Page 365 - Invited Paper Session (IPS) - Volume 1

IPS169 Markku L.
that is, what the institution is – what it represents for the individual in question
(Kestilä-Kekkonen & Söderlund 2016, 141). For example, one can trust
specifically the present government coalition or more generally the national
government. To evoke trust, an institution needs to be perceived as
competent, sincere, transparent, reliable in keeping its promises, and
demonstrate ability to deal with mistrust and to avoid mismanagement or
entanglement in political scandals (Holmberg & Weibull 2017, 39; Laurian
2009, 383-384; Tuler & Kasperson 2010). The independence of statistical
offices is therefore vital and seen as an essential part of the very identity of
statistical institutions. Earlier research has shown that in situations of
longstanding institutional mistrust, attempts at trust-building, e.g. via
participation and openness, can initially undermine trust (Gouldson et al. 2007;
Laurian 2009). The above statements from Davies and Desrosières express
concern for loss of trust in statistical offices and statistics as an institution.
The distinction between specific and diffuse support to an institution helps
us to understand the foundations of institutional trust, yet if indeed specific
support stems from evaluation of the performance of an institution, the
question arises: on which grounds do citizens judge the performance of a
statistics institution? Is scientific robustness enough or does trustworthiness
also build upon features such as the sensitivity of the institution to the plurality
of contexts, “both in terms of the relevant social spaces and values” (Rafols
2019)? Statistical institutions are faced by a dilemma between independence
and scientific robustness on one hand, and relevance on the other. To be
‘argumentatively effective’ (Desrosières 2015), indicators need to be seen as
scientifically reliable, but when this implies foregoing considerations relating
to the needs and values of the potential users, indicators risk being seen by
the public and stakeholders as increasingly obsolete, irrelevant, and even
misleading. Ignoring the feedback and the performative effects of indicators
risks hiding underlying value commitments, uncertainties, and the regimes of
power’ underpinning the indicators behind seemingly neutral discussions
concerning the appropriate methodological choices (Merry 2016). Jasanoff &
Simmet (2017, 754) draw attention to the “wider realities in which contested
public facts are embedded”, and argue that any “truth needs to be performed
and accepted” (ibid., 756). These insights help to put into perspective the
“truths” conveyed by composite indicators and the alleged downsides from
the declining trust in indicators and statistics. According to this line of thinking,
what is at stake is the ability and moral authority of the state to produce
‘serviceable truths’, that is, “robust statements about the condition of the
world, with adequate buy-in from both science and society” (Jasanoff &
Simmet 2017, 756).
The ideological dimension of trust is relatively seldom addressed in social
science research, yet it is perhaps the most informative from the point of view

354 | I S I W S C 2 0 1 9