This article contains a critical reflection on eXplainable Artificial Intelligence (XAI): the idea that AI based decision-making using AI should be transparent to people faced with these decisions. We discuss the main objections to XAI. XAI focuses on a variety of explainees, with different expectations and values; XAI is not a neutral activity, but very value-sensitive; AI is dynamic and so XAI quickly becomes obsolete; many problems are ‘wicked’, which further complicates XAI. In addition, the context of XAI matters – a high level of politicization and a high perceived impact of AI-based decisions, will often result in much criticism of AI and will limit the opportunities of XAI. We also discuss a number of alternative or additional strategies – more attention to negotiated algorithms; to competing algorithms; or to value-sensitive algorithms, which may contribute to more trust in AI-based decision-making.

Theorieën en methoden uit de gedragswetenschappen betreden steeds nadrukkelijker de beleidsscene. Gedragsinzichten en nudging beloven beleid te verrijken en te versterken. Het begin van deze gedragswetenschappelijke omslag of behavioural turn laat men doorgaans samenvallen met de publicatie van het boek Nudge van Richard Thaler en Cass Sunstein in 2008. In dit artikel plaatsen we nudging in perspectief en argumenteren we dat het concept zowel een zegen als een vloek betekent, en zowel een katalysator als een rem is voor de bredere toepassing en verankering van gedragsinzichten in beleid. Ondanks het aantrekkelijke narratief botst nudging op functionele limieten en ethische bezwaren. Om de gedragswetenschappelijke, experimentele en evidence-based beleidsbeloften alsnog in te lossen, zien we een strategie van steeds verdere verbreding. Het programma van de Behavioural Insights-beweging op basis van vijf pijlers leek in eerste instantie een oplossing te bieden, maar kampt door een eendimensionale interpretatie met interne spanningen. De nog bredere en ambitieuzere Behavioural Public Policy-agenda biedt nieuwe perspectieven, maar moet op functioneel en ethisch vlak nog verder onderbouwd worden.

The discipline of public administration is critiqued for its lack of social relevance. Three issues prevent public administration scholars from social relevant research: (1) the assumption of universal knowledge; (2) the assumption of neutrality; and (3) distance from practice. These issues obscure the role of power processes in both the administrative practice under study and in conducting research, while paying attention to power processes is key to a more socially relevant public administration. I propose a phronetic research approach as an alternative to research that lacks social relevance. Phronetic public administration (1) acknowledges and deliberately uses the complexity of different perspectives of actors; (2) makes power processes visible; and (3) chooses a position in a dialogue about the problem under study by imagining alternative future scenarios. Phronetic research offers a promising, critical perspective on public administration by specifically aiming for a social contribution.

Q-methodologie is een nog relatief onbekende onderzoeksmethode, met veel potentieel voor beleidsonderzoek en -analyse. De benaderingen, doelen en onderzoeksvragen in verschillende toepassingen lopen uiteen, maar vertonen ook duidelijke overeenkomsten. In dit artikel beschrijven we de belangrijkste theoretische en analytische bouwstenen van de methode, en een praktijkgericht 10-stappenplan waarmee men snel zelf aan de slag kan met Q-methodologie. Op basis van een aantal toepassingen van Q-methodologie in Nederland en Vlaanderen laat dit artikel op inzichtelijke wijze zien wat Q-methodologie toevoegt aan de toolbox van beleidsonderzoekers. Naast de theoretische achtergrond van de methode biedt deze bijdrage een praktisch stappenplan voor het gebruik van de methode in de praktijk.

Public Administration as a scientific discipline can increase its scientific and social relevance if it takes a more design-oriented approach. With that I go back to the classical notion of Herbert Simon that public administration is a design science. I advocate revitalizing that idea. This is necessary if we look at the problems governments are struggling with, but also in view of the developments in the environment of our field. But it is also promising because it offers opportunities to achieve innovative, scientific knowledge. The design sciences contain numerous insights into how a public administration design process can be set up, with room for creativity, abduction and intuition. I divide a public design process into five rounds: the round of understanding the issue, defining or demarcating it, coming up with possible solutions, testing and refining these solutions and distilling the more generically applicable elements. This is ideally done through fast, iterative rounds of developing, testing and refining designs. In such a design process, disciplinary knowledge can play an important role in understanding the issue and coming up with possible designs. And so designing can also become a way to validate this knowledge.

A large number of people, institutions, journals and approaches have contributed to the history of (local) administrative sciences in the Netherlands. Initially (around 1914) the legal approach was dominant; from 1964 onwards, political science would become the dominant approach; and from 1990 onwards, Public Administration would increasingly profile itself as an independent discipline. This essay concentrates on the influence on this development of sociology and its, typically Dutch, predecessor sociography. The starting point here is the promotion tree of the founder of the Dutch sociology Sebald Steinmetz. Through him various lines (via his doctorates Nicolaas ter Veen and Jakob Kruijt) go to modern Public Administration. This essay tells the story of the influence of sociography and sociology on the development of the administrative sciences and modern Public Administration in six acts, in which two persons from the promotion tree are discussed (via Sjoerd Groenman, who is promoted by Nicolaas ter Veen there are two different lines again). The line via Jakob Kruijt contains Aris van Braam (he wrote in 1957 what is considered the first Dutch empirical study in Public Administration) and Jos Raadschelders. The first line via Sjoerd Groenman contains Henk Brasz (the first full-time professor in Public Administration in the Netherlands), Fred Fleurke and Ko de Ridder. The second line via Sjoerd Groenman contains Joop Ellemers, Geert Braam (professor at the first regular Dutch Public Administration programme in Twente) and Wim Derksen. These acts are framed with short intermezzos about the other sociological key figures who played an important role in the story of sociography, sociology and Public Administration. In conclusion, the author of this essay discusses the continuing relevance of sociology for modern Public Administration.

In Nederland komen tal van wicked problems (WP) voor. Ze worden ook wel aangeduid als weerbarstige problemen. Kenmerkend is het unieke karakter, het feit dat kennis over probleemaspecten beperkt is en verschillende perspectieven op een probleem en meerdere waardenoriëntaties een rol spelen. Naast cognitieve en normatieve complexiteit bestaat er ook nog sociale complexiteit. WPs zoals voetbalvandalisme spelen zich af in een beleidsnetwerk met tal van elkaar afhankelijke actoren. Kennis, preferenties, handelingsvermogen en middelen blijken gespreid en niet in de hand van één actor. Om toch tot een bevredigend resultaat te komen moeten de actoren gezamenlijk optrekken. Daarom is interorganisationeel netwerkmanagement en deliberatie met burgers een wenselijke responsestrategie. Bij WPs zijn actoren dus tot elkaar veroordeeld. Centralistische besluitvorming werkt niet. Een centrale actor kan overigens nog wel een rol vervullen als initiator, facilitator of regisseur van overleg en discussie; als arbiter bij botsende perspectieven; of beslisser na netwerkberaad en inbreng van burgers.

Een lerende evaluatie combineert leren en verantwoorden. Deze methode past bij complexe beleidsopgaven, waar meerdere actoren en overheden bij betrokken zijn die een behoefte hebben om te leren van elkaars ervaringen. In de evaluatie van het Natuurpact is deze methode op nationale schaal voor het gedecentraliseerde natuurbeleid toegepast. Uit deze casus blijkt dat een succesvolle toepassing vraagt om een voortdurende en zorgvuldige aansluiting van het onderzoek op de beleidspraktijk. Dit is nodig voor het betrekken van en interactie met stakeholders, het combineren van leren en verantwoorden, de wetenschappelijke onafhankelijkheid, het bestuurlijk mandaat en de beheersbaarheid van het onderzoek.

Het ontwikkelen van succesvol overheidsbeleid vraagt om een systematische benutting van kennis uit wetenschap én praktijk. In dit artikel wordt beschreven op welke manier die systematische benutting kan worden bewerkstelligd. Als de afstand tussen kennis en beleidsontwikkeling kan worden verkleind en tevens wordt gekozen voor een proces van open beleidsontwikkeling, zal de benutting van kennis tijdens het beleidsproces beter verlopen. Daarbij dienen tegelijkertijd diverse voorwaarden te worden gesteld aan een adequate programmering van het corresponderende beleidsonderzoek.

Wicked problems ask for new, smart forms of governance beyond a singular focus on hierarchy, market or community. Based on the case studies presented in the individual articles of this special issue, this concluding article describes what smart governance could entail and discusses its strengths and weaknesses, both as a concept and a practical tool.

Policy decisions taken now can determine the room for manoeuvre of future generations for a very long term. Politicians and civil servants often only seem to be interested in short-term implications, however. A major focus on the short term does not provide sufficient stability and impetus for long-term structural measures. This is primarily an institutional issue. In this article, the authors discuss how to instil a long-term focus in day-to-day processes of policymaking.

De preoccupatie met evidence based beleid miskent de exploratieve waarde van wetenschap voor beleid. In veel en de meest uitdagende beleidsprocessen, zoals de grote transities aangaande energie, zorg, landbouw en circulaire economie, moet de overheid beslissingen in onzekerheid nemen. Het beleidsonderzoek zou zich minder moeten bezighouden met wat zeker is en meer met wijs handelen in onzekerheid. Wetenschappers kunnen de besluitruimte voor politici duiden, en kunnen aangeven wat mogelijk effectieve maatregelen zijn. Dit bescheiden, voorzichtig aftasten van de werkelijkheid vereist arena’s waar wetenschap, bestuurders en volksvertegenwoordigers in gesprek zijn over de toepassing van wetenschappelijke inzichten in de praktijk van het hier en nu. Op diverse plekken in Nederland wordt geëxperimenteerd met dergelijke arena’s.

Societal and technological developments (such as the digital and energy revolutions) move faster than existing institutions can keep up with. The developments may lead to a metaphorical institutional void, which brings questions about the nature of the void, the changing rules, practices and responsibilities, and about the strategies to deal with the void. The concept has been around for a while but (again) seems relevant to understand current socio-technological innovations and challenges, that also allow us to further conceptualise the institutional void. In this introduction to the issue, we discuss the concept of an institutional void and explore how it is used in various domains of study, including public administration. We argue for how the concept is relevant today and therein also introduce the topics that are discussed in this special issue.

Scientific knowledge plays a pivotal yet problematic role in identifying, assessing and evaluating climate impacts, and hence in their governance. This raises questions of how knowledge for adaptation policy is made. This article studies the production of authoritative and meaningful knowledge claims in the Delta Committee, regional water management and urban warming. It is argued that the conventional supply-and-demand conceptualisation with its notion of ‘knowledge transfer’ has fundamental flaws. This study shows how the wicked issue of climate change is tamed and made tractable in climate adaptation. In these processes knowledge of climate change transforms. This article presents a conceptual apparatus to study transformation. Transformation has a Janus face. While transformation brings climate change in conversation with localised meaning to create concrete adaptation responses, it also closes down and becomes blind to particular climate risks. Transformations are affected by the goals and institutions of policy fields. To overcome problems of blindness and cognitive path dependencies, more institutional change is necessary than the current piggyback approach of mainstreaming and knowledge co-creation entails.

Reflection and debate initiates academically inspired discussions on issues that are on the current policy agenda.

Social media monitoring is a topical and relevant phenomenon. It enables civil servants and politicians to gauge the sentiments voiced on social media, on the basis of which they are in a better position to take into account the wishes and needs of citizens. Social media monitoring is primarily used for rational and strategic purposes. In terms of democratic legitimacy, it may enhance the quality of the processes on the output side of the political system, i.e. authorities can be more responsive and can fine tune public policies. There also threats for the relation between citizens and government. When citizens communicate on networks they perceive as private, social media monitoring can be seen as an intrusion into their private sphere. This not only concerns individual privacy but also an interpersonal private sphere in terms of the right that people have to define a domain within which they can exchange experiences with peers. Transparency and accountability are therefore important conditions for the application of this instrument.

Government cybersecurity requires action from many public and private actors. Both collective knowledge and collective priority are needed to ensure cybersecurity at a government level. This makes collective learning essential. There is a system of arrangements that includes all kinds of governmental organisations and private parties. How can learning be stimulated in this system? And what is the need for steering here? This article provides answers to these questions, based on the contributions in this special issue. We conclude that both central control and self-regulation are essential to cybersecurity, even if they are in conflict. We coin the term ‘complimentary self-regulation’. We also conclude that many arrangements have been developed or are under development, however, it is difficult to institutionalise the coherence between these initiatives. There is a long road ahead in terms of gaining a collective understanding. Cybersecurity and its organisation will probably not vanish from the administrative agenda any time soon.

For this article, part of a series on the future of the discipline in the Netherlands, the author has talked to ten newly appointed professors in the field of public administration. We discussed their background, how they see their role and position within university and society and how they view recent developments in our field of study and our discipline. The young professors share their view on how our discipline could and should develop and what their role will be in this process.

Science and practice are largely agreed on the importance of interorganizational cooperation in the approach of tackling complex societal problems. Organization transcending innovation through this type of cooperation however appears to be complicated. Based on an analysis of the literature about partnerships, the authors distinguish three challenges: coping with the tension between old and new accountability structures, building good working relationships and developing capabilities for problem-oriented working. Starting from these insights they designed action research into problem-oriented partnerships in the safety domain (safety chain). Their main question was: what is the most important obstacle for innovation through problem-oriented interorganizational cooperation? Over a period of nine months, they watched eight teams of professionals from different organizations. Their task was to develop and implement innovative approaches to tackle persistent organized crime. Although all three challenges identified in the literature indeed played a prominent role, problem diagnosis and problem definition appeared to be the main obstacle for the teams. In this article the authors describe the action research and explore, on the basis of the results and the literature, how partnerships could cope in practice with the challenge of problem definition and problem analysis. They conclude the article with suggestions for the design of a follow-up round of the action research.