Expert accountability: What does it mean, why is it challenging—and is it what we need?

When the Norwegian Parliament opened southeastern Barents Sea for petroleum activity in 2013, it did so based on an impact assessment report prepared by Ministry of Petroleum and Energy (Norwegian Energy, 2012–2013). In 2016 first licenses to drill this area were awarded, two environmental organizations sued state awarding them. To prepare court case commissioned economists with expertise resource economics review these quickly discovered major errors. For one, not discount expected costs future incomes—a standard procedure such reports. Discounting a 4% real interest rate would have reduced estimated net income from opening approximately 38%. This make loss (Greaker & Rosendahl, 2017a, 2017b). The also gross scenario low findings be twice what underlying numbers showed These mistakes together optimistic estimates employment oil price made seem like there was no economic risk associated opening, may contributed Parliament's decision open spite resistance both environmentalists fishing industry 2017a; NTB, 2017; Tomassen, 2017). A remedy when are political life is find someone hold account. But experts involved governance processes they rarely held account their mistakes. sometimes seen as legitimacy problem expert-reliant governance, warnings given about handing over more power “unaccountable” experts. So, solution accountable? popularity “accountability” all kinds problems has led inflation fragmentation meanings attributed term, uncritical uses accountability measures. Still, article argues that remains crucial addressing brought up increasing expert dependency. Yet, instead proposing new type regime applicable contexts take part, takes step back asks: What exactly does mean “accountable”? And distinct challenges accountability? There several studies relation particular institutions, public agencies (e.g., Busuioc, 2013; Schillemans et al, 2021), central banks Heldt Herzog, judicial Contini Mohr, 2007), parliaments Crum, Eriksen Katsaitis, 2020). typically difficult compare because define different ways or leave definitions implicit underspecified. They rely taxonomies subtypes, different, unclear, assumptions account-holding good for. Furthermore, focus types roles differ will relevant can serve appropriate balanced set relations. We believe need general systematic approach accountability, contrast some previous attempts Holst Molander, Moore, 2017), our contribution connects detail literature vast tease out finer complexities accurate manner. aim provide conceptual analytical tools facilitate further comparative evaluative research who part processes, many do. exemplify distinctions , but illustrate challenges, tradeoffs, holding clarifies use term “expert,” looks at wide range context governance. second works narrow differentiated conception helps us see how subtypes draw opposite directions. third demonstrates allows map relations precise way. Part four criteria saying is, not, sufficiently accountable. five analyzes Two special identified—the epistemic asymmetry, biases influence other forum drift eyes hands. most proper response vary institutional goal account-holding, horizontal forms accountability—peer professional reputational particular—are key settings. However, bias, insularity, prevent functioning properly. moreover tensions tradeoffs between which designed ensure democratic control abuse power. Finally, emphasized alone cannot address normative Norway areas lies Parliament. required law before parliamentary made. must sent consultation, sum study inputs consultation white paper should presented thus in-house ministries agencies, addition contracted consultancy. Political parties, media, stakeholders—from local communities NGOs industry—also consulted topics relating opening. short, various sitting capacities, sides table, could exert significant amount far regard. decisionmaking standardly involves experts, well plethora bodies, banks, advisory committees courts. Commentators characterized “expertization” society policy making (Turner, 2003), “rise unelected”—the development branch government those knowledge cuts across traditional separation powers (Vibert, conventional civil society, private (Moore 2021). discuss extent answer we least working definition expertise. line Alvin I. Goldman's influential approach, considerable knowledge, than others, domain, able skillfully novel (2001(2011), p. 114). makes scientist academic important category. often terms “experts” “scientists” used interchangeably, drawn disciplinary background (economists, lawyers, engineers, etc.). identified capacities (Grundmann, 2016), sources scientific training, especially experiences becoming knowledgeable something through practical engagement certain issues time (Collins Evans, 2007). Experienced servants judges, example, possess kind practically gained skills; same apply groups substantive sector shall refer them here, only level competence area. Their called (Gundersen, 2018). Knowledgeable people become then, according definition, considered relevant, guidance asked within category socially recognized knowers, here included making. Roughly speaking, adds class “unelected” knowers arguably increasingly present day politics (see Habermas, 1992 early diagnosis “expertocracy”). Accordingly, “governance experts” mind work service, courts, consultancy, organizations, associations, institutions universities; information providers, advisors, agenda setters, even decisionmakers—as judge decides case, bank board rates. ask, varied, distinctive, Accountability sense denote persons organizations’ inner responsibility moral values norms, willingness responsible, transparent, responsive, engage dialog justification (Behn, 2001, 3–6; Bovens, 2010, 946; Dubnick, 2007; Mulgan, 2000, 555). allude arrangements powers, constitutionalism, (Mulgan, 563). Such broad approaches tend however miss distinctive starting point therefore rather been core sense, meaning “answerability performance” (Romzek, 2015, 28), precisely, “the process being 'to account’ authority one's actions” (Bovens, 2007, 450). understood refers then specific social relation11 Others speak “process” “a mechanism,” “set mechanisms control.” actor22 Also referred “agent,” “accounter,” “agent institution account.”, forum33 “account-holder” “accountee,” “master,” “principal,” “authority.”, where actor under obligation, formally informally, explain justify his her conduct, request explanation actor's actions subsequently sanction trying delineate means, observation mind, despite internal diversity, generally scope discretion virtue comparatively agent autonomy (Lindberg, 209). if aspects performance, foremost mechanism aimed controlling constraining scope. furthermore ex post facto, accountable past (and future) performance. trusted domain. acts using discretionary powers. requests regarding expert's actions. provides (or not). follows questions answers. passes its judgment action provided information. sanctions66 Sanctions informal reprimanded, getting bonus renewed contract. Importantly, stand relationship, things listed do place. happen, aware this, presumably shapes expectations behavior.77 Hall al. (2017); Tetlock (1992). Narrowing down way characterizations discussion are, any taxonomy sub-types grasp accountholder accountee. suggest, however, useful start Staffan Lindberg (2013) supplement Bovens (2014). Lindberg's advantage placing axes simultaneously, while differentiating vertical importance discussing accountability. Specifically, focuses three characteristics relationship88 Note “agent” “actor,” “forum.”: First, external organization? Second, forum's degree weak orstrong? third, upward form researchers project leader misusing findings? Is downward top bureaucrat holds lower-level advisor analyst account? Or one members peers? Combining dimensions, sub-specifications, yields 12 (Figure 1). Bureaucratic example specified hierarchical vertical, organization, high control. Peer hand, horizontal, internal, marked dimensions missing taxonomy. particular, Lindbergh's specify aspect for, nor standards. analyze aspects, Schillemans, Robert Goodin's supplement. questions: (1) Who (corporate, hierarchical, collective, individual accountability). (2) nature whom rendered? (political, bureaucratic/managerial, legal, professional, (3) conduct he for? (financial, procedural, product (4) standards forum? (bureaucratic, accountability) al., 2014). whether source strength direction preferrable taxonomy; yet brings valuable additional nuances dimensions. instance, alerts fact always single assessing work, assess (cf. Willems Van Dooren, 2012). forum, say Minister, seek adherence rules procedures (bureaucratic accountability), legal norms (legal demands preferences (political Below supplementing needed. It said unaccountable, closer look multiple relationships. Take economist whose includes writing summarizes reports consequences Sea. Obviously, bureaucratic administrative superiors Ministry, relationship upwards primarily abiding (procedural money spent, expenses travel allowances budgets (financial auditing), outputs outcomes (product legally every job regulated laws, summoned harassment, criminal corruption. Also, audited internally fiscally accountable, both, research. economy direct voters. theory committee, unless done punished law, Minister ultimately ministry (a relation). principal strong content report. While formal politically leadership likely outcome experts’ considerations. Formally, interfere economist's judgments, unlikely principals heightened scrutiny vis-à-vis go against principal's wishes. If so, stands relation. peer organization. Peers assesses addition, comes peers outside apparatus, professors academia. Civil media try societal relationship.99 Lindbergh “diagonal accountability.” Activist ask existing Greenpeace Nature Youth tried responsible report, including it, pursuing reputational, suing decision, hiring help document mistakes, simultaneously bringing attention attempt capture incomplete suffices embark mapping relationships of. mappings sort directly worry connected policy. Yes, de facto relations, Or, put differently, “accountability deficit”? Let once concretize going case. Some criticized noticed already increased carbon emissions. Other omissions organizations. had failed incomes, (with small findings) double expect. latter mistake traced Excel-typo Directorate 2017b; Taraldsen, methodological choices strengthened impression very profitable, instance projection remain 120 USD per barrel. By awarded 45 USD. found into 2013b). result risks underplayed, influenced open. situation, wrote Impact Assessment Excel-typo, democratic, popular, corruption, power, arbitrary exercise enhance standards, learning, contribute effectiveness achieving stated goals. preparing reflect relied on: main ensuring Prevent corruption abuse, secure quality, professionalism induce learning cycles combination goals values? general, conducive securing rule possibility democratically elected forums, parliaments, actors mix audit, bureaucratic, contributing preventing learning. Thus, natural priorities accountabilities. wants achieve goals, see, accountabilities easily come conflict create overloads. disagree account: became known, calls heads roll: Director (an organization), others demanded hearing Standing Committee Scrutiny Constitutional Affairs (hierarchical One practice reviewed Office General Auditor audit relation) (NTB, voiced activists politicians critical activity. critics saw skewed possible deliberate fraud, purpose continued production Norway, framed report's problem, arguing inadequate basis actors, Directorate, focused Excel typo “human mistake,” impact1010 mistaken place 25).—a characterization measures improve quality reports.1111 publicly known authors changes routines As seen, voters, give performance authorized minister committee. indirect control, cases minister's proxies, compromise independence, undermine qualities legitimizing role Including “filter” ensures “truth-sensitivity” decisions, reflecting primary involving (Christiano, 2012): defensible consult delegate decisions area, extra enlighten processes. On assumption, regimes (Holst requires emphasis shifts toward accountability—such review—may resisted, ministry, reduce outcomes. agreement right balance aims, occur them, stems expert/nonexpert relationship. other, well-known features behavior. Besides, intensify expert, drift, eyes, asymmetry characterizes forum–actor principal–agent (which may, principal) consists nonexperts profound. With “expert” understand harder effective develop. asymmetries forums variable, seldom significant. hard d