A team at Chulalongkorn University recently undertook a research project to examine the factors that increase public participation in anticorruption efforts, so as to develop a more effective communication strategy to promote public participation. (The final paper is currently only available in Thai, though an English translation is in progress, and a summary of the work can be found here.) The study employed a concept used in marketing research called “segmentation,” which seeks to identify latent classes of people—sorted by various characteristics and indicators—who will be more responsive to particular kinds of messaging. In marketing research, the idea is to identify which potential consumers will be most responsive to certain marketing strategies. The same research techniques can be used to classify different segments of the public by their likely responsiveness to anticorruption messaging (or to different kinds of anticorruption messaging).
This research project included two related but separate studies. The first study used a nationwide survey, with 720 participants evenly distributed by age group and region of residence, to identify individuals who were most likely to get involved with anticorruption activities. The survey sought to measure four aspects of each respondent’s interest in, and commitment to, anticorruption: (1) the respondent’s awareness that corruption is a problem; (2) the respondent’s personal commitment not to engage in corrupt activities; (3) the extent to which the respondent shares information about anticorruption with others in their communities and urges them to be honest (“communication and prevention”); and (4) whether the respondent reports corruption that they witness and actively participates in anticorruption activities (“active engagement”). Each of these dimensions was evaluated separately and then averaged into an aggregate measure of “anticorruption orientation.”
The survey also included questions about respondent’s demographic characteristics (such as age, gender, occupation, and income) as well as questions designed to measure latent psychological traits. The researchers then identified the factors that are most correlated with a strong anticorruption orientation—much as traditional market researchers use similar techniques to segment the customer base so as to identify those consumers with the strongest potential demand for the product. Interestingly, the factors that were most useful in identifying individuals with a strong anticorruption orientation were not demographic factors, but rather a certain psychological profile, which only weakly correlated with traditional demographic characteristics. In particular, anticorruption orientation was strongest among those who (1) report higher feelings of personal guilt from paying bribes; (2) believe that they have more control over the outcomes of events in their lives; (3) tend to abide strictly by guidelines; (4) are more “collectivist” as opposed to “individualist” (that is, give the welfare of the group priority over the individual); and (5) advocate higher societal equality (both income equality generally, and gender equality specifically). In fact, the results of the survey suggested that the population can be sorted into roughly four segments, which the researchers labeled the Frontline (~17% of respondents), the Exemplar (~28%), the Mass (~45%), the Individualist (~10%). Frontline and Exemplar respondents have similar overall levels of anticorruption orientation, though the former category scores higher on the “active engagement” dimension, while the latter scores higher on the “communication and prevention” dimension. The other two segments have significantly lower average anticorruption orientation, though their psychological profiles differ somewhat.
However, while this first study was able to identify the psychological traits that correlate with strong anticorruption orientation, that information alone is not sufficient to tell us what kinds of messaging strategies are likely to have the greatest impact in mobilizing these individuals to participate in anticorruption activities. The second study therefore used a lab experiment, on a sample of 240 university students, to see whether messages with different tones and framing would have different impacts.
Each subject first read an introductory message, followed by a series of questions regarding how the respondent would behave in different situations. The questions were designed to elicit responses along the same four dimensions used in the first survey—subjects were asked, for example, about whether some kinds corrupt activities were acceptable, whether they personally would be willing pay bribes under certain circumstances, whether they would spread anticorruption messages in their community, and whether they would report corruption cases. But the subjects were randomly given with introductory messages. All of the messages had similar content—which included not only instructions on completing the survey, but also some anticorruption messages—but they differed in tone and framing. Half of the participants received introductions that included anticorruption messages expressed in a serious authoritative tone—with sentences like “you need to take corruption problem seriously!”, “if we don’t fix this now, our nation will fail!”, “we need to punish the culprits, if not, our society will fail,” and “participants in this study must strictly follow the rule at all times!”. The other half of the participants received messages with softer tone, with words like “please”, “thank you”, and “let’s do this together,” and with sentences like “corruption is a problem but there are solutions” and “we should work together to solve it and develop our country.” Two weeks later all the same subjects were asked to participate in the experiment again, but this time each subject received a message that was the opposite of the one they got on the first time.
Interestingly, and perhaps counterintuitively, the players who received less intense introductory messages, with a less authoritative tone, displayed a higher average level of anticorruption orientation, especially on the active engagement dimension. Additionally, those who got the serious, intense introductory message in the first round and then the opposite in the second round tended to show greater average anticorruption orientation in the second round. The reverse was also true—participants who had received a softer version of the message in the first round, and the more intense version the second round, exhibited lower anticorruption orientation the second time they completed the survey questions than they had previously. This finding suggests that anticorruption messages with a less serious and intense messages with less authoritative tone are more likely to persuade participants to actively engage in anticorruption activities.
These findings have implications for the design of effective communication strategies to encourage more public participation against corruption. Most anticorruption agencies, especially those in developing countries, do not have abundant resources for marketing and advertisement, which makes it important for campaigns to target the segment of the population that could be more easily persuaded to become more actively engaged in anticorruption initiatives. Anticorruption messages should be tailored to those who already have a high baseline level of anticorruption orientation so that it has the highest marginal impact—that is, to maximize the increase in anticorruption orientation. In the case of Thailand, it appears that Exemplar segment is the most likely to show increased anticorruption orientation in response to messaging, so the focus should be on tailoring messages to appeal to that segment. Perhaps even more importantly, the mood and tone of messages must be carefully designed to maximize their persuasiveness to the target population. One might have supposed that messages that use powerful, authoritative, moralistic rhetoric would be most effective in mobilizing individuals to care about corruption and help to fight it. But at least in the Thai context, the research described above suggests that the messages that are most likely to be effective—at least for a substantial segment of the persuadable population—are those that eschew of serious, intense, and authoritative rhetoric, and instead use softer and encouraging emotional appeals.