criteria were chosen in the reviewed studies (e.g., based on the programs' specification, on participatory processes, on expert processes). Finally, it reflects on the tension between evaluation routines, field requirements (research traditions, international standards), and the calls recently made to develop innovative criteria able to meet the challenges of democracy and sustainable transition.