Data Loading...
The Effect of Transparency in Decision Making for Public ... Flipbook PDF
1 The Effect of Transparency in Decision Making for Public Perceptions of Legitimacy in different Policy-areas Jenny de
115 Views
26 Downloads
FLIP PDF 335.57KB
The Effect of Transparency in Decision Making for Public Perceptions of Legitimacy in different Policy-areas
Jenny de Fine Licht Ph.D. student University of Gothenburg Sweden
Abstract Building on the findings of recent empirical studies that transparency in political decision making may affect public legitimacy beliefs negatively, this article suggests that transparency has different effects depending on policy area. Specifically, transparency is hypothesized to be less effective in the case of policy decisions that involve trade-offs relating to questions of human life and death or well-being. Using an experiment that involves 1,032 participants, the effect of transparency is tested in two policy areas that represent a case of routine priority setting (culture and leisure) and a case of policy decisions implicitly relating to human life (traffic security). Results indicate that transparency has the power to increase public acceptance of political decisions, but this effect is moderated by the type of policy area involved. Furthermore, a limited type of transparency in which decision makers provide justifications for their decisions can result in benefits while avoiding potential costs.
1
Introduction The hypothesis—or even assumption—of a positive relationship between increased transparency in political decision making and public legitimacy and trust has a long history in political philosophy and is well grounded in normative accounts on how people should to evaluate political decision making in a democracy (e.g., Heald 2006; Hood 2006). However, the empirical support for such a relationship is not only scant but also contradictory (e.g., Grimmelikhuijsen 2012a for discussion), with several studies showing null or even negative effects (e.g., Bauhr and Grimes 2013; de Fine Licht 2011; Grimmelikhuijsen 2010; 2012a; Worthy 2010). The question on why transparency does not seem to be the obvious prescription against declining public trust and legitimacy that democratic theories suggest has recently received increasing scholarly interest. Empirical studies have, for example, showed that the initial level of trust and knowledge affects how people experience transparency (Grimmelikhuijsen and Meijer 2012), that people from different cultures tend to react differently to transparency (Grimmelikhuijsen et al. 2013), and that there is a translation problem between actual decision-making procedures and public perceptions of procedures that makes transparency less efficient in improving the public’s legitimacy beliefs (de Fine Licht forthcoming; see also Esaiasson 2010). In line with efforts to increase our understanding of transparency effects, this article makes a first test of whether transparency has different effects in different policy areas. Inspired by the theory of taboo trade-offs—which argues that trade-offs between sacred values such as human life and secular values such as money are particularly difficult for people to accept—the article suggests that one reason for why transparency fails in empirical test is that studies have tended to focus on policy areas that handle issues related to human wellbeing, such as health care and air quality. Building on data from an original experiment that involves 1,032 participants, the analysis examine the effect of increased transparency for public perceptions of legitimacy is evaluated in two different policy areas: one that involves political choices relating to human life and well-being (public spending on traffic security) and one that involves routine-like decisions (public spending on culture and leisure). In each policy area, three degrees of transparency are manipulated. The article’s contributions are both theoretical and practical. Its theoretical contribution lies in enhanced understanding of how transparency effects might be conditional, which cast light on the diverging results in recent empirical studies. Its practical contribution lies is providing further insights into how transparency reforms can affect public acceptance and trust. As difficult policy decisions where important values have to be traded off against each other are not likely to be fewer in the future, this is important knowledge for public administrators and political decision makers.
Transparency as a driver of perceived legitimacy Transparency in political decision-making—defined as information about decisions and decisionmaking procedures that is provided or available to the public—is generally regarded as a golden 2
tool in policy making; as increasing or causing a wide range of positive qualities such as economic efficiency, accountability, and anti-corruption (e.g., Heald 2006). This article explores one of the most commonly mentioned benefits of transparency, namely that it will positively affect public perceptions of political decisions and decision makers (e.g., Heald 2006; Roberts 2006). In doing this, the article will use perceived legitimacy as an overarching term encompassing public trust for authorities’ handling of a policy area as well as public willingness to accept decisions and decision-making procedures. These beliefs will all serve as indicators that the public perceives authoritative decision making to be rightful and worthy of consent and obedience; i.e., that it is legitimate (see e.g., Levi, Sacks, and Tyler 2009 for discussion). This means that the internal relations between trust and acceptance will not be explored; i.e., the article does not take a stand on whether trust causes acceptance or acceptance causes trust. Further, ―perceived‖ is included to highlight that the article’s focus is empirical and not normative. This means that transparency might very well make decision making more legitimate (in a normative sense) without the public experiencing it as such. The literature on transparency has identified several reasons for why increased transparency can have a positive influence on public perceptions of legitimacy. For example, transparency may increase people’s sense of control by making the decision makers accountable for their actions and may increase the public’s understanding of decisions and the decision makers (see de Fine Licht et al. forthcoming, for an overview). However, research has also identified several potential problems that result from increased transparency which may reduce the presumed positive effect. For instance, increased transparency may induce public disappointment with regard to how political decision making is actually conducted (Grimmelikhuijsen 2010), or may lead to information overload and confusion (O’Neill 2006). Further, decision makers may need relatively closed doors to do their work efficiently. For instance, they may need to say things that can be considered as fraternizing with the enemy, or they may need to explore alternatives that they will later on repudiate (Mansbridge 2009). The so far limited empirical studies on this topic have found support for both a positive (e.g., Cook, Jacobs, and Kim, 2010; de Fine Licht et al. forthcoming; Tolbert and Mossberger 2006) and negative (e.g., Bauhr and Grimes 2013; de Fine Licht 2011; Grimmelikhuijsen 2010; 2012a; Worthy 2010) relationship between transparency and legitimacy beliefs such as decision acceptance or trust. Overall, the skeptical position or the belief that transparency has hardly any effect on trust or trust-related measures (Grimmelikhuijsen and Meijer 2012) seems to have the most support, which indicates a need to explore potential moderators of transparency effects.
Policy area as a potential moderator of transparency effects As transparency effects are highly context dependent (de Fine Licht et al. Forthcoming), addressing all possible conditions that can affect transparency’s relationship with perceived legitimacy is not possible. Nevertheless, this article will explore one potential moderator, namely the type of policy area. This suggestion is based on an observation that experimental studies that have shown negative effects of transparency have tended to focus on policy areas that are clearly related to public health and well-being. For example, de Fine Licht (2011) uses priority setting in public health care 3
as a case, i.e., decisions on which medical treatments should be publicly funded and which are too expensive compared with other needs. de Fine Licht (2011) finds that transparency does not affect acceptance of decisions or decision-making procedures and that transparency, in fact, generally results in a negative effect on trust in health care. Similarly, Grimmelikhuijsen (e.g., 2010; 2012a) uses a public policy concerning local air quality as a case and finds significant negative effects of transparency on indicators of trust in local governments, such as perceived competence of authorities. This article argues that the explicit or implicit trade-offs that are typically made in these policy areas offer one potential explanation for why transparency has resulted in negative effects on perceptions of legitimacy. The rationale behind this suggestion is based on the theory of taboo trade-offs (e.g., Fiske and Tetlock 1997; Tetlock 2000; 2003; Tetlock et al. 2000), which proposes that decisions relating to human life and well-being that involve monetary considerations are especially difficult for people to handle. According to the theory, human life and wellbeing is, like love, friendship, and honor, considered to be ―sacred,‖ and cannot be traded off against values that are ―secular,‖ such as money. Trade-offs that violate this taboo will be resisted not only because they are difficult to make, but also because they are considered strange, unthinkable, immoral, or ridiculous to even consider (Fiske and Tetlock 1997). The implication for transparency theory is that if people perceive political decisions as trade-offs between human lives or wellbeing and money, increased transparency that makes these choices more salient may trigger negative feelings toward decisions and decision makers. Tetlock et al. (2000) provide a test of the theory of taboo trade-offs that shows how taboo tradeoffs differ from other trade-offs. They differentiate between routine trade-offs (e.g., choices between secular values, such as money and consumption goods), tragic trade-offs (e.g., choices between values that are considered sacred, such as human lives), and taboo trade-offs (e.g., choices between values considered to be secular, such as money, and values considered to be sacred, such as human lives). In an experiment, they manipulated whether a director of health care management had a choice of saving one of two very sick boys with the only liver available (a tragic tradeoff) or a choice between saving a sick boy by means of a liver transplant and saving a large amount of money for the hospital (a taboo trade-off). Furthermore, the authors manipulated whether the Director saw the decision as an easy one or if he gave it much thought and contemplation. Finally, they manipulated whether the director decided to save the boy named Johnny or not. The results of the experiment showed that in the taboo trade-off case, the participants tended to punish the director for choosing money over Johnny after giving the case a thorough thought, whereas the director was judged least negatively when he quickly chose Johnny. In the tragic trade-off case, the director evoked a low level of outrage regardless of whether he chose to save Johnny or the other child when he contemplated on the case (Tetlock et al. 2000). These findings indicate that tragic trade-offs differ from taboo trade-offs in peoples’ minds. The decision maker who thinks a long time on whether to choose the life of a child or a large amount of money was judged as morally suspect regardless if he finally made the ―right‖ decision (i.e., to save the child). In a tragic trade-off, in which only one life can be saved, the decision maker who carefully contemplated on the case was judged as showing respect for the difficult nature of the decision. A 4
choice between two sacred values (e.g., children’s lives) is indeed tragic, but people can accept the fact that a choice has to be made. Contrary, the choice between a sacred value (a child’s life) and a secular value (money) is not ―allowed.‖ How does earlier research on transparency’s empirical effects relate to the reasoning on taboo trade-offs? In the examples mentioned above—priority setting in health care and air quality—no explicit comparisons between sacred and secular values are made. This means that they are no perfect examples of taboo trade-offs. Especially the priority setting case (de Fine Licht 2011) could actually be seen as a tragic trade-off, i.e., where two potentially sacred values (health needs) are compared. However, contrary to the tragic trade-off earlier discussed where there was simply one liver to allocate (Tetlock et al. 2000), decisions on which medical treatments that should be available is a consequence of scarce financial resources. As people generally are very reluctant to believe that there are not money enough to finance all medical needs (e.g., Syrett 2007, 6), these decisions are likely to be perceived, at least to some extent, as a choice between money and human well-being, i.e., as taboo trade-offs. The same reasoning can apply to policy decisions regarding air quality (Grimmelikhuijsen 2010; 2012a). Aside from the fact that environmental protection is potentially classified as a sacred value for many people, it has well-known effects on human health. Therefore, politicians involved in air quality policy might be perceived as stingy actors who are not willing to provide sufficient resources to secure the environment and public health, i.e., trading money for human lives and well-being. In sum, the theoretical insights from the theory of taboo trade-offs may add understanding to the findings in previous research which note that transparency undermines perceived legitimacy. We can hypothesize that transparency will have zero or even negative effects in policy areas that involve trade-offs where human life and well-being are at least implicitly weighed against money. In cases of more routine like trade-offs, transparency will, on the other hand, have a positive effect on perceived legitimacy, given, of course, that reasonable arguments are provided. The empirical part of this article will provide a first test of how transparency affects perceived legitimacy in different policy areas.
An empirical test of transparency effects Experimental design In order to explore whether the nature of the policy area moderate transparency effects, this article uses an experimental approach to explore the effect of transparency in decision making on perceived legitimacy in two different policy areas. Although experiments are by definition manipulating the reality, they allow us to isolate and explore causal effects of interest in a way that other methods cannot. This means that we can get indications of what effects specific aspects of a proposed reform will bring, that can later be explored in the field. Inspired by the theory of taboo trade-offs, one policy area—decisions regarding public spending on local culture and leisure activities—was chosen to represent an area of rather unproblematic 5
decisions from a moral perspective. Although people may have highly diverging interests and opinions on how common resources should be spent, people are likely to accept trade-offs in general and are not morally offended by the fact that priorities need to be made. The other policy area—decisions regarding public spending on traffic security—was chosen to represent a case where policy decisions typically include elements of taboo trade-offs. As questions regarding public spending on traffic security involve calculations of how many accidents that can be avoided by a certain investment, these decisions clearly involve considerations where money is—as in the health care case—implicitly traded against human lives and well-being. However, as it is likely that people are more used to the idea that security is something that can weighed against its cost than to the idea that access to certain kinds of health care can be denied, traffic security should probably be seen as in between health care priority setting and culture and leisure investments; at least given that no exceptional descriptions of traffic victims are not provided. The actual decisions that were presented to the participants were fictitious but inspired by real stories in local newspapers. In the culture and leisure case, a municipality decided to close a local library branch to increase financial support to local non-profit associations. In the traffic security case, the municipality decided not to support the funding of road dividers on a local road and instead invest in the construction of a safe crossroad in the city. As transparency should probably not be understood in terms of a binary opposition—as something that is either present or not (Florini, 1998)—three different degrees of transparency were manipulated in each policy area. While transparency proponents sometimes tend to assume that ―the greater the transparency, the better,‖ Mansbridge (2009) suggests that full transparency may not the best practice in policy making. Instead, a limited amount of transparency that focuses on decision makers providing reasons for their choices can be a good alternative when transparency may have costs, such as preventing decision makers from acting in an effective way. In such a case, ―we should favor not extreme transparency in process (for example making all committee meetings public), but instead transparency in rationale—in procedures, information, reasons, and the facts on which the reasons are based‖ (Mansbridge 2009, 386. Emphasis in original). Inspired by Mansbridge’s (2009) conceptualization, the transparency in process condition consisted of a rather lengthy description of the decision-making process that explained both the formal procedure and the deliberation process among decision makers where they presented their arguments for different positions. The less transparent condition, transparency in rationale, consisted of a justification for the decision. Finally, in the condition of no transparency, only the final decision was announced, and no justification or procedural description was provided. Although a shorter piece of information can indeed be easier to grasp and therefore more comprehensible, transparency in process is here considered more transparent than transparency in rationale as it includes the arguments presented by the decision makers and adds procedural information. The information about the policy decisions was presented to the participants in the form of newspaper articles. This means that the experiment manipulated mediated transparency (Meijer 2009), and not a situation where citizens actually observed the decision-making situations themselves. In the condition of no transparency, the information resembled a short news telegram that simply reported the decision. In the condition of transparency in rationale, the text resembled a 6
news note that reported the decision and justifications. Finally, in the condition of transparency in process, the participants read a full article that presented the decision and a part the deliberation among the decision makers, including their arguments and justifications. The articles were written with the guidance of professional journalists. They strive to present a neutral description of the justifications and, in the condition of transparency in process, the deliberation among decision makers. For example, no interviews with affected citizens are included, and the articles do not take a position on the validity of the arguments. The justifications presented were the same in the conditions, although the justifications were expressed in plain text in the condition of transparency in rationale and in the form of statements from politicians in the condition of transparency in process. Justifications were based on the arguments advanced in real life debates of these issues. For example, the justifications in the culture and leisure case included references to a declining number of visitors for the library and the important work for the youth and the elderly carried out by nonprofit associations. The decision in the traffic security case was motivated by reference to a comparatively small number of commuters using the road in contrast to the large number of people, especially pedestrians, put at risk on the city crossroad. The description of the decision-making process (in the transparency in process condition) included details on the various stages of the procedure, including a professional stage in which the representatives for different functions made their priorities, and a political stage in which politicians made the final decisions. Moreover, the articles stated that the minutes of the meetings were available on the Internet, and that an evaluation of consequences was scheduled. Although the same structure was used for the articles in both policy areas, they differ for obvious reasons in terms of the particular arguments used. Therefore, the effect of transparency will be analyzed separately in the condition of culture and leisure and the condition of traffic security, and no statistical comparisons between the policy areas will be made.
Experimental procedure 1032 subjects participated in the experiment. The participants were drawn from an electronic panel of Swedish citizens which is available from the MOD group at University of Gothenburg. This panel is self-selected which means that frequencies, such as the mean value for various indicators of perceived legitimacy, cannot be generalized to the larger population. For experimental purposes, in which the effects of different treatments are in focus, the sample is, however, suitable. Compared to the Swedish population in general, participants in the panel are highly educated and men are overrepresented (c.f., Martinsson, Andreasson, and Petterson 2013). The demographic data of the participants are presented in table 1.
7
Table 1: Demographics Gender (%)
Age (years)
Level of education (%) High school
Male Female Mean Min Max not completed
Graduated from
University degree
high school but
or PhD
no university degree 58.5
41.5
48.05
19
86
7.8
41.7
50.5
Note: ANOVA showed no significant differences between the experimental groups when it comes to how often subjects indicated that they visited a library, how many association they were active and passive member of, interest in cultural issues, sex, education, political orientation from left to right, general trust in politicians, and general interest in politics (for culture and leisure), and how often subjects indicated that they go by car, how often they go by bicycle, how often they go by public transport, interest in traffic issues, sex, education, political orientation from left to right, general trust in politicians, general interest in politics (for traffic security).
The participants were invited via e-mail to join the experiment and were, upon acceptance, randomly assigned to one of the two policy areas (culture and leisure or traffic security) and within that policy area, to one of the transparency conditions (no transparency, transparency in rationale, or transparency in process). They were asked to read an article about a policy decision in an unnamed municipality, and thereafter, to answer a series of questions on the decision and the decision-making process. In the end, the participants were informed that the decision was fictitious and were thanked for their participation.
Measures We will analyze the effect of transparency for a number of indicators of perceived legitimacy that are often used in theoretical as well as in empirical research on legitimacy and trust. First, we will evaluate public reactions to the decision making procedure, i.e., how the decision was made. Evaluations of the fairness of procedures have shown to be strongly correlated with assessments of how decisions and decision makers are evaluated (e.g., Tyler 2006), and is frequently used as an indicator of beliefs in legitimacy in empirical research (e.g., Persson, Esaiasson, and Gilljam Forthcoming). In this study, Procedure acceptance was measured by an index that consisted of the following three questions: What do you think of how the decision was made?; How fairly do you think the decision was made?; and How fair do you think you, as a citizen, were treated when the decision was made? Answers were given on a scale from 1 (very bad/not fair) to 7 (very good/very fair). Cronbach’s alpha for the index was 0.924 for culture and leisure and 0.936 for traffic security. Secondly, we will evaluate public perceptions of the actual policy decision; the outcome of the procedure. For example Tyler (1997, 2006) connects legitimacy with voluntary willingness to obey and accept authoritative decisions, which means that a stated willingness to accept a decision can be treated as an empirical indicator of legitimacy. This study measured Decision acceptance by an index that consisted of the following three questions: What do you think of the decision to close a 8
library branch to invest in increased funding for associations?/What do you think of the decision not to support the funding of road dividers and instead investing in the construction of a safe crossroad?; How fair do you think the decision is?; and How willing are you to accept the decision? Answers were given on a scale from 1 (very bad/not fair/not willing) to 7 (very good/very fair/willing). Cronbach’s alpha for the index was 0.928 for culture and leisure and 0.899 for traffic security. As decision acceptance measure a rather passive reaction to a political decision, the self-reported likeliness of protesting against the decision was also included as a separate measure: How likely do you think it is that you will protest against the decision, for example, by writing a letter to an editor or by contacting a politician? Answers were given on a scale from 1 (not likely) to 7 (very likely). Finally, trust is frequently associated with legitimacy or voluntary compliance (e.g., Levi 1997) and is also used as an operationalization of legitimacy in empirical studies (e.g., van der Toorn, Tyler, and Jost 2011). While acknowledging that there is a debate on whether trust is a single or multidimensional construct (e.g., Grimmelikhuijsen 2012a, Kim 2005), this article opted for a single question measure of trust in the general management of the policy area: How much trust do you have in the work for local public culture and leisure promotion in Sweden?/How much trust do you have in the work for traffic security in Sweden? Answers were also given on a scale from 1 (no trust) to 7 (high trust).
Stimulus control The stimulus control of the transparency manipulation consisted of the question: To what extent do you feel that you were informed about how the decision was made? ANOVA showed a significant main effect of transparency in both the policy area of culture and leisure (F2,486=19.354, p