Algorithms are starting to play an increasingly prominent role in government organizations. The argument is that algorithms can make more objective and efficient decisions than humans. At the same time, recent scandals have highlighted that there are still many problems connected to algorithms in the public sector. There is an increasing emphasis on ethical requirements for algorithms and we aim to connect these requirements to insights from public administration on the use of technologies in the public sector. We stress the need for responsible algorithmization – responsible organizational practices around the use of algorithms – and argue that this is needed to maintain the trust of citizens. We present two key components of responsible algorithmization – value-sensitivity and transparency – and show how these components connect to algorithmization and can contribute to citizen trust. We end the article with an agenda for research into the relation between responsible algorithmization and trust.

Algorithms applied in public administration are often criticized for lack of transparency. Lawmakers and citizens alike expect that automated decisions based on algorithmic recommendations to be explainable. The focus of this article is the organizational context behind the idea of transparent algorithms. Transparency is portrayed as one of numerous values that are at play when algorithms are applied in public administration. The article shows that applying algorithms may lead to conflicts between these values. Such conflicts often result in trade-off decisions. Looking from the organizational perspective, we describe how such trade-offs can be made both explicitly and implicitly. The article thus shows the complexity of algorithmic trade-offs. As a result of this complexity, we not only call for more transparency about algorithms, but also more transparency about trade-offs that take place in public administration. Finally, we present a research agenda focused on studying the organization of trade-offs.

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.

Hoeveel nieuwe methoden en technieken van sociaalwetenschappelijk onderzoek zijn er sinds de jaren zestig van de vorige eeuw bij gekomen? Moderne ICT heeft het allemaal wel sneller en makkelijker gemaakt, maar in wezen gaat het nog steeds om dezelfde methoden, wellicht op een enkele uitzondering na. Maar misschien levert deze column reacties op die dit tegenspreken.

Nederland staat voor forse en complexe beleidsopgaven. Deze opgaven vragen om een bijzondere beleidsaanpak met een aansluitende wijze van beleidsevaluatie – namelijk één die leren ondersteunt om iteratief de kwaliteit van het beleid te verbeteren en de weg naar de beleidsambities te vinden. Beleidsonderzoekers en beleidsbetrokkenen werken in lerende evaluaties samen om kennis te produceren voor het gelijktijdig verantwoorden en leren van beleid. Verondersteld wordt dat de kwaliteit en bruikbaarheid van de geproduceerde kennis met deze benadering groter zijn dan bij reguliere, op verantwoording georiënteerde, evaluatiemethoden. Als gevolg daarvan zou lerend evalueren meer impact hebben op beleid voor complexe opgaven. In dit artikel wordt aandacht besteed aan de waarde van lerend evalueren vanuit het perspectief van beleidsbetrokkenen en beleidsonderzoekers van de lerende evaluatie van het Natuurpact (2014-2017), uitgevoerd door het PBL en de WUR. Geconcludeerd wordt dat lerend evalueren de kwaliteit, bruikbaarheid en impact (minder aantoonbaar) van de geproduceerde kennis vergroot, maar onder specifieke voorwaarden: namelijk wanneer onderzoekers erin slagen om leren en verantwoorden, met de bijbehorende rollen en kwaliteitsstandaarden, te benaderen als wederzijds versterkend in plaats van tegenstrijdig. Onderzoekers hebben voelsprieten nodig voor de wisselwerking tussen het proces van kennisproductie en de politiek-bestuurlijke context waarin deze kennis wordt gebruikt. Zowel in de beleids- als onderzoekspraktijk is ruimte nodig voor een verbrede kijk op de functie van beleidsevaluatie om lerend evalueren toe te kunnen passen.

De focus op effectiviteit en vooral ook efficiëntie in beleid leidt wellicht tot beleidsuitvoering die efficiënt is voor de overheid, maar of dat ook zo is voor andere stakeholders, wordt misschien uit het oog verloren. Dat laat de huidige coronacrisis zien, maar blijkt ook uit andere voorbeelden. Voor beleidsonderzoek is het relevant om beleid te evalueren binnen de bredere maatschappelijke context en met de vraag voor wie beleid efficiënt is.

In this article, we investigate the moderating role of political sophistication on the vote for populist parties in Belgium. Building on the literature about the diverse determinants of populist party support, we investigate whether issue considerations and populism-related motivations play a bigger role in the electoral calculus of politically sophisticated voters.
Using data from the 2019 general elections in Belgium, we focus on the cases of Vlaams Belang (VB) and Parti du Travail de Belgique- Partij van de Arbeid (PTB-PVDA). We find evidence suggesting that political sophistication enhances the impact of populism-related motivations on populist party support, although the effects are contingent on the party. Moreover, we show that, for issue considerations, the moderation effect only comes into play for VB voters: the impact of anti-immigrant considerations is greater at increasing levels of political sophistication.

Recent research demonstrates that political parties in western Europe are generally structured along one dimension – and often take more or less similar ideological positions on the economic and cultural dimension – whereas the policy preferences of voters are structured two dimensionally; a considerable part of the electorate combines left-wing stances on one dimension with right-wing stances on the other. These ideologically ‘unserved’ voters are the main focus of this study. Using data from a large-scale survey in Flanders and Wallonia, we demonstrate how the salience of the two dimensions explains whether these unserved voters ultimately end up voting for a right-wing or a left-wing party. Specifically, we show that these voters elect a party that is ideologically closest on the dimension that they deem most important at that time. To summarise, the findings of this study confirm that salience is a key driver of electoral choice, especially for cross-pressured voters.

This study investigates how protest attitudes and ideological considerations affected the 2019 election results in Belgium, and particularly the vote for the radical right-wing populist party Vlaams Belang (VB) and for the radical left-wing populist party Partij van de Arbeid-Parti du Travail de Belgique (PVDA-PTB). Our results confirm that both protest attitudes and ideological considerations play a role to distinguish radical populist voters from mainstream party voters in general. However, when opposed to their second-best choice, we show that particularly protest attitudes matter. Moreover, in comparing radical right- and left-wing populist voters, the article disentangles the respective weight of these drivers on the two ends of the political spectrum. Being able to portray itself as an alternative to mainstream can give these parties an edge among a certain category of voters, albeit this position is also difficult to hold in the long run.

Recently, studies have burgeoned on the link between populism and demands for democratic reforms. In particular, scholars have been debating the link between populist citizens or voters and support for referendums. In this article, we examine voters of populist parties (Vlaams Belang (VB) and Parti du Travail de Belgique-Partij van de Arbeid (PTB-PVDA)) in Belgium in 2019 and we look at their attitudes towards various types of democratic reforms. We find that voters of populist parties differ from the non-populist electorate in their support for different kinds of reforms of representative democracy. Voters of VB and PTB-PVDA have in common stronger demands for limiting politicians’ prerogatives, for introducing binding referendums and for participatory budgeting. While Vlaams Belang voters are not significantly different from the non-populist electorate on advisory referendums, citizens’ forums or technocratic reform, PVDA-PTB voters seem more enthusiastic.

This article digs into the relationship between voters’ political resentment and their electoral choice in 2019 by focusing on the respondents’ emotions towards politics. Using the RepResent 2019 voter survey, eight emotions are analysed in their relation to voting behaviour: four negative (anger, bitterness, worry and fear) and four positive (hope, relief, joy and satisfaction). We confirm that voters’ emotional register is at least two-dimensional, with one positive and one negative dimension, opening the possibility for different combinations of emotions towards politics. We also find different emotional patterns across party choices, and more crucially, we uncover a significant effect of emotions (especially negative ones) on vote choice, even when controlling for other determinants. Finally, we look at the effect of election results on emotions and we observe a potential winner vs. loser effect with distinctive dynamics in Flanders and in Wallonia.

Adriaan Koelma fits in with the list of legal scholars who helped to shape the early history of the (local) administrative sciences in the Netherlands, which was dominated by a legal approach to local administration. In that respect, he was not only a follower of the first Dutch public administration scholar, Gerrit van Poelje, but also his successor. He held the chair in Public Administration in Rotterdam, which Van Poelje vacated in 1933, first as a lecturer and later as a professor (from 1946 onwards). Nowadays, Koelma is mainly remembered for the state commission named after him: he (in vain) advocated the introduction of districts (next to municipalities). He was chairman of this state commission that was installed by Minister Beel on 19 December 1946. He fulfilled his scientific activities in addition to a career in the Dutch civil service. Koelma was a typical ‘self-made man’ who worked his way up from junior employee at the municipal clerk’s office of Dordrecht to municipal clerk and, if only briefly, mayor of Alkmaar. His experiences in the Second World War had a great influence on his later life. Due to a war-related illness, he had to give up the chairmanship of the Koelma Commission in 1947 and in 1948 his professorship and role as mayor of Alkmaar. This war also gave him insight into the pernicious influence of Nazi ideology on governance theory and governance practice. He could not have suspected how hard the German occupier would put the Dutch administration and its servants to the test during his public lesson of 1934, because at that time the Nazi regime in Germany had not yet shown its true nature at the local level of government.

This paper analyzes the impact of linguistic characteristics of accountability reports on public sector organizations. It does so by analysing hundreds of accountability reports by four public sector bodies using the linguistic tool LIWC. The research question is: what linguistic characteristics of accountability reports are related to a bigger impact on the evaluated organization? The impact of three strategic choices is assessed. First of all, the impact of strategic positioning. Authors of texts can maintain a position of power in the choice of language (high clout) and speak top down to the recipient or they can take a more egalitarian, face to face, position. Secondly, authors can choose to use many complex linguistic phrasings, with causal reasoning for instance, or they can opt for simpler texts. Finally, the text can be littered with emotional, positive and negative, wordings or can be set in a neutral tone. Our analyses suggest that more emotional accountability reports are consistently related to a better reception and seem to have more impact. This has important consequences both theoretically and practically, which are discussed in the paper.