ArticlePDF Available

The effects of judicial transparency on public trust: Evidence from a field experiment

Abstract

Trust in judges is needed for voluntary acceptance of judicial decisions, and judicial transparency is thought to strengthen trust. It exposes the public to symbols that embrace a ‘myth of legality’ which is expected to have a positive effect on trust. We assess a specific understanding of transparency and trust by looking at the moderating effect of knowledge and predisposition to trust. We report on a field experiment which investigates the effect of a Dutch television series on trust. Findings show that judicial transparency indeed has a positive effect on trust. Moreover, our analysis demonstrates that it has the strongest effects on individuals with medium prior knowledge about the judiciary. However, higher predisposition to trust mitigated the effect of transparency, indicating a ceiling effect. This sustains the idea that the unique traits of visual judicial transparency expose typical judicial symbols that imply that impartiality which increases trust in judges.
1
Please refer to as:
Grimmelikhuijsen, S.G., and Klijn, A. (2015). The effect of judicial transparency on public
trust: evidence from a field experiment. Public Administration (in press). Article first
published online: 14 JAN 2015 | DOI: 10.1111/padm.12149
2
The effects of judicial transparency on public trust: evidence from a field experiment
Authors:
dr. Stephan Grimmelikhuijsen (corresponding author)
Assistant professor
Utrecht University School Governance, Utrecht University
s.g.grimmelikhuijsen@uu.nl
+31302531433
dr. Albert Klijn
Scientific Advisor
Training and Study Centre for the Judiciary
Abstract
Trust in judges is needed for voluntary acceptance of judicial decisions; judicial transparency
is thought to strengthen trust. Judiciary transparency exposes the public to symbols that
embrace a ‘myth of legality’ which is expected to have a positive effect on trust. Furthermore
we assess a specific understanding of transparency and trust, by looking at the moderating
effect of knowledge and predisposition to trust. We report on a field experiment which
investigates the effect of a Dutch television series on trust. Findings show that this indeed has
a positive effect on trust. Moreover, our analyses demonstrate that judicial transparency has
strongest effects on individuals with moderate prior knowledge about the judiciary. However,
higher predisposition to trust mitigated the effect of transparency, indicating a ceiling effect.
This sustains the idea that the unique traits of visual- judicial transparency expose typical
judicial symbols that imply impartiality and non-political decision-making increases trust in
judges.
3
1. INTRODUCTION
Legitimacy of judges is acknowledged to be crucial for voluntary acceptance of judicial
decisions and maintaining social order (Tyler 2006) and trust is acentral component of this.
Transparency is seen as an important means of strengthening citizen trust (Gibson and
Caldeira 2009; Worthy 2010; Meijer, Curtin and Hillebrandt 2012). However, there is intense
debate in the academic literature whether transparency really contributes to trust or whether it
deteriorates citizen trust in institutions (Meijer 2009). Indeed, an emerging body of empirical
work has offered some evidence that transparency can have negative or at best subdued
effects on citizen trust (Worthy 2010; De Fine Licht 2011; Grimmelikhuijsen 2012), which is
supposedly due to the exposure of the inherent processes of political decision-making that
people dislike, for instance, bargaining, brokering and deal-making in political institutions
(Hibbing and Theiss-Morse 1995; 2002; Gibson and Caldeira 2009; Grimmelikhuijsen 2010).
Most work in this area exclusively focused on the transparency of political institutions and
political decision-making, while at the same time researchers have acknowledge the context-
dependency of the effects of transparency on trust and legitimacy (Grimmelikhuijsen et al.
2013; Meijer 2013; De Fine Licht 2014). Therefore, we want to further our understanding by
investigating the effects of transparency on trust in a crucial and contextually very different
institution: the judiciary. The judiciary works differently from political institutions on a
crucial point, as it is expected to make impartial, independent and non-political decisions.
Transparency exposes people to symbols that embrace this ‘myth of legality’ (Gibson and
Caldeira 2009). These symbols are of particular importance as citizens often lack in-depth
knowledge about the judiciary and therefore make symbolic evaluations (Tyler 1998, 849;
Van de Walle 2009).
Typical symbols such as a gavel, a courtroom or a gown, convey the message that courts are
‘special’ and different from other institutions. The message that these symbols put across is
that not bargaining or compromise, but impartial and non-political decision-making take
place within courts. For instance, Hibbing and Theiss-Morse (1995) found that more
awareness of the Congress led to less support, whereas increased awareness of the Supreme
Court led to more support for this institution. Based on prior work by Gibson and Caldeira
(2009) on media exposure of the Supreme Court, we expect that exposure to these symbols
triggers a positivity bias. This bias entails that citizens have certain beliefs about the judiciary
that are generally positive. From early life people are taught that judges are impartial and
4
different. When exposed to typical judicial symbols, these latent positive attitudes of citizens
are ‘woken up’ by the symbols that remind citizens of the impartiality of the judiciary
(Gibson, Caldeira, and Baird 1998; Gibson and Caldeira 2009).
This study seeks to contribute to our understanding of transparency in two ways. First, we
conducted a general analysis to assess whether judicial transparency indeed has a positive
effect on trust in judges. The second goal of this article is to test a more specific
understanding of the effect of transparency. Based on theoretical notions we test how
attitudinal changes are affected by pre-existing attitudes and knowledge (Petty and Caccioppi
1984; Bohner 2001; Grimmelikhuijsen and Meijer 2014). Individuals process information
differently if they have certain beliefs or attitudes concerning the relevant topic. We
hypothesize that citizens with a great deal of knowledge about judges are less receptive to
change their trust in this institution. Furthermore, we expect that transparency confirms
existing predispositions, i.e., positively predisposed citizens become even more trusting,
negatively predisposed individuals become less trusting.
The central question we deal with in our article reads as follows: To what extent does
transparency affect trust in judges and how is this effect related to individual predispositions
to trust and knowledge?
In order to assess the question we rely on a field experiment that was carried out in
cooperation with a Dutch District Court. The study we report on consisted of a television
show about a District court in the Netherlands that allowed a public television to cover its
daily work. Day-to-day cases were taped in combination with footage of commentary,
deliberations and personal interviews with judges were broadcasted on eight evening shows.
In this sense the series was different from regular mass media coverage mostly focusing on
high profile cases.
2. TRANSPARENCY AND TRUST IN JUDGES
Many definitions and operationalizations of trust are being used by researchers in the social
sciences. However, across disciplines it is generally acknowledged that two conditions are
important with regard to trust: (1) a degree of ‘risk’ and (2) ‘interdependence’ (Zucker 1986;
Luhmann 1979). Risk and interdependence are crucial because trust is a threefold
relationship, in which A trust B to do X (Hardin 1993). In other words, A depends on B to do
5
a certain thing which is in his/her interest. This interdependence yields a risk as citizens are
uncertain as to whether an institution (e.g. judiciary) actually carries out its tasks the way
they are entrusted to. Risk is invoked because government institutions exert a certain degree
of power over citizens, which can be used properly or not. Without uncertainty about what an
organization does with the trust placed in them there would be no need for trust (Lewis and
Weigert 1985). On the other hand, some knowledge is thought to be needed in the process of
building trust (Hardin 2002).
Based on these conditions Rousseau et al. (1998) created a much-used multidisciplinary
definition of trust: trust is ‘a psychological state comprising the intention to accept
vulnerability based upon positive expectations of the intentions or behavior of another.’
(Rousseau et al. 1998, p. 395). With regard to judges, these expectations are to a large extent
connected with the notion of ‘procedural fairness. The centrality of procedural fairness as a
driver for trust and legitimacy of courts has been recognized in the psychological literature.
Psychologists have provided evidence that people base their trust in judges not so much on
the outcomes they produce, but on whether they feel individuals are treated fairly (Tyler et al.
1997; Tyler 2001). Thibaut and Walker (1975) originally showed the importance of process-
based evaluation in the willingness to accept court decisions. Making the judiciary more
transparent contributes to making these processes more visible and assuming that (both)
parties involved generally receive a fair process this will increase trust in judges.
The second main concept of this study, transparency, is often related to the extent to which an
entity reveals relevant information about its own decision processes, procedures, and
performance (Heald 2003; Gerring and Thacker 2004; Welch, Hinnant, and Moon 2005;
Curtin and Meijer 2006; Grimmelikhuijsen 2012). Building on this literature we use the
following definition: Transparency is the availability of information about an organization or
actor allowing external actors to monitor the internal workings or performance (adapted
from Grimmelikhuijsen 2012, p. 55; Grimmelikhuijsen and Welch 2012; Meijer 2013).
Based on this definition, the type of transparency in this study (i.e., informative television
series about a Dutch district court) can be characterized as follows. First, the notion of
availability of information in the definition can refer to both pro-active and passive
transparency (Meijer, Curtin and Hillebrandt 2012). Traditionally, judicial transparency has
been attained by the openness of cases: in principle everyone is allowed to attend court cases
6
whenever they want. In this study, transparency is taken one step further by bringing it
proactively to citizens by broadcasting cases and deliberations on public television. The
television series extends the traditional openness in another way and shows how judges
assess a variety of cases in the court room and letting them explain decisions afterwards.
Although this a broadening of traditional openness, one can argue that this does not constitute
‘full’ transparency and in some respects is restriction of transparency, because a series is
directed and edited. On the other hand, the series was an independent journalistic production
and the film makers were given the freedom to tape the cases they wanted. A number of
judges were asked if they wanted to participate in the series and if they consented their cases
could be used in the series. Because the court room had a special back wall with a moving
camera installed the consenting judges did not know which of their cases would be used for
broadcasting. These are important features that increase transparency.
It is important to note that the visual nature of this type of transparency contributes to the
accessibility of the information. Visual information is becoming more important and cultural
theorists even argue we live in a ‘visual culture’ (e.g. Mirzoeff 1999). The visual nature is in
contrast with most transparency initiatives and studies, which tend to focus on textual
information, for example, on websites or on paper (e.g. Evans & Campos, 2013). Visual
information, and especially in a television series broadcasted nationally on public television
may be more appealing and makes information much easier to comprehend and access than
text based information tucked away on websites.
In sum, this means that this form of judicial transparency can be conceptualized as pro-active
and visual transparency. The next section will explain how this relates to the expected effects
on citizen trust in judges.
3. POTENTIAL EFFECTS OF JUDICIAL TRANSPARENCY ON TRUST
The effects of transparency has been heavily debated in the literature (Meijer 2009). Recent
empirical work has suggested a subdued and sometimes negative effect of transparency on
trust in public institutions (Worthy 2010; De Fine Licht 2011; Grimmelikhuijsen 2012).
However, it seems that the effect of government transparency depends on the type of policy
area, as some policies have a more controversial nature than others (De Fine Licht 2014).
7
It has therefore been suggested that government institutions suffer from a ‘negativity bias’.
Negative performance has a much more pronounced effect on citizen attitudes than good
performances (Baumeister et al. 2001; James 2011). In this article we argue that judicial
transparency is a peculiar case, and different from government transparency because the
judiciary is expected to profit from a positivity bias instead.
Research on judicial exposure to media attention suggests that the judiciary has a different
position than other democratic institutions (Gibson and Caldeira 2009). Based on an analysis
of longitudinal data they found that citizens respond positively to media exposure about the
judiciary. Based on positivity bias theory they explain how media exposure increased
perceived legitimacy of the Supreme Court, even when it concerns contentious confirmation
hearings. As Gibson, Caldeira, and Baird (1998, p. 356) explain:
“Generally, to be aware of a court is to be supportive of it. This positivity bias may be
associated with exposure to the legitimizing symbols that all courts so assiduously
promulgate.”
Positivity bias occurs because from early life citizens learn that the judiciary is ‘different’,
i.e., they use non-political, impartial, forms of decision-making and as such adhere the ‘myth
of legality’. If the judiciary is exposed, as transparency does, the symbols that are associated
with this distinctness and impartiality are exposed (Gibson and Caldeira 2009, p. 9). As a
result, exposure of the judiciary, even though it might be negative, makes these dormant
positive attitudes towards the judiciary salient.
In sum, transparency research has produced mixed results regarding the effect of government
transparency, but nevertheless transparency in particular areas is believed to contribute to
greater trust (De Fine Licht et al. 2014; Grimmelikhuijsen and Meijer 2014). Judicial
transparency is expected to profit from a positivity bias and therefore positive effects on trust
in judges are expected (Gibson and Caldeira 2009). We therefore postulate the following
hypothesis:
H1: Judicial transparency is expected to have an overall positive effect on trust in judges.
However, we believe that that the relationship between judicial transparency and trust is more
complicated and depends on various psychological factors. To better understand how
8
transparency might affect trust, we need to understand how people process information at the
micro-level. This is needed because people process information differently if they already
have certain beliefs or knowledge concerning the topic in question. Information will be
related and interpreted in terms of their pre-existing knowledge (Bohner 2001). The
Elaboration Likelihood Model (ELM), which was developed in the field of social psychology
by Petty and Cacioppo (1986), provides a theoretical basis for our analysis.
The ELM-study states that attitude change can take place both through effortful and effortless
processing of information. The central route of ELM - related to effortful processing of
information - proposes that people who are exposed to a persuasive message relate the
message to their existing knowledge about the message topic (Bohner 2001, p. 253).
Persuasion through the central route is a process that involves deliberative and active
information processing in terms of scrutiny.
The ELM also proposes a peripheral route based on social identification, conditioning, and
the use of heuristics. This route is chosen when people spend limited time or have little
knowledge to process a persuasive message. Likewise, people with little knowledge about
judges can be expected to change their attitude more easily as they cannot relate the new
information to existing patterns. Individuals who already have a great deal of knowledge
about the judiciary will relate new information to this existing knowledge. As a result, they
are expected to have more stable attitudes towards judges and their attitude will be more
difficult to change. An application of ELM in the field of public administration has been
carried out by Grimmelikhuijsen and Meijer (2014). Following the same logic as proposed
here, they found that the attitudes of people with high levels of knowledge were not
influenced by government information, whereas citizens who reported have little knowledge
did change their attitude after reading government information.
Based on the above mentioned theory and findings, we hypothesize that knowledge of the
judiciary moderates the relationship transparency and trust. High levels of prior knowledge
are expected to weaken this relation, whereas low levels of knowledge are expected to
strengthen it.
H2: High levels of prior knowledge weaken the effect of judicial transparency on trust in
judges.
9
Secondly, the psychological theory of confirmation bias provides a theoretical basis for the
idea that predispositions play a central role in judging information. Confirmation bias is the
phenomenon that people use information that is consistent with their beliefs, and that
evidence or information that does not fit the beliefs of a person may even be denied. As a
result, confirmation bias theory proposes that people tend to search for and use information
only if it confirms their own beliefs (Baron 2000; Nickerson 1998; Kunda 1999), which
results in overconfidence in these personal beliefs. Confirmation bias also applies to the
biased interpretation of information by different individuals.
People seek and use information that confirm their prior beliefs. This might also apply to
transparency. This means that the way the same informative television show is interpreted by
various people can differ (cf. Lord, Ross & Lepper 1979). One crucial variable that colors
someone’s perceptions of information is people’s predisposition to trust the judiciary system
in general. An individual’s propensity affects the likelihood that an individual will trust
something or someone. Someone’s predisposition to trust is different from specific trust in
the sense that it is a rather stable personal trait which differs from person to person and can be
seen as the general willingness to trust (Mayer et al. 1995, p. 715; McKnight et al. 2002, p.
336). People with different developmental experiences, personalities and cultural
backgrounds vary in their predisposition to trust. This will influence how much trust one will
place in a specific instance.
When we apply this theory to this study, somebody’s general trust in the judicial system at
large can be seen as a relatively stable predisposition, shaped by someone personal
development experience (see also Gibson and Caldeira 2009). This, in turn, is expected to
influence the effect of transparency. For instance, people with a weak predisposition to trust
(i.e. have low levels of trust in the judiciary system) are expected to either disconfirm the
information that comes to them through a television show, or will see their negative beliefs
endorsed. Instead, those with a strong propensity to trust (have high levels of trust in the
judiciary system at large already) may feel affirmed in their thoughts that the judiciary is
indeed trustworthy. This expectation is supported by a study by Avery (2009) who
investigated the moderating role of pre-existing political trust. Those citizens with low levels
of political trust already do not become more trusting nor less trusting, whether they are
watching television news or reading newspapers. In contrast, citizens with higher trust levels
10
become more trusting following exposure to newspapers but less trusting following television
news exposure. Avery’s findings suggest that pre-existing attitudes play an important role in
the relation between transparency and trust. Specific research on the judiciary also elucidated
the importance of predispositions with regard judicial legitimacy (Gibson and Caldeira 2009).
H3: High levels of predisposition to trust the judiciary system strengthen the effect of judicial
transparency on trust in judges.
4. METHOD
4.1 Institutional context and design
The judiciary in the Netherlands is one of the least corrupt in the world (listed 14th by
Transparency International in 2007). Judges in the Netherlands are appointed for life (until
their 70th birthday) to strengthen their independent position. They are formally appointed by
the head of state (the King) based on the advice of a committee consisting of members of the
judiciary, the public prosecutor and persons who play a societal role. There are no political
and open appointment procedures, no ‘dissenting opinions’ and there is no role for lay judges
in the Dutch judicial system.
It might be this closed system of appointments and decision-making that has recently sparked
the debate about a more transparent judiciary in the Netherlands. Scientists and other
commentators argue that transparency is necessary for high quality decisions and to restore
trust in the judiciary (Sterk 2013; Prins 2013). Public trust in the Dutch judiciary is relatively
high at the moment - 69 percent of the population express (a great deal of) trust in judges
whereas other public and political institutions consistently receive lower trust score.
However, in recent years (late 90s and mid 00s) trust rates dropped at around and sometimes
even below fifty percent (Croes 2011). This is one of the reasons the Utrecht District court
issued the development of a television series on public television by an independent
broadcast organization.
The series had an average of 900,000 watchers each episode, which is a rather largeaudience
for Dutch television. In comparison, on an average evening without any big televised events
(e.g. an important soccer match) the best-watched show is the evening news on the public
television reaching an average daily audience of 1,860,000 (in 2010). Of the treatment group,
40 percent watched all eight episodes and 66 percent watched six or more episodes.
11
To investigate the influence of the series on individual attitudes this study employed a field
experiment and consisted of two groups that both completed a pre-test and post-test, which is
a suitable design for making causal inferences (Shadish, Cook, and Campbell 2002).
4.2 Manipulation
The manipulation consisted of an instruction to watch the series. The control group was given
no such instruction. More detailed information on the manipulation can be found in the
Appendix. However, in the end we are not interested in the effect of our instruction as such,
but in the effect of the television series that people watched as a result of the instruction.
Therefore, we will elaborate on the content of series in more detail below.
The television series consisted of eight episodes. In each episode several judges (men and
women, young and old) were televised in different types of cases. A typical episode had a
duration of 25 minutes and covered three very different types of cases. For instance, one
episode reported about a nightly stabbing after a night out, a divorce, and an amateur soccer
club suing the national soccer association. So, different realms of the judicial practice were
presented to the public. These are interesting cases, but far from sensationalist. Each case is
presided by a different judge. However, given the total number of cases, some judges were
showed more than once in the full series. An episode contained video material from the case
itself, but judges also give personal testimonies about their earnest deliberations in this
particular case. The series was developed relatively free from influence of the court. Judges
could self-enroll to participate in the series, but did not know beforehand which cases would
be used and broadcasted, nor did the court have influence on how the material was edited.
Although informing the public about the actual work of judges (e.g. providing information
and deliberation in a variety of cases) was an important goal of the series, the episodes go
beyond merely providing information. The episodes in the series clearly express the symbols
and values of the judiciary that undergird the ‘myth of legality’ (cf. Gibson and Caldeira
2009). For example, judges are always shown in their robe, the use of the gavel is shown and
the ‘rise’ of the court room is shown repeatedly in each episode. Moreover, the title sequence
at the start of each episode shows these symbols repeatedly.
4.3 Data Collection and Sample
12
The sample was drawn from online respondent panel managed by a research company
(Mediatest) that regularly carries out public opinion research projects, administering the
questionnaires in the pre- and posttest to an online panel. The total panel consists of
respondents that are representative for the Dutch population on a number variables, such as
age and sex.
A random group from this panel was asked to participate in the project. They were informed
about the start of the series and were instructed to watch them as well as filling out surveys
that were (and would be) digitally provided by the company. To watch the series we refer to
the following website: www.uitzendinggemist.nl/programmas/1890-de-rechtbank (accessed
May 23, 2014, only in Dutch).
The control group was asked to complete a questionnaire and were told they would be
approached a later moment in time. Of course, no mention was made about the television
show; they were told about the councils’ wish to get better informed about the trust in the
judiciary and the judges as well. Because of the longitudinal design of the research and the
required discipline of the participants in watching the show, we expected a higher drop-out
rate and so we oversampled the treatment group (see Table 1).
--TABLE 1 around here--
Altogether 2148 people were approached for the treatment group of which 1515 (70.5%)
completed the pretest questionnaire. 585 respondents (27%) completed the posttest
questionnaire. A reason for the low percentage of dropouts may be that respondents that had
not watched lost interest at an early stage and refrained from completing the posttest
questionnaire. In addition, after four episodes (out of eight in total) there was a break of a few
weeks in the series because of the Christmas holidays. In this break we sent a reminder to all
treatment group respondents to ask them if they watched the series already and that the series
would restart with four remaining episode in the new year. For the control group 872 people
were sampled of which 93% responded to the pretest questionnaire. 464 respondents
completed the posttest questionnaire.
The declining response rates in Table 1 indicate that there may be some selection bias in the
process of sampling. First, there is a relatively high dropout rate in the treatment group.
13
Presumably, respondents who received the nudge to watch the show, were more likely to
refrain from completing the pretest questionnaire in the first place. Only 70.5 percent
completed the pretest questionnaire, whereas 93.0 percent of the control group did do so.
We checked the dataset (n=1049) for comparability on three key background traits: age,
education and prior experience. Tyler (2001) found a positive correlation between age and
trust in judges, Bovens and Wille (2011) asserted that citizens with university degree tend be
more trusting towards government and other public institutions than citizens with lower
degrees. Prior experience with courts was taken into account, because prior analyses have
shown that this negatively albeit marginally affects perceptions of judges (e.g. Van de
Walle 2009).
--TABLE 2 HERE--
Next we tested for significant differences between groups on these background traits. Ideally,
there are no significant differences between the experimental groups, because respondents
were assigned to experimental groups randomly. Nevertheless, a significant difference was
found regarding the average age of between the control and treatment group. Uneven
distribution of background traits in experimental groups can be problematic as it may
confound the results of the experiment. However, we think that the impact of this bias on the
results of this study is limited. First of all as Table 2 in the ‘Results’ Section will point out
only has a very weak negative correlation with pre- and post-test trust in judges. Differences
between the control and treatment group are therefore attributable to age to a very limited
extent. Furthermore, although the age difference between control and treatment group is
significant, the difference is less than five years which will make the potential confounding
impact of age even smaller. To be certain that this confounding effect is eliminated we took
age into account as a covariate (control variable) in our analyses, to control for its potential
confounding effect.
4.4 Measures
Trust
In measuring trust, we followed Tyler’s operationalization of trust in judges (Tyler 2001). He
measured trust in terms of judges are honest, court decisions are fair, courts protect citizen
rights, decisions are fair, court guarantees everyone a fair trial (Tyler 2001, p. 218).
14
According to Tyler, “this produces an overall index of confidence or institutional trust. The
items we used were similar to Tyler’s; participants were asked to indicate their level of trust
on a 5-point scale in: the dedication, the competence, the integrity, the impartiality and the
decisions of judges (Cronbach’s alpha pre-test = 0.91, Cronbach’s alpha post-test = 0.92).
Predisposition to trust
Predisposition to trust was measured using generalized trust in the criminal justice system as
a proxy. This was operationalized using two items: The Dutch justice system is effective in
battling crime and The Dutch criminal justice system performs adequately. (Cronbach’s alpha
= .748). Both items were assessed on a 5-point scale. Based on these two items participants
could have a mean predisposition score of 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5 and 5.0. To
assess the moderation effect of predisposition we devised three subsets of respondents. To
determine the borders of the middle category, we used the mean score (2.65) and median
score (2.5) on these two items. Therefore, respondents in the 2.5 (median) category and 3.0
(mean) category constituted the moderate predisposition group (N=468). Participants in the
1.0-2.0 group were assigned to the “low predisposition” group (N=365) and the others (3.5-
5.0) to the high predisposition group (N=216).
Knowledge
Knowledge was measured using an item measuring self-assessed knowledge: How do you
assess your knowledge about criminal justice in the Netherlands? The scale ran from 1 (very
low) to 5 (very high). To assess the moderator effect of knowledge we used three subsets of
respondents: one set of low knowledgeable respondents (score of 1 or 2), one set of medium
knowledgeable respondents (score of 3) and one set of high knowledgeable respondents
(score of 4 or 5). This operationalization of knowledge was based on self-assessment and was
only measured by one item. In the absence of validated objective measures of judicial
knowledge we considered this the second-best option to serve as a proxy for actual
knowledge. To make sure knowledge was not an effect of the treatment, we measured
knowledge before the series started. 198 respondents were assigned to the high knowledge
subset, 509 in the medium knowledge subset and 342 in the low knowledge subset.
4.5 Analysis
We used the Instrumental Variables (IV) method to analyze our field experiment.
Instrumental variables are useful in case there is concern that the treatment and unobserved
15
factors that might affect the outcome…are correlated in some significant way…the issue is
not assigned treatment, but rather the actual treatment as received and experienced by the
subject (McDermott 2011, p. 31). What would this mean for this study? According to
Angrist (2006), the IV method should be applied when treatment compliance is not perfect, as
is the case in our field experiment. In other words, several participants who were assigned to
the control group still watched the television series.
IV methods can be applied to our study as follows. IV sees the intention to treat as an
instrument to the actual explanatory variable, which is in our case the number of television
shows participants had watched. This can be tested with a two-stage least squares (2SLS)
regression analysis which test the relation between instrument on explanatory variable, and in
the second stage the effect of the explanatory variable on the dependent variable. We
constructed a difference variable of pre-test trust minus post-test trust in judges and entered
this as the dependent variable in the equation. We report the betas for the second stage, which
designates the effect of television episodes watched on trust in judges: whole population, For
further details on IV methods in social science we refer to the work of Angrist (2006).
5. RESULTS
5.1 Overall analysis
We included a correlation table between key background variables, the two moderating
variables (knowledge and predisposition) and the pretest and posttest trust means in the
Appendix. The highest correlations occur between predisposition to trust the system and trust
in judges. As judges are part of the justice system this makes sense. It should be noted,
however, that even though these correlations are high compared to other statistics in the table,
they are not anywhere near values that indicate multicollinearity (above .8 or .9). For more
details we refer to the appendix.
First, we analyzed the effect of the series on the overall sample. Looking at the regression
analysis on the overall population, we found a main effect of the television series on trust in
judges (F=22.54, β=.254, p <.001, R2=.041). This indicates a moderate effect of watching the
television series: participants who watched more episodes tend to have more trust in judges.
4.1 percent of the change in trust can be explained by the series. Next, separate analyses were
carried out to show how knowledge and predisposition moderates the relationship between
16
judicial transparency and trust. First we will show the results for predisposition and then for
knowledge.
How is the low predisposition group affected by watching the show? The 2SLS regression
indeed shows that watching the show affects (β=.365, p <.001). This indicates a significant
and rather strong effect, indicating that transparency explains 9.1 percent of the variation in
trust in judges. The next analysis provides results on how the moderate predisposition group
is affected by the series. When we applied 2SLS regression again detect a clear effect of the
transparency manipulation, although less strong (β=.234, p <.001), explaining 3.5 percent.
Finally the high predisposition group is analyzed. The effect remains significant, but
considerably weaker compared to the moderate and low predisposition groups (F=4.30,
β=.175, p <.05). Tables 3-5 display these results.
1
TABLE 3, 4 AND 5 AROUND HERE
Secondly, three separate analyses were carried to determine the effect of prior knowledge on
watching the show. In the low knowledge subset we found an effect of the watching the
television series (β=.229, p < .01) with a reasonable effect size. The size of the effect
increased when we carry out the 2SLS regression on the medium knowledge participants
(β=.269, p <.001). Finally, also in the high knowledge group we found a difference (β=.225, p
<.01), but slightly more subdued than in the medium knowledge subset. In conclusion, the
television shows seems to have a stronger positive effect on people that have some substantial
knowledge of the judiciary, compared to participants with (very) little knowledge. Tables 6-8
show of the analyses of all three subsets.
TABLES 6, 7 and 8 AROUND HERE
Overall, a slightly weaker effect can be found in the low knowledge group. A weak effect
was found in the high knowledge subset, however, there was a significant difference in pre-
1
We carried out an additional regression analyses in which the predisposition variable was included as an
interaction term in a ‘regular’ OLS regression analysis. In this additional regression (DV=difference between
pre- and posttest) analysis we found similar results: a negative effect of the moderating variable predisposition
(β=-.174, p<.001).
17
test attitudes between control and treatment group which means we should interpret the
findings from this subset with some caution.
2
To conclude the results section, we carried out an intention to treat (ITT) analysis to test the
robustness of our results. ITT is a way to analyze the data the way it was intended in the
experiment (cf. Angrist et al. 1996). For instance, respondents in the control group who
watched the television series anyway are still treated as being in the control group (i.e., the
way the treatment was intended by the researcher). This analysis is useful to assess how
sensitive our results are for a selection bias that may be caused by the unintended migration
of respondents from one experimental group to another. We used ANCOVAs to assess the
effect of our intended treatment on the change in trust in judges. The results are listed in
Table 9:
TABLE 9 AROUND HERE
These results in Table 9 are very similar to those reported in the 2SLS regression. Again the
results show the mitigating effect of one’s predisposition to trust the judicial system. With
regard to self-assessed knowledge the ITT analysis show something new. While the treatment
group mean increased almost as much as in the medium knowledge subset the overall
difference is smaller because of an increase in trust in the control group. The overall
conclusion from the 2SLS regressions, i.e, that the treatment has the strongest effect for those
with a medium level of knowledge, is upheld by this auxiliary analysis nevertheless.
6. CONCLUSION AND DISCUSSION
The results of our study show that there is an overall positive effect of judiciary transparency
on trust in judges. This means that hypothesis 1 is supported by our conclusion. Our positive
result is in contrast with recent studies that found a negative effect of transparency of
2
In the high knowledge subset we found large pre-test differences for trust (2.83 in the control group and 3.22
in the treatment group (F=9.20, p=.003) and increase in trust in both the control and the experimental group.
This indicates there is some self-selection bias present. Critical and knowledgeable respondents were less likely
to decide to watch the show as compared to trusting and knowledgeable respondents. The second issue an
increase in trust in both the treatment and control group suggests that exogenous factors have relatively large
influence on respondents with high knowledge as compared to other respondents. Highly knowledgeable
citizens may follow the news about the judiciary much more closely by means of other sources resulting in a
marginal influence of the television series.
18
government and political institutions on citizen trust (Worthy 2010; De Fine Licht 2011;
Grimmelikhuijsen 2012). On the other hand, the positive result does resonate with the study
of Gibson and Caldeira (2009) who found that the judiciary is ‘different’ from other
institutions in the sense that it merits from a positivity bias, which is aroused by exposure to
symbols that capture the ‘myth of legality’: the judiciary is impartial and has non-political
mechanisms of decision making. In the debate on transparency this is a significant finding as
researchers have only started to discover the influence of various contextual variables on the
relationship between transparency, trust and legitimacy (e.g. Grimmelikhuijsen et al. 2013;
De Fine Licht 2014).
Our study adds two important elements to this general finding, namely the influence of prior
knowledge and predisposition to trust the criminal justice system. The second hypothesis
predicted a mitigating influence of knowledge on the relation between transparency and trust.
This was partly supported by our results. The least pronounced effect was found for the high
knowledge group. However, this finding should be interpreted with caution, because we
found rather large differences in the pretest trust score of participants. We further found that
the effect of transparency was the strongest for people with medium knowledge, which
indicates that transparency works best for those citizens who have some basic knowledge
about the judiciary. This makes sense as basic knowledge about the judiciary is needed to
place the information the series has to offer in a proper frame of reference. Overall, the ELM
model for processing information (Petty and Cacioppo 1986) seems to work quite well in
predicting the moderating effect of knowledge. Based on this model we predicted that
citizens that do not know that much about the judiciary will more easily add new information
to their current patterns of thinking, which will lead to faster attitude change compared to
citizens with high levels of knowledge. This was especially true for those who had moderate
knowledge of the judiciary.
This is a relevant finding for practitioners, too. The fact that the type of transparency under
study is most effective on medium knowledgeable citizens seems to be promising.
Traditional, heavy text-based types of transparency often reach only a limited audience,
presumable highly knowledge and well-educated citizens. However, in this case of visual
transparency a different and broader audience is reached, and specifically an audience that is
more likely to gain trust by this type of transparency: approximately 75 percent of our sample
19
reports to have ‘little’ to ‘medium’ knowledge. This means that transparency can be valuable
to reach out to a broader public that is otherwise less knowledgeable about public institutions.
Hypothesis 3 is rejected based on our findings. Individuals with a high predisposition to trust
showed less increase between reported trust in judges before and after, disconfirming
hypothesis 3. Based on the theory of confirmation bias (e.g. Nickerson 1998) we predicted
that people with a positive predisposition towards the justice system were more likely to
increase their trust. Since our results proved otherwise we need an alternative explanation for
our findings, i.e., that citizens with high predispositions were less likely to have an increase
in their trust. A simple alternative explanation is the existence of a ceiling effect to trust.
Individuals with a high predisposition concerning the justice system will not easily gain even
more trust by watching a television show. Citizens with relatively little trust have much more
room to increase their trust in judges.
A limitation is our focus on the immediate influence of a one-off series and we did not
investigate what specific cues were most beneficial for trust in judges. These limitations
provoke some new questions that should be tackled in future research. For instance, to what
extent will the positive effect of watching the series last? To what extent can the positive
effect of watching the series counter negative experiences with courts, for instance, negative
reports by the press? Furthermore, we investigated the influence of the series as a whole.
More rigorous laboratory experiments would be needed to assert this. That said, a field
setting more closely aligns with a real life setting in which people watch a series, at home on
their couch, and not in a laboratory. A final question for future research would be to assess
cultural differences. In other countries where people hold less favorable views regarding the
judiciary, for instance, because of a deeply corrupt judicial system, positivity bias might not
apply. More work is needed to further our understanding on these contextual influences on
trust and judicial transparency.
That said, the main conclusion of this study is fairly optimistic and shows that transparency in
the judicial context can contribute to trust. This can be explained by the unique traits of visual
judicial transparency: the exposure to typical judicial symbols that imply impartiality and
non-political decision-making triggers a positivity bias among citizens. Transparency seems
to be fit to trigger these positive symbolic evaluations in order to win the hearts and minds of
the public.
20
21
REFERENCES
Angrist, J.D. 2006. ‘Instrumental variables methods in experimental criminological research:
what, why and how.’ Journal of Experimental Criminology, 2, 23-44.
Angrist, J.D., Imbens, G. W. and Rubin, D. B. 1996. Identification of causal effects using
instrumental variables. Journal of the American Statistical Association, 91, 434, 444
455.
Avery. J.M. 2009. Videomalaise or Virtuous Circle? The Influence of the News Media on
Political Trust’, International Journal of Press/Politics, 14, 410-33.
Baron, J. 2000. Thinking and deciding. 3rd edn, New York: Cambridge University Press.
Baumeister, R.F., Bratslavsky, E., Finkenauer, C. and Vohs, K.D. 2001. Bad is stronger than
good. Review of General Psychology, 5, 4, 323-70.
Bohner, G. 2001. ‘Attitudes’, in M. Hewstone and W. Stroebe (eds), Introduction to Social
Psychology. Malden: Blackwell Publishing, pp. 239-82.
Bovens, M. and A. Wille. 2011. Diplomademocratie: Over de spanning tussen meritocratie
en democratie [Diploma democracy: about the tension between meritocracy and
democracy]. Amsterdam: Prometheus.
Croes, M.T. 2011. ‘Heeft de burger vertrouwen in de rechter?’ [Do citizens trust judges?], in:
M. Hertogh, & H. Weyers (eds.), Recht van onderop. Antwoorden uit de
rechtssociologie, Nijmegen: Ars Aequi Libri, 301-324
Curtin, D. and A.J. Meijer. 2006. Does Transparency Strengthen Legitimacy? Information
Polity, 11, 109-23.
De Fine Licht, J. 2011. Do We Really Want to Know? The Potentially Negative Effect of
Transparency in Decision Making on Perceived Legitimacy.’ Scandinavian Political
Studies, 34, 183-201.
De Fine Licht, J. 2014. Policy Area as a Potential Moderator of Transparency Effects: An
Experiment’, Public Administration Review, 74, 3, 361-371.
De Fine Licht, J., Naurin, D., Esaiasson, P. and M. Gilljam. 2014. When Does Transparency
Generate Legitimacy? Experimenting on a Context-Bound Relationship’, Governance,
27, 111-34.
Evans, A.M. and A. Campos. 2013. ‘Open government initiatives: challenges of citizen
participation’, Journal of Policy Analysis and Management, 32, 1, 172203.
Gerring, J. and S.C. Thacker. 2004. Political Institutions and Corruption: The Role of
Unitarism and Parliamentarism’, British Journal of Political Science, 34, 295-330.
22
Gibson, J.L. and G.A. Caldeira. 2009. Citizens, Courts, and Confirmations: Positivity Theory
and the Judgments of the American People. Princeton University Press, Princeton
Gibson, J.L., Caldeira, G.A. and V.A. Baird. 1998. ‘On the legitimacy of national high
courts’, The American Political Science Review, 92(2), 343-358.
Grimmelikhuijsen, S.G. 2010. ‘Transparency of Public Decision-Making: Towards Trust in
Local Government?’ Policy and Internet, 2, 1, article 1.
Grimmelikhuijsen, S.G. 2012. Transparency and trust. An experimental study of online
disclosure and trust in government. Dissertation, Utrecht University.
Grimmelikhuijsen, S.G. and E.W. Welch. 2012. Developing and Testing a Theoretical
Framework for Computer-Mediated Transparency of Local Governments’, Public
Administration Review, 72, 562-571.
Grimmelikhuijsen, S.G. and Meijer, A.J. 2014. The effects of transparency on the perceived
trustworthiness of a government organization: evidence from an online experiment.
Journal of Public Administration Research and Theory, 24, 137-57.
Grimmelikhuijsen, S.G., Porumbescu, G. Hong, B., and T. Im. 2013. ‘The ffect of transparency
on trust in government. A cross-national experiment.’ Public Administration Review, 73,
4, 575-586.
Hardin, R. 1993. The street-level epistemology of trust’, Politics and Society, 21, 505529.
Hardin, R. 2002. Trust & Trustworthiness. Volume in the Russell Sage Foundation series on
trust. Russell Sage Foundation: New York.
Heald, D. 2003. ‘Fiscal transparency. Concepts, measurement and UK practice’, Public
Administration, 81, 4, 723-59.
Hibbing, J.R., & Theiss-Moore, E. (1995). Congress as public enemy: Public attitudes
toward American political institutions. Cambridge: Cambridge University Press.
Hibbing, J.R., & Theiss-Moore, E. (2002). Stealth democracy: American’s beliefs about how
government should work. Cambridge: Cambridge University Press.
James, O. 2011. ‘Managing Citizens’ Expectations of Public Service Performance: Evidence
from Observation and Experimentation in Local Government. Public Administration
89, 4, 141935.
Kunda, Z. 1999. Social Cognition: Making Sense of People. Cambridge: MIT Press.
Lewis, J.D. and A. Weigert. 1985. ‘Trust as a Social Reality’, Social Forces, 63, 967-985.
Lord, C.G., Ross, L. and M.R. Lepper. 1979. ‘Biased Assimilation and Attitude Polarization:
The Effects of Prior Theories on Subsequently Considered Evidence’, Journal of
Personality and Social Psychology, 37, 2098109.
23
Luhmann, N. 1979. Trust and power. Chichester: Wiley.
Mayer, R., Davis, J.H. and Schoorman, D. 1995. An integrative model of organizational trust.
Academy of Management Review, 20(3), 709734.
McDermott, R. 2011. Internal and external validity, in J.N. Druckman, P.G. Green, J.H.
Kuklinski, and A. Lupia (eds.), Cambridge handbook of experimental political science,
pp. 27-40.
McKnight, D.H., Choudhury, V. and Kacmar, C. 2002. Developing and validating trust
measures for e-commerce: An integrative typology. Information Systems Research,
13(3), 334-359.
Meijer, A.J. 2009. Understanding Computer-mediated Transparency’, International Review of
Administrative Sciences, 75, 255-69.
Meijer, A.J. 2013. ‘Understanding the Complex Dynamics of Transparency.’ Public
Administration Review, 73, 429-439.
Meijer, A.J., Curtin, D. and M.Z. Hillebrandt. 2012. Open government: connecting vision and
voice’, International Review of Administrative Sciences, 78, 10-29.
Mirzoeff, N. 1999. An introduction to visual culture. Routledge: New York.
Nickerson, R.S. 1998. Confirmation Bias. An Ubiquitous Phenomenon in Many Guises.
Review of General Psychology, 2, 175220.
O’Neill, O. 2002. A Question of Trust: The BBC Reith Lectures 2002. Cambridge: Cambridge
University Press.
O’Neill, O. 2006. ‘Transparency and the Ethics of Communication’, in C. Hood and D. Heald
(eds), Transparency: The Key to Better Governance? Oxford: Oxford University Press.
Petty, R.E. and J.T. Cacioppo. 1986. Communication and Persuasion: Central and Peripheral
Routes to Attitude Change. New York: Springer-Verlag.
Prins, C. 2013. Rechtspraak en transparantie: inbreng voor een noodzakelijk debat [The
judiciary and transparency: input for a necessary debate]. Rechtstreeks, 9, 16-37
Rousseau, D., Sitkin, S., Burt, R., and C. Camerer. 1998. Not so Different After All: A Cross-
Discipline View of Trust’, The Academy of Management Review, 23, 393-404.
Shadish, W.R., Cook, T.D., and D.T. Campbell. 2002. Experimental and quasi-experimental
designs for generalized causal inference. Houghton Mifflin Company: Boston/New
York.
Sterk, K. 2013. Naar een meer open karakter van de rechtspraak [Towards a more open
character of the judiciary]. Rechtstreeks, 9, 12-5.
Thibaut, J. and Walker, L. 1975. Procedural Justice. Hillsdale, NJ: Erlbaum.
24
Transparency International. 2007. Global corruption report 2007. Cambridge: Cambridge
University Press.
Tyler, T.R. 2006. Why people obey the law. Princteon: Princeton University Press.
Tyler, T.R. 2001. Public trust and confidence in legal authorities: what do majority and
minority group members want from the legal and legal institutions? Behavioral
Sciences and the Law, 19, 215-235.
Tyler, T.R. 1998. Public mistrust of the law: A political perspective. University of
Cincinnati Law Review, 66, 847-876.
Tyler, T.R., Boeckmann, R., Smith, H.J. and Y.J. Huo. 1997. Social justice in a diverse
society. Denver, CO: Westview.
Van de Walle, S. 2009. Confidence In The Criminal Justice System: Does Experience
Count? British Journal of Criminology, 49, 384-398.
Welch, E.W., Hinnant, C.C. and M.J. Moon. 2005. Linking Citizen Satisfaction with E-
Government and Trust in Government. Journal of Public Administration Research and
Theory, 15, 371-391.
Worthy, B. 2010. More Open but Not More Trusted? The Effect of the Freedom of Information
Act 2000 on the United Kingdom Central Government’, Governance, 23, 561582.
Zucker, L.G. 1986. Production of trust: Institutional sources of economics structures, 1840
1920’, in B.M. Staw and L.L. Cummings (eds), Research in organizational behavior.
Greenwich, CT: JAI Press, pp. 53-122.
25
Table 1 Response rates
Approached
Completed
pretest
Completed
posttest
Overall
response rate
Treatment
2148
1516
585
27.2%
Control
872
811
464
57.2%
26
Table 2 Sample information for key background variables
Age
Education
(1=high)
Experience
(1=prior case)
Control (N=464)
46.7 (13.8)
0.38 (.49)
0.37 (.48)
Treatment (N=585)
42.9 (13.5)**
0.46 (.50)
0.32 (.47)
*** p < .001 **p <.01
27
Table 3 2SLS regression for low predisposition subset (n=365)
Assumption of multicollinearity not violated (VIFs do not exceed 1.024). Standardized betas
are displayed. *p<.05 ** p<.01 *** p< .001
Step 1:
Treatment condition
Number of shows watched
Step 2:
Number of shows watched
Trust in judges
Number of shows watched
N/A
.365***
Treatment condition
.832***
N/A
Age (control)
-.017
-.038
F 416.85***
R2 .697
F 19.04***
R2 .095
28
Table 4 2SLS regression for moderate predisposition subset (n=468)
Step 1:
Treatment condition
Number of shows watched
Step 2:
Number of shows watched
Trust in judges
Number of shows watched
N/A
.234***
Treatment condition
.799***
N/A
Age (control)
-.052
.044
F 429.56***
R2 .649
F 8.37***
R2 .035
Assumption of multicollinearity not violated (VIFs do not exceed 1.009). Standardized betas
are displayed. *p<.05 ** p<.01 *** p< .001
29
Table 5 2SLS regression for high predisposition subset (n=216)
Step 1:
Treatment condition
Number of shows watched
Step 2:
Number of shows watched
Trust in judges
Number of shows watched
N/A
.175*
Treatment condition
.791***
N/A
Age (control)
-.049
.000
F 187.69***
R2 .638
F 2.15
R2 .020
Assumption of multicollinearity not violated (VIFs do not exceed 1.017). Standardized betas
are displayed. *p<.05 ** p<.01 *** p< .001
30
Table 6 2SLS regression for low knowledge subset (n=342)
Step 1:
Treatment condition
Number of shows watched
Step 2:
Number of shows watched
Trust in judges
Number of shows watched
N/A
.229**
Treatment condition
.795***
N/A
Age (control)
-.025
.-013
F 294.54***
R2 .635
F 5.85**
R2 .028
Assumption of multicollinearity not violated (VIFs do not exceed 1.002). Standardized betas
are displayed. *p<.05 ** p<.01 *** p< .001
31
Table 7 2SLS regression for medium knowledge subset (n=509)
Step 1:
Treatment condition
Number of shows watched
Step 2:
Number of shows watched
Trust in judges
Number of shows watched
N/A
.269***
Treatment condition
.805
N/A
Age (control)
-.049
.016
F 505.95***
R2 .667
F 12.23***
R2 .046
Assumption of multicollinearity not violated (VIFs do not exceed 1.043). Standardized betas
are displayed. *p<.05 ** p<.01 *** p< .001
32
Table 8 2SLS regression for high knowledge subset (n=198)
Step 1:
Treatment condition
Number of shows watched
Step 2:
Number of shows watched
Trust in judges
Number of shows watched
N/A
.225**
Treatment condition
.834***
N/A
Age (control)
-.043
.132
F 236.09***
R2 .708
F 4.56*
R2 .045
Assumption of multicollinearity not violated (VIFs do not exceed 1.021). Standardized betas
are displayed. *p<.05 ** p<.01 *** p< .001
33
Table 9 ITT analyses using ANCOVAs
Table shows the mean change between the pre-test and post-test, on a 5-point scale. * differs
at p <.05 ** differs at p < .01 *** differs at p < .001. All analyses include ‘Age’ as a
covariate.
Mean change in trust
in control group
Mean change in trust
in treatment group
General analysis
.078
.322***
F 45.20***
Partial η 2=.041
Low PD
.120
.519***
F 36.19***
Partial η 2=.091
Moderate PD
.101
.307***
F 16.74***
Partial η2=.035
High PD
-.074
.092*
F 4.095*
Partial η2=.019
Low Knowledge
.055
.262**
F 11.65***
Partial η2=.033
Medium Knowledge
.081
.350***
F 23.86***
Partial η2=.045
High Knowledge
.127
.341**
F 7.04**
Partial η2=.035
... Although some argue that transparency is key to creating an atmosphere of openness that facilitates trust, others argue that transparency creates and only feeds existing mistrust [24]. Although existing studies do demonstrate a positive effect of transparency on trust [25][26][27][28], these studies also show that the relationship between both constructs is highly dependent on the context [27,28]. Therefore, the results of most of these studies cannot directly be transferred to the context of technology and/or AI. ...
... Although some argue that transparency is key to creating an atmosphere of openness that facilitates trust, others argue that transparency creates and only feeds existing mistrust [24]. Although existing studies do demonstrate a positive effect of transparency on trust [25][26][27][28], these studies also show that the relationship between both constructs is highly dependent on the context [27,28]. Therefore, the results of most of these studies cannot directly be transferred to the context of technology and/or AI. ...
... The experimental findings show that both the "why" and "why not" scenario increase the level of understanding reported by respondents. The increased level of understanding also positively influenced the level of trust confirming previous studies outside the realm of AI that have reported a similar effect [27], but only in the "why not" scenario. ...
Preprint
Full-text available
This study uses an experimental setting to study the effect of transparency concerning the use of an AI tool as an alternative to explanations that are common in the field of Explainable AI. It specifically addresses the impact of transparency on the level of understanding and trust among non-expert users.
... By contrast, some studies analyse trust using indicators that refer to the performance of courts and the attributes of judges by asking respondents whether they think that courts protect citizen rights, that judges are honest, that court decisions are fair, and that the courts guarantee everyone a fair trial (e.g. Tyler 2001; Grimmelikhuijsen and Klijn 2015). This seems problematic because it blurs the concept of trust and the drivers for trust. ...
... The link between trust in the Constitutional Court and the expectation of effective output in the form of expert judgments, points to the importance of functional process evaluations that are not usually captured by surveys utilized for measuring public trust (Grosskopf 2008). This supports the idea that trust in courts is different from trust in political institutions because of the specific characteristics attributed to courts (trustee-related drivers), such as impartiality, independence and expertise (Grimmelikhuijsen and Klijn 2015). In a recent study Garoupa and Magalhaes (2020) show that public trust in the judicial systems is positively related with fundamental institutional properties such as judicial independence and judicial accountability. ...
Technical Report
Full-text available
This deliverable offers a systematic and comprehensive review of the literature on trust and regulation and their relations in three parts. Section 1 provides a brief overview of trust and distrust; their relationship; and antecedents (drivers) and positive and negative consequences. This second part of provides an overview of phases and processes of trust building, maintenance and repair as well as a review of how processes of distrust building or trust-reduction work. The third part suggests that the unfolding of trust relationships varies according to the type of involved actor, i.e. an individual, organisation, or system (such as regulatory regimes) and level of analysis. Section 2 moves to regulation. It reviews the main concepts of regulation and clarifies some of the most important questions around it. This should allow us, in later sections of this deliverable, to point on some of the possible directions that trust and regulation research may take. It starts with definitions and a distinction between narrow and broad approaches to regulation. It then clarifies the concept of regulatory regime; the regulatory agency, the regulatory state and regulatory capitalism. It concludes with an actor centered analysis of the regulation. Section 3 deals with trust and distrust in government. It starts with a summary of the drivers of the dis/trust relations. The main sub-section deals with reviews of the literature on trust and distrust in (a) political institutions and actors; (b) public administration; (c) among courts; (d) citizens by government and (e.) between public organizations. Section 4 brings out the critical aspects of the review. It focuses on the relations between trust and regulation and offer a new conceptualization of their relations. One that will serve as the basis for section 1.3. The first part of the section moves the discussion of the relations forward in an attempt to looking at the relations beyond the current literature by distinguishing between four types of relations: independent, competitive, substitutive and supportive. The second part of the section focuses on the relations of trust between the main actors of the regime. The third discusses the relations between trust and regulation, when they touching on explanations for the processes. The fourth revolves around on assessment of outcomes of the relations. Section 5 deals with the operationalization and measurement of trust. It deals with the ways in which trust between actors in regulatory contexts has been measured in recent scholarship. It informed by, and seeks to expand upon, the current knowledge by more specifically examining measures used in the study of regulatory contexts. It starts with a description of the systematic methodology used. It then summarizes the purely descriptive findings of the review and an overview of the number of studies on trust in regulatory regimes published in recent years. It then presents separate discussions on how trust in different kinds of actors in regulatory regimes has been measured. Section 5 closes with a reflective and critical discussion of the literature. We analyse prominent limitations and gaps in existing empirical work.
... This finding is explained with an effect of procedural justice (Gallagher and Ashford 2021). Support for the judiciary is also increased in participants who are induced to watch a Dutch TV series that makes judicial procedure more transparent (Grimmelikhuijsen and Klijn 2015). ...
Preprint
Full-text available
Judges are human beings. Is their behavior therefore subject to the same effects that psychology and behavioral economics have documented for convenience samples, like university students? Does that fact that they decide on behalf of third parties moderate their behavior? In which ways does the need matter to find a solution when the evidence is inconclusive and contested? How do the multiple institutional safeguards resulting from procedural law, and the ways how the parties use it, affect judicial decision-making? Many of these questions have been put to the experimental test. The paper provides a systematic overview of the rich evidence, points out gaps that still exist, and discusses methodological challenges.
... 9 In addition, the development of views on the role of the judiciary in society has led to the use of such words as publicity, accessibility, and transparency. 10 This was significantly influenced by international regulations, including the Universal Declaration of Human Rights (Art. 10), 11 the International Covenant on Civil and Political Rights (Art. ...
Article
Full-text available
The processes of transition to democracy that post-Soviet countries underwent in the early 1990s predetermined different directions for their further development. The author presents and proves the hypothesis that in the context of post-Soviet civil society, judicial transparency arose as a response to a social demand at a certain historical moment of crisis of public authority. The idea of transparency in post-Soviet countries appeared only at a certain level of development of political institutions and public law, pointing out the democratic transition of power. At the same time, its emergence established information asymmetry and the poor quality of state institutions of power. The need to ensure the transparency of judicial activity, in addition to the natural process of the democratic transition of power, is also driven in post-Soviet countries by two important factors. The first is that in the modern world, the judiciary is increasingly becoming involved in the process of law-making, which requires the transformation of existing ideas about the system of checks and balances. The second is related to a global tendency in the fight against corruption, which has been a key problem for the countries of this region for many years. Although it has become the de facto rule for developed democracies, transparency affects the development of the legal culture of populations in transitional democracies differently. It performs various functions, including educational, preventive, stimulating, communicative, protective, and others. The article pays special attention to the unique forms of communication between courts and the public that have arisen in post-Soviet countries with an unstable political situation. In studying them, the author highlights the transformation of transparency from a factor of the development of civil society into one of its results.
... In this vein, insecurity of using the app may decrease and trust in the public institutions may increase. In general, transparent communication of the (prosocial) reasons underlying certain behavioral regulations (Grimmelikhuijsen & Klijn, 2015) and epidemic adequate behavior of policy representatives could facilitate trust and, thus, increase people's prosocial usage of a tracing app. ...
Article
Full-text available
Contact tracing apps have been identified as a promising technology to curb the spread of COVID-19. To be effective, a sufficient number of individuals needs to install the app and to disclose information like a COVID-19 infection to such an app. Yet, usage data demonstrates that a large amount of app users does not disclose a COVID-19 infection to the app. Hence, in two studies (overall N = 1,522), we investigate factors related to individuals’ willingness to actively disclose information to such an app. In a preregistered online-experiment conducted two months before the app launch onto the German market, we find that disclosure willingness increases when the app’s prosocial benefit or a social-life-enabling benefit is emphasized (vs. no benefit emphasized). In a subsequent, quota-representative survey study conducted two months after the app launch onto the German market, we adapted and extended the Technology Acceptance Model 2 (TAM2) to the context of prosocial information sharing in tracing apps. We find that the perceived prosocial benefit of the app, trust in public institutions, and fear of COVID-19 are relevant predictors. Moreover, we demonstrate that the relation between perceived prosocial benefit and disclosure willingness is moderated by perceived ease of use. Results are discussed with regard to effective implementation and communication strategies for a tracing apps, and the general role of prosocial concerns for technology usage to address major societal challenges.
... Over recent decades, the field has seen the adoption of theories, concepts and methods from psychology, behavioural economics and cognitive sciences. Labelled as "behavioural public administration" (Grimmelikhuijsen et al., 2017), this latest advance brings a range of new concepts, including cognitive bias and nudging, and new insights into existing public administration scholarship on transparency, performance information use and trust (e.g., Grimmelikhuijsen & Meijer, 2014;Grimmelikhuijsen & Klijn, 2015;Walker et al., 2018). It has also enhanced the methodological repertoire of the field by introducing experimental designs . ...
Preprint
Scholars of public administration are increasingly using experimental research to develop more robust causal inferences and greater methodological capacity. Against this backdrop, we examine the extent to which experimental research has taken hold in the Asia-Pacific region and assess regional capacity. Our review of 30 articles published by scholars based in the Asia-Pacific region in the public administration section of the Web of Science's Journal Citation Reports finds that the regional capacity for experimental research is concentrated in a small number of institutions and strongly supplemented through international collaboration. Topics studied reflect the advent of behavioural public administration. Although progress is being made in reporting experimental designs, much work is needed in the region to bring greater transparency to scholarship. We conclude by encouraging scholars to more robustly implement and report experimental research and by outlining future directions.
... The judiciary and the administration of justice need to generate a certain degree of confidence (Trontti, Harakka, & Pyöriä, 2019). Trust in the judiciary is needed for the voluntary adoption of judicial decisions (Grimmelikhuijsen & Klijn, 2015). Evidence has shown that the citizens' trust in court is not very high, but seems to be adequate in some way (see Table 4.1). ...
Book
Full-text available
Trust in public and political institutions is an indication of democratic governance. The legitimacy of public and political institutions is heavily reliant on citizens’ trust, which is critical for a healthy democracy. In this regard, this study aimed to examine the pattern of citizens’ trust in the public and political institutions of Nepal. The data for this study was taken from Nepal National Governance Survey 2017/18, and analyzed with an ordered logit model. This study shows that citizens’ trust in public and political institutions is not satisfactory. Citizens' trust in political institutions is relatively lower than that for public institutions. Citizens have a low level of trust in ministers, political parties, and parliamentarians. This indicates that citizens are not satisfied with the performance, actions, and behavior of the political institutions. However, citizens have more trust in public institutions, as compared to political institutions. Some public institutions have a relatively high level of trust, while others have it at a low level. Public institutions such as the army and the media have a higher level of citizens’ trust, while police and civil servants/bureaucrats have a lower level of citizens’ trust. It is noteworthy that the citizens' trust in public and political institutions is also influenced by demographics, social associations, and other variables.
Chapter
This chapter seeks to improve our understanding of the critical role trust plays in enhancing good governance. It is generally argued that citizens’ trust in government is dwindling, which in turn poses a serious threat to representative democracy. Studies on trust and its implications for good governance have concentrated mostly on developed countries, with little focus on the developing world or the African continent in particular. This chapter provides an insight into the relationship between trust and other democratic values such as freedom of expression, judicial independence, control of corruption, accountability, and approval of democratic tenets. The increasing need for good governance has brought to the fore the significance of trust that the citizens have for their governments in promoting good governance. The chapter will also discuss the utility of the concept of the new public management in contributing to the discourse on trust and good governance. Using a qualitative research design, specifically the narrative approach, this chapter examines citizens’ trust for government. The study draws examples from Ghana, South Africa, and Bostwana to support the discussions. The study found that the declining levels of trust in Africa can be attributed to some factors, including low approval rate of democracy, lapses in control of corruption, and low levels of transparency and accountability coupled with low levels of freedom of expression. This falling trend calls for the strengthening of institutions to ensure public sector accountability, enhancing the role of the media as watchdogs in the political scene and enhancing the participation of citizens in the governance process.
Article
Full-text available
Confirmation bias, as the term is typically used in the psychological literature, connotes the seeking or interpreting of evidence in ways that are partial to existing beliefs, expectations, or a hypothesis in hand. The author reviews evidence of such a bias in a variety of guises and gives examples of its operation in several practical contexts. Possible explanations are considered, and the question of its utility or disutility is discussed.
Article
Introduction The Psychology of Social Justice Relative Deprivation Is Justice Important To Peoples Feelings And Attitudes? Distributive Justice Procedural Justice Retributive Justice Behaviorial Reactions To Justice And Injustice Psychological Versus Behavioral Responses to Injustice Behavioral Reactions to Injustice Why Do People Care About Justice? The Nature of the Justice Motive When Does Justice Matter? Social Structural Influences Culture
Article
We outline a framework for causal inference in settings where assignment to a binary treatment is ignorable, but compliance with the assignment is not perfect so that the receipt of treatment is nonignorable. To address the problems associated with comparing subjects by the ignorable assignment - an "intention-to-treat analysis" - we make use of instrumental variables, which have long been used by economists in the context of regression models with constant treatment effects. We show that the instrumental variables (IV) estimand can be embedded within the Rubin Causal Model (RCM) and that under some simple and easily interpretable assumptions, the IV estimand is the average causal effect for a subgroup of units, the compliers. Without these assumptions, the IV estimand is simply the ratio of intention-to-treat causal estimands with no interpretation as an average causal effect. The advantages of embedding the IV approach in the RCM are that it clarifies the nature of critical assumptions needed for a causal interpretation, and moreover allows us to consider sensitivity of the results to deviations from key assumptions in a straightforward manner. We apply our analysis to estimate the effect of veteran status in the Vietnam era on mortality, using the lottery number that assigned priority for the draft as an instrument, and we use our results to investigate the sensitivity of the conclusions to critical assumptions.
Article
This article asks how Internet use, citizen satisfaction with e-government, and citizen trust in government are interrelated. We first review the literature on trust and explore how radical information technologies may work to alter the production or maintenance of trust. We then develop hypotheses about how citizens' experience with e-government, satisfaction with e-government and government Web sites, and trust in government are interrelated. Moreover, the model for e-government and Web site satisfaction incorporates citizen perspectives on electronic transaction, transparency, and interactivity. Using data obtained from the Council on Excellence in Government, we then develop and test a two-stage multiple-equation model that simultaneously predicts experience, satisfaction, and trust. Findings indicate that government Web site use is positively associated with e-government satisfaction and Web site satisfaction and that e-government satisfaction is positively associated with trust in government. We also find that while citizens are generally satisfied with the electronic provision of information (transparency), there is some dissatisfaction with the transaction and interactivity of Web sites. We conclude that electronic government strategies-transaction, transparency, and interactivity-are important factors that directly affect e-government satisfaction and indirectly affect trust. Individuals who use government Web sites are not only critical consumers but also demanding citizens.