Is transparency the key to trust in government?

Some empirical studies into the effects of transparency have been carried out (e.g. Tolbert & Mossberger 2006; Cook et al. 2010; De Fine Licht 2011, 2013; Grimmelikhuijsen 2011, 2012; Grimmelikhuijsen et al. 2013) but the results have been mixed: positive effects were found in some cases and evidence for adverse effects of transparency on trust in others. A shortcoming of these studies is that the effect of transparency on trust in government is discussed in general terms and one could argue that this approach flattens out all differences in the population. The mentioned studies do not distinguish between different groups of citizens and therefore cannot identify effects that depend on prior attitudes and the political awareness of citizens. To develop a more sophisticated understanding of the relationship between transparency and trust, we have investigated the moderating effects of certain characteristics of citizens on the relationship between transparency and the perceived trustworthiness of government.

Investigating the moderating effects of prior knowledge and attitude of citizens

We employed two theories from psychology that can provide insights into the way individuals process information. Social psychology teaches us that individuals process information differently if they have certain beliefs or attitudes concerning the topic and the information is related and interpreted in terms of pre-existing knowledge (Bohner 2001).Two specific theories from social psychology can help to understand the moderating effects of prior knowledge and the attitude of citizens:

• Cognitive dissonance theory (Festinger et al. 1956) states that individuals will reduce the stress that comes from contradictory beliefs, ideas, and/or values by avoiding situations and information which would be likely to increase the dissonance. Translated to the relation between transparency and trust in government, this theory provides the basis for the assumption that those who already trust the government will gain more trust ¿ while those with little trust are expected to have even less trust in government.
  
  
• The elaboration likelihood model (ELM) (Petty & Cacioppo 1986) postulates that persuasion through information is a process that not only involves deliberate and active information processing, but can also follow a so-called 'peripheral route' by processing through heuristic and easily recognisable cues. People with prior knowledge on a given topic tend to adopt a more deliberative style of information processing and more strongly relate the information to their existing knowledge. For the moderating effect of prior knowledge on the relation between transparency and trust, this theory suggests that prior knowledge is expected to weaken the effect of transparency on trust, since these people will be influenced more strongly by the knowledge they already possess about the policy topic.

The following research question is central: to what extent does the prior knowledge and attitudes of citizens affect the relation between transparency and trust in government?

Figure 1. Research Model

We approached this question by employing a large scale online experiment (N=570). When testing causal effects, few social science methods can live up to the rigour and level of control of an experimental design. In addition, the large scale of the experiment allows us to distinguish between groups of citizens and this enables us to develop a more sophisticated model of the relation between transparency and trust.

An online experiment

An experiment was carried out over the Internet. We used the design of a government website to give it the look and feel of a real website. The content of the website was then adjusted for the purposes of this study. This resulted in three websites (see Table 1).

Table 1.

Random attribution of participants to these experimental conditions caused an equal distribution of relevant background variables across the experimental groups, such as age, gender, education and political preference. It should be noted that we only presented relatively positive government information to participants.

Findings: moderated relations between transparency and trust

Firstly, we found no effect of transparency on people who had high levels of self-reported knowledge. High or low transparency had no significant effect on the perceived trustworthiness of government in the three dimensions (perceived competence, benevolence, and honesty). Transparency does not affect trust when citizens already have a good knowledge of government.

Secondly, the results for individual with little prior knowledge about government varied. From the three dimensions of trustworthiness taken into consideration (perceived competence, benevolence, and honesty) only competence (Table 2) and benevolence (Table 3) were affected.

Table 2. Group comparison for perceived competence

Table 3. Group comparison for perceived benevolence

Tables 2 and 3 show distinct patterns for participants with little knowledge. Our analysis shows that prior knowledge and general disposition moderate the transparency and trust relation. There is a decline in perceived competence when confronted with limited transparency for people with little knowledge. A rise in perceived benevolence occurs only in one specific group: people with limited prior knowledge and little trust in government in general. This effect disappears when analysing the whole population.

Implications for theory and practice

What does this mean for the debate on transparency and trust? We would like to raise three main points:

• Transparency has no effect on trust in a specific government organisation if the prior knowledge of citizens is high. Based on the elaboration likelihood model, we argue that people with high prior knowledge are less likely to be persuaded by new information. Once (perceived) prior knowledge is high, it becomes the primary driver for perceived trustworthiness of a government organisation and citizens are not 'persuaded' by the outcome information disclosed on websites.
  
  
• For individuals with little prior knowledge, trust is much less based on prior views and new information may change their attitudes. 'Naïve trustees in government organisations lose their trust if the government does a bad job in creating transparency ¿ while naivety has an opposite effect when it comes to perceived benevolence. Individuals with little prior perceived knowledge and a low predisposition to trust government were pleasantly surprised about government benevolence when confronted with a high degree of transparency.
  
  
• These findings can form the basis for better targeted transparency policies. Citizens with little knowledge about government are the most difficult to reach and the most interesting in terms of strengthening trust in government. These citizens may need to be approached more actively and cannot be expected to visit government websites spontaneously: the use social media such as Twitter and Facebook may entice them to access government information. Targeting transparency efforts at specific groups of citizens helps government organisations strengthen trust in their performance and intentions.