1What is the role of experts and the value of expert advice in the decision-making process? The world is becoming increasingly complex, requiring decision-makers to make increasingly quick decisions. To make these policy decisions, individuals weigh decision options based on the availability of information as well as their interests, preferences, and beliefs. Expertise is said to play a particularly important role as part of the ‘information bank’ that factors into policy-making decisions. However, is expertise actually considered valuable and, more importantly, is it heeded by policymakers?
2Previous research has stipulated that the perceived value and willingness to adopt expertise depends on a variety of factors. Studies have focused on the content, context, and/or perceived trustworthiness of expert advice as factors weighing on whether decision-makers adopt it. However, there is limited scholarship on investigating whether the expert’s personal attributes/characteristics influence how decision-makers value this advice and whether they are willing to adopt it. This exploratory study seeks to understand the extent to which an expert’s gender and national identity influence how valuable policy decision-makers consider expert advice (and are willing to adopt it) in the Armenian policy context.
3Research posits that implicit bias influences prejudicial judgment, decisions, and behaviors without conscious intent. People act on the basis of internalized schemas and preconceptions of which they are unaware (Pritlove et al., 2019; Boroch, 2024). Implicit bias is not always psychological but rather a social phenomenon that affects everybody. Stereotypes (gender, ethnic, racial, etc.) ‘feed’ implicit biases and arise from historical and cultural understandings that may be outdated but nonetheless extremely powerful. For example, gender stereotyping, which restricts gender to certain roles and levels of authority based on preconceived notions of femininity and masculinity, can influence behaviors and meaning-making (Heilman & Caleo, 2018). The same applies to ethnic and/or national stereotypes, which are a set of compounding preset beliefs that compartmentalize members of a given community based on certain traits, status, and cultural characteristics. The overarching question, therefore, is whether gender and national stereotyping influences how people consider and adopt advice in public policy decision-making.
4Examining the lack of gender diversity in the legal profession, Justin D. Levinson and Danielle Young (2010) find that implicit gender bias does influence the continued discrimination and subordination of female lawyers. However, their study also found that subjects are increasingly aware of such biases and are able to resist making discriminatory decisions influenced by gender (female) stereotyping. Such hope in overcoming implicit gender biases quickly diminishes when examining the prejudices women face in politics. Politics and political decision-making seem rife with “sexism” (Deber, 1982). For example, Cecile H. Mo finds that “implicit gender attitudes impact vote choice” (Mo, 2015, p. 389). Put simply, female candidates need to be perceived as more qualified than their male counterparts in order to earn the support of voters and actually win elections. This uphill battle women face in politics is further elaborated by Donelson R. Forsyth, Michele M. Heiney, and Sandra S. Wright whose study shows that gender stereotypes cause women leaders to be “often evaluated more negatively than men leaders” (Forsyth et al., 1997, p. 101). The reason for such a negative evaluation of female leaders is attributed to implicit gender bias whereby the characteristics of leadership seemingly conflict and are incongruent with the “conventions concerning appropriate female behavior” (Eagly et al., 1992, p. 5).
5It must be recognized, however, that a significant amount of research argues against the gender bias narrative. Barbara C. Burrell finds that American attitudes towards women as political leaders are much more nuanced than previously thought. She argues that scarcity, rather than prejudice, is to blame for the lack of women in politics in the United States (Burrell, 1996). Similar sentiments are shared by Richard A. Seltzer, Jody Newman, and Melissa V. Leighton who seek to debunk the myth of female discrimination in U.S. politics. In their text, Sex as a Political Variable, the researchers argue against the “myth” of a so-called gender gap in politics, claiming that female political candidates have equal opportunities and face similar challenges as their male counterparts when running for public office (Seltzer et al., 1997). Such conclusions are further corroborated by quantitative studies focused on election survey data that suggest that candidate sex alone has little to no effect on election outcomes (Darcy & Schramm, 1977).
6The existence of implicit national bias is perhaps less contentious than the existence of gender bias. In fact, the nationalist position makes implicit bias explicit. The very definition of nationalism implies “special obligations” to conationals that should affect all decision-making and behaviors of a political community (Wellman, 2003, p. 267). Such national, ethnic, and even racial favoritism (if not outright prejudice) plays a controversial role in identity politics. Mental associations about specific social groups influence how people come to perceive the ‘other.’ This “othering” is important from the perspective of forming national identity – separating the ‘we’ (ingroup) from the ‘they’ (outgroup) for the sake of maintaining distinctiveness and galvanizing a political community (Traindafyllidou, 1998). Further empirical studies from the judicial policy arena also suggest the persistence of implicit racial biases. For example, Justin D. Levinson (2007) found that racial biases influence how judges and jurors interpret information, stories, and evidence. For example, white judges are more inclined to remember aggressive facts about an African-American story character than a Caucasian story character, thereby confirming the existence of implicit biases.
7There is a temptation to believe that expertise is dead. In his aptly titled book, The Death of Expertise: The Campaign against Established Knowledge and Why It Matters, Tom Nichols argues that there is a new obsession with the worship of ignorance. According to Nichols, there is an evident trend in which opinions trump expert advice. “To reject advice of experts is to assert autonomy” (Nichols, 2017, p. x). Highlighting the paradox that a stubborn public ignorance is becoming all the more prevalent in the midst of the Information Age, Nichols argues that the ideal of expertise is dying. The relationship between experts and citizens in a democracy is eroding, and with it the conversation between the two parties. While democracy inherently challenges orthodox beliefs, this growing distrust of intellectual authority and intellectual self-reliance has possibly gone too far. However, for Nichols, the biggest issue is the change in public discourse which is not “indifference to established knowledge; it’s the emergence of a positive hostility to such knowledge” (Nichols, 2017, p. 20).
8Christoph Kehl, Steffen Albrecht, Pauline Riousset, and Arnold Sauter reaffirm the notion that the role of expert advice is increasingly being questioned. They point to the growing polemics surrounding what constitutes trustworthy scientific advice. Studying the governing role of the Office of Technology Assessment at the German Bundestag (TAB), these authors observe a general shift “towards a so-called post-truth regime which leads to mistrust in science and erosion of established democratic values” (Kehl et al., 2021, p. 12). They conclude that to effectively operate, an expert-driven agency must first navigate the increasingly questioned authority paradox.
9Expertise alone no longer constitutes authority, nor does it guarantee public adherence to expert advice. Instead, establishing the trustworthiness of expertise comes before invoking expert authority (Geiger, 2020; Cairney & Wellstead, 2020; Bennett, 2020). Nathaniel Geiger investigates whether people perceive expert sources as more trustworthy and knowledgeable than nonexpert sources. The findings are nuanced and demonstrate differences across policy areas. Geiger’s study reveals that experts are generally perceived as more knowledgeable but cautions that people consider experts to be equally trustworthy as nonexperts. Therefore, being a perceived expert on a topic means very little in terms of perceived trustworthiness and ultimately, the ability to sway public opinion on a certain issue. To address this concern, Paul Cairney and Adam Wellstead stress the need for trust at all levels of the policymaking process and between policymakers, experts, and the public at large. Comparing the US and UK Covid-19 policies, these researchers stress the vital role of (political) culture in influencing the extent to which expert advice is adopted (Cairney & Wellstead, 2020, pp. 5-7).
10However, a general political culture that nurtures trust may not be enough. Matthew Bennett argues that a specific type and level of trust is needed for expert advice to be adopted. Bennett believes that adopting expert advice requires more than mere trust in specific expertise. In fact, adopting expert advice requires action on the part of the consumer. Hence, Bennett stresses the need to cultivate and support “recommendation trust” (Bennett, 2020). His cynicism towards expert adoption lies in the difficulty of nurturing this type of recommendation trust which not only requires a transparent relationship between the expert and the public , but also an expansion of the expert’s role to include convincing the public that it is in their rational self-interest to heed specific advice.
11Trustworthiness of expert advice is important, as is the context in which expertise is divulged and the language (actual content) experts use (Islam, 2010; Gerblinger, 2021). Drawing from his experience advising the Bangladeshi government, Nurul Islam emphasizes the importance of distinguishing between solicited and unsolicited advice. He stresses that the private versus public domain (setting) through which expertise is divulged factors into whether policymakers accept expert advice. Sound expert policy analysis can sway policy decision-making, but only if it is shared in an appropriate setting. However, the setting in which advise is shared is only part of the puzzle, as Christiane Gerblinger discovered when she turned her attention to the language experts use. While it is tempting to believe that experts are being ignored and branded untrustworthy, Gerblinger argues that the problem of adopting expertise lies in the policy language used by experts, which avoids political context and employs strategies of impersonality (Gerblinger, 2021). Simply put, expert advice is not adopted due to a perceived lack of value to policymakers, rather than a lack of trustworthiness. Examining expert advice in various Australian policy arenas, Gerblinger finds an increasing disconnect between how experts construct advice and how the public processes it. In other words, the repeated uncertainty that characterizes the content of expert advice, as well as the lack of deeper expert insights, makes such ‘advice’ nearly unusable to layperson policymakers. According to Gerblinger, to solve this problem, changes must be made to how expert advice is communicated in order for it to be more readily adopted.
12While the level of perceived trustworthiness, the nature of expert language, and the setting in which advice is divulged influence the extent to which expertise is adopted, the element of (expert source) culture is yet another possible factor to consider in this complex equation. If culture refers to commonly held beliefs and values within communities, public policy that takes into account these value systems is perceived to be more legitimate. Culture does not determine the outcome of policy, but it does affect the legitimacy of policy success, which transcends performance metrics, since acceptance is as important as impact (Muers, 2018, p. 9). But cultural considerations are not limited to the actual policy itself; they can also be ascribed to policymakers and policy advisers who contribute to these processes. How important is the cultural background of the advisor (source of expertise) in influencing policy adoption? Not very, according to Ryan E. Carlin, Roberto Gonzalez, Gregory J. Love, Daniel Andres Miranda, and Patricio D. Navia. Investigating whether ethnicity (as a sense of belonging to a shared culture) influences levels of trust more than shared policy preferences, Carlin et al. found inconsistent evidence that coethnicity improves intergroup trust (and, ultimately, the increased likelihood for expertise adoption). Instead, policy discrimination has more to do with whether a policy stance is shared rather than the actual ethnic background of the policy advisor (Carlin et al., 2021, p. 17). These findings carry promise for all those who believe that expert content actually matters more than the cultural background of the expert when deciding whether to accept policy advice.
13While an expert’s culture/ethnic identity may influence whether policymakers adopt policy advice, another element that possibly factors into whether expertise is adopted is the gender of the expert source (Neal, 2014; Greve-Poulsen et al., 2021; Ribeiro et al., 2020). Studying women as expert witnesses in the US judicial system, Tess M.S. Neal finds that male experts [compared to female exerts] may have a slight advantage overall in terms of perceived credibility and persuasiveness to jurors” (Neal, 2014, p. 176). Neal contends that while evidence is still rather weak and also issue/context specific, there is some indication that an expert’s gender affects how advice is perceived and adopted. Gender role expectations are especially important in maintaining perceived expert credibility, even though women have the added task of appearing likeable. Neal, therefore, concludes that “while gender differences in how men and women experts may be perceived do exist, most of these differences are relatively small and contextually determined,” and can be overcome by modulating behaviors (Neal, 2014, p. 177). However, Vinicius Farias Ribeiro, Adriana Victoria Garibladi de Hilal, and Marcos Goncalves Avila are far more convinced that expert gender plays a role in advice adoption. Ribeiro et al. seek to identify the circumstances in which gender and advice justification influence advice-taking in the Brazilian business setting. Their findings suggest that analytic justification is valued more than intuitive justification, and that this type of justification favors male advisors (Ribeiro et al., 2020, p. 10). These authors, however, do contend that the gendered cultural perspective that male values, rather than male gender itself, might be more relevant in explaining why male expertise is perceived as more credible and trustworthy that female expertise in policy adoption.
14Conversely, Katrine Greve-Poulsen, Frederik K. Larsen, Rasums T. Pedersen, and Erik Albæk (2021) challenge this idea and find that gender differences among experts actually do not play a role in how the public perceives expertise. Investigating how the public perceives expertise in the Danish media, these researchers found that gender imbalances persist only in quantity and not in quality. While the number of male experts far exceeds that of female experts, the public does not actually “perceive the competence of male and female experts differently” (Greve-Poulsen et al., 2021, p. 17). According to Greve-Poulsen et al., experts do persuade media audiences to varying degrees of success, but this ability has little to do with the expert’s gender.
15While numerous studies have reaffirmed the idea that the adoption of expert advice depends on the perceived level of trustworthiness, style of expert language, and the context in which expertise is shared. However, scholarship is more nuanced (and conflicted) regarding the extent to which an expert’s cultural background and gender factor into this dynamic as implicit biases that unconsciously impact perceptions of expert advice. Most research on this topic is confined to the medical and judicial arenas of decision-making processes. This exploratory study adds to this body of knowledge by investigating whether and how these two variables of gender and national identity impact public policy decision-making more generally. This preliminary research, therefore, focuses solely on the source of expertise and seeks to understand how the expert’s gender and national identity influence policy decision-makers’ perception of the value of given advice and the likelihood of adoption. The research is driven by two overarching research questions:
(R1) Does the expert’s gender influence how valuable policy decision-makers perceive that expert advice to be and how likely they are to adopt it?
- 1 The term “national identity” is used here to distinguish between local Armenians, diaspora Armenian (...)
(R2) Does the expert’s national identity influence how valuable policy decision-makers perceive that expert advice to be and how likely they are to adopt it?1
- 2 Our research indicates that 65.4% of the current members/parliamentarians of the National Assembly (...)
16To examine whether the source of expertise influences how people consider and adopt advice when making policy decisions, a survey was conducted with a total of 50 current graduate students from the American University of Armenia, Yerevan State University, Armenian State University of Economics, the Public Administration Academy of Armenia, and the Diplomatic School of Armenia. Participants were required to be actively enrolled in political science, international relations, economics, and/or public administration programs at their respective institutions. These five institutions were targeted for two reasons: First, they are the most prominent universities in Armenia with the relevant majors; and second, the majority of the current policymakers in Armenia are graduates of these universities.2 The study intentionally targeted graduate students rather than current policymakers since these subjects are less likely to be bound by their official position, professional experience, and/or party affiliation when considering expert advice. Graduate students, as potential future policymakers, were considered the ideal target group, acting as a type of tabula rasa that controlled for confounding variables such as political party loyalty, professional/institutional affiliation, and training. The intention was to survey subjects who were willing to freely share their perceptions and were likely not trained to recognize and control their biases. Ultimately, the aim of this study was to test for possible gender and national identity biases when considering expert advice rather than the actual content of that advice. Moreover, this study focused on this population specifically because their educational background prepares them to participate in public decision-making processes in Armenia in the near future. They are also the most likely to become policymakers in this country. Thus, while this study is exploratory in nature, it targeted a key demographic studying at the most relevant institutions in Armenia.
17Participants were recruited either through in-person invitations or via known contacts. A combination of convenience and purposive sampling methods was employed that relied on personal contacts, but with predefined selection criteria: being a graduate student in a specific degree program at one of the above-listed institutions of higher learning in Armenia. Participation in this study was completely voluntary and without compensation. The survey was not timed and was administered either in-person or online depending on the institution in question. On average, it took participants about 10 minutes to complete the survey. The survey was conducted between April 1 and May 15, 2022. All responses were anonymous.
18It consisted of a single questionnaire containing 12 statements by fictional experts on wealth redistribution and text structures. The names of fictional experts were given next to each statement to provide cues about the speakers’ identities. First, the names included the professional academic title prefix “Dr.” to denote the experts’ credentials. Second, the name indicated the expert’s gender and whether they were local Armenians, diaspora Armenians, or foreigners. Common Armenian and Western female and males names were chosen to indicate the gender of the experts. The spelling of the surname suffixes(“-yan” for locals and “-ian” for diaspora) was manipulated to distinguish between the national identity of local Armenian and diaspora Armenian experts. This is a common way for local and diaspora Armenians to distinguish between each other, even today. The differences in naming strategies for diaspora Armenians have extensive historical and political roots, tracing back to the Ottoman Empire, where Armenians, as an ethnic and religious minority, had to navigate a precarious nationalist social landscape (Bayar, 2021; Türköz, 2018; Adalian, 2010). Thus, there were six categories of experts: local Armenian male, local Armenian female, diaspora Armenian male, diaspora Armenian female, foreign male, and foreign female. The questionnaire included two statements by experts from each category, for a total of 12 statements.
19The research design intentionally included the creation of fictional experts as opposed to using the names and statements of real experts. First, using fictional experts and statements avoids activating any previous associations a respondent may have with an actual expert; respondents may choose to trust the advice of an expert that they are familiar with for many reasons, many of which may be unrelated to their status as an expert on a particular topic or their identity. Statements by recognizable or popular experts may activate associations with things like their political affiliation, past statements about other issues, personal life, or other traits. By highlighting the expert status of the speaker and manipulating their identity, we were able to more clearly observe how the identity of a speaker may influence a respondent's level of trust in the expert’s advice. This is the central focus of this study.
20The expert statements addressed wealth redistribution and tax structures in Armenia. This specific topic was selected because it is arguably a policy issue about which the participants would be somewhat knowledgeable but is not so divisive so as to overshadow the influence of the advice and the characteristics of the fictional expert providing it. Expert statements were framed either in support of or against the need to address wealth redistribution and restructure the tax system in Armenia. Statements for and against these positions were divided equally by gender and national identity. The rationale behind evenly splitting statements both in support of and against wealth redistribution across gender and national identity was to isolate these two variables/potential biases and specifically test for changes within each group, thereby controlling for all other confounding variables (content of policy position). To control for the influence of participants’ policy position bias on their judgement of each statement, the fictional expert sources in each category presented claims both in support of and against wealth redistribution. For example, a fictional local Armenian female expert made a statement in support of wealth distribution, while another local Armenian female expert made an opposing claim rejecting the need to address it and restructure the Armenian tax system. The same was repeated for all categories of experts. Below are examples of these opposing statements made by imaginary local Armenian female experts:
“Major socioeconomic differences persist in Armenia. This socioeconomic gap will only increase if the Armenian government does not do something to help poor Armenians overcome poverty, which should be of primary concern if this country is to become stronger and more successful.” - Dr. Anahit Ohanyan
“Armenians living in poverty are not the concern of the more successful members of society. The Armenian government should not compel richer Armenians to support those living in poverty by increasing tax rates. Such a move is not in tune with the principles of liberal democracy and will only serve to stifle individual motivation and free competition.” - Dr. Mane Margaryan
21Respondents were asked to imagine themselves in the role of Armenia’s policymakers, to consider each of the statements made by experts in their respective fields, and to rate each statement according to how much value they placed on the advice and their likelihood of using it in their decision-making. A five-point Likert scale was used to rate each statement on a continuum ranging from completely valueless (1) to extremely valuable (5). The full questionnaire including all expert statements and rated responses is available in Appendix A.
22Quantitative data analysis was used to examine whether the characteristics of the expert giving the advice influenced the degree of the value respondents placed on that advice. The objective was to examine how gender and national origin influenced the value respondents placed on given advice. Put differently, quantitative data analysis was deemed to be the most effective in unearthing possible differences in how respondents rated statements made by male compared to female experts, and to compare statements made by local Armenians, diaspora Armenian, and foreign experts.
23The mode of responses was analyzed to determine the extent to which expert advice was valued and how likely it was to inform respondents’ decision-making process. In other words, the mode provides an overview as to whether respondents generally valued policy expertise, regardless of the expert’s gender, national identity, or even policy position). Out of the 12 statements, the most frequent response to nine of the statements was “4 – Somewhat Valuable.” Only three statements had a mode of “3 – “Neutral”. Interestingly, two of the three statements whose mode was “Neutral” were related to the male diaspora Armenian category of expert.
Figure 1: Mode for each of the 12 expert advice statements.
Expert Category
|
Mode (Perceived Values of Expert Statement)
|
MD (PRO)
Male Diaspora (Pro-redistribution)
|
3 – Neutral
|
MD (ANTI)
Male Diaspora (Anti-redistribution)
|
3 – Neutral
|
FA (ANTI)
Female Armenian (Anti-redistribution)
|
3 – Neutral
|
FA (PRO)
Female Armenian (Pro-redistribution)
|
(4) – Somewhat Valuable
|
MA (PRO)
Male Armenian (Pro-redistribution)
|
(4) – Somewhat Valuable
|
MA (ANTI)
Male Armenian (Anti-redistribution)
|
(4) – Somewhat Valuable
|
FD (PRO)
Female Diaspora (Pro-redistribution)
|
(4) – Somewhat Valuable
|
FD (ANTI)
Female Diaspora (Anti-redistribution)
|
(4) – Somewhat Valuable
|
MF (PRO)
Male Foreign (Pro-redistribution)
|
(4) – Somewhat Valuable
|
MF (ANTI)
Male Foreign (Anti-redistribution)
|
(4) – Somewhat Valuable
|
FF (PRO)
Female Foreign (Pro-redistribution)
|
(4) – Somewhat Valuable
|
FF (ANTI)
Female Foreign (Anti-redistribution)
|
(4) – Somewhat Valuable
|
Source : the Author
24Generally, the mode of responses for each of the expert statements indicates a high regard for expertise. Respondents judged the statements made by experts to be valuable in their decision-making process, regardless of expert source characteristics.
25The mean for each of the 12 expert statements was calculated in order to rank each expert category by how valuable respondents judged them to be in their decision-making process. Such a ranking indicates how respondents perceived statements made by male compared to female experts, and statements made by local Armenian experts compared to diaspora Armenian experts compared to foreign experts.
Figure 2: Raking table of means for each of the expert advice statements.
Rank
|
Average Mean
|
Expert Category
|
1
|
3.78
|
FA (PRO)
Female Armenian (Pro-redistribution)
|
2
|
3.52
|
FD (PRO)
Female Diaspora (Pro-redistribution)
|
3
|
3.46
|
MA (PRO)
Male Armenian (Pro-redistribution)
|
4
|
3.44
|
MA (ANTI)
Male Armenian (Anti-redistribution)
|
5
|
3.42
|
FF (PRO)
Female Foreign (Pro-redistribution)
|
6
|
3.34
|
MD (ANTI)
Male Diaspora (Anti-redistribution)
|
7
|
3.22
|
FF (ANTI)
Female Foreign (Anti-redistribution)
|
8
|
3.18
|
MD (PRO)
Male Diaspora (Pro-redistribution)
|
9
|
3.18
|
MF (ANTI)
Male Foreign (Anti-redistribution)
|
10
|
3.14
|
FA (ANTI)
Female Armenian (Anti-redistribution)
|
11
|
3.20
|
FD (ANTI)
Female Diaspora (Anti-redistribution)
|
12
|
2.92
|
MF (PRO)
Male Foreign (Pro-redistribution)
|
Source: the Author
26Generally, respondents seemed to value the two categories of local Armenian and diaspora Armenian expertise more than they valued foreign expert advice. However, it must be stressed that these four categories of experts were ranked only slightly higher than the other two (male and female) categories of foreign expert. This lends weight to the argument that the national identity of the expert influences how valuable their advice is perceived to be. On the other hand, the gender of the expert did not seem to influence how respondents perceived expert advice. The mean rankings clearly show that respondents did not have a preference for either male or female expertise. However, the mean ranking does reveal one interesting trend - respondents generally ranked pro-distribution statements higher than anti-redistribution statements.
27Correlations were calculated for each possible statement pairing. Out of 66 possible pairings, only 16 relationships were found to be statistically significant. This means that there were only 16 cases in which an evidenced (positive or negative) relationship between two expert statements was not due to chance or error (within the standard 5% probability threshold).
Source: the Author
28These correlation findings indicate two important facts: a) respondents considered each policy statement carefully and independently (rather than employing some random sequence to responses); and b) every single statistically significant pairing indicated policy consistency. More specifically, if a respondent rated a pro-redistribution high from one expert source, they were more likely to value another pro-distribution statement from another expert source highly also. On the other hand, if a respondent valued a pro-redistribution statement high, they were more likely to rank an anti-redistribution statement lower. This trend is evidenced for all 16 statement pairings.
29The figure below highlights only the 16 cases of statistically significant relationships between policy statements. Cases show high levels of policy consistency evidenced in respondents’ answers. Policy position (pro-distribution compared to anti-distribution) is underlined for each of the pairings.
Figure 4: Correlation table presents the 16 cases of statistically significant relationships between policy statements.
- 3 An abbreviation key for all the statement categories is located in Appendix B.
Expert Category Pairing3
|
Pearson r / Significance (p-value)
|
FF (PRO) - FA (ANTI)
|
-0.464 / 0.001
|
MA (ANTI) - MD (ANTI)
|
0.314 / 0.026
|
MA (ANTI) - FD (PRO)
|
-0.308 / 0.030
|
MA (ANTI) - MF (ANTI)
|
0.412 / 0.003
|
MD (ANTI) - MF (ANTI)
|
0.323 / 0.022
|
MA (PRO) - MF (PRO)
|
0.443 / 0.001
|
MA (PRO) - FD (PRO)
|
0.499 / 0.000
|
MA (PRO) - FA (ANTI)
|
-0.315 / 0.026
|
FD (ANTI) - FF (ANTI)
|
0.414 / 0.003
|
FA (ANTI) - MD (PRO)
|
-0.337 / 0.017
|
FD (ANTI) - FA (ANTI)
|
0.321 / 0.023
|
FD (ANTI) - FF (ANTI)
|
0.326 / 0.021
|
MF (PRO) - FD (PRO)
|
0.512 / 0.000
|
FF (ANTI) - FA (ANTI)
|
0.306 / 0.031
|
FF (ANTI) - MF (ANTI)
|
0.281 / 0.048
|
FA (ANTI) - MF (ANTI)
|
0.287 / 0.043
|
Source: the Author
30Most importantly, however, there are no statistically significant relationships between gender and/or national identity pairings. Statistically significant relationships are confirmed only for policy position consistency, not for the categories of expert gender and national identity. This further establishes that correlation testing provides no evidence that expert gender or national identity factor into how valuable policy decision-makers perceive expert advice to be or how likely they are to adopt it.
31A two-sample t-test was performed to compare two groups of variables: male experts versus female experts and local Armenians versus diaspora Armenians compared to foreign experts. The two-sample t-test was especially useful for testing gender and national identity categories in isolation while keeping all other variables constant. Results show that, for both gender and national identity categories, there were no statistically significant differences between gender and national identity (except for two isolated cases). This means that respondents generally did not perceive the value and usefulness of policy expertise differently if it came from a male or female expert, or from experts representing different national identities.
- 4 What is especially interesting to highlight here is not only that this sole pair shows a statistica (...)
32The figure below shows the calculations for a two-sample t-test of the gender variable. Gender is underlined for each pairing tested (while holding national identity and policy position constant). Out of six possible pairings, only one pair showed a statistically significant difference in mean values: FF (PRO) = 3.42 - MF (PRO) = 2.92. This indicates that respondents only perceived a difference in the value of expert advice when judging policy statements made by a female foreign expert compared to a male foreign expert.4 For all other categories, however, gender did not influence how respondents perceived the value of expert advice and their willingness to adopt it.
Figure 5: Two-sample t-test results table for the variable of gender.
Expert Category Pairing
|
Significance (p-value)
|
FF (PRO) - MF (PRO)
|
0.018*
|
FF (ANTI) - MF (ANTI)
|
0.842
|
FA (PRO) - MA (PRO)
|
0.088
|
FA (ANTI) - MA (ANTI)
|
0.121
|
FD (PRO) - MD (PRO)
|
0.097
|
FD (ANTI) - MD (ANTI)
|
0.209
|
Source: the Author
33Another two-sample test was performed for the variable of national identity as presented in the following figure below. National identity was compared between local and diaspora Armenians, as well as between local Armenians and foreign experts. Each pairing was tested while holding gender and policy position constant. Out of 12 possible pairings, only one statistically significant relationship was calculated. This means that respondents only perceived a difference in the value of expert advice when judging policy statements made by a local Armenian male compared to a foreign male expert. For all other categories, national identity distinctions did not influence how respondents perceived the value of expert advice and their willingness to adopt it.
Figure 6: Two-sample t-test results table for the variable of national identity.
Expert Category Pairing
|
Significance
(p-value)
|
Expert Category Pairing
|
Significance
(p-value)
|
Expert Category Pairing
|
Significance
(p-value)
|
MA (ANTI) –
MD (ANTI)
|
0.574
|
MA (ANTI) –
MF (ANTI)
|
0.145
|
MD (ANTI) –
MF (ANTI)
|
0.405
|
MA (PRO) –
MD (PRO)
|
0.204
|
MA (PRO) –
MF (PRO)
|
0.003*
|
MD (PRO) –
MF (PRO)
|
0.233
|
FA (ANTI) –
FD (ANTI)
|
0.823
|
FA (ANTI) –
FF (ANTI)
|
0.659
|
FD (ANTI) –
FF (ANTI)
|
0.491
|
FA (PRO) –
FD (PRO)
|
0.175
|
FA (PRO) –
FF (PRO)
|
0.095
|
FD (PRO) –
FF (PRO)
|
0.617
|
Source: the Author
34The limited quantitative results of this exploratory study suggest that the gender and national identity of experts do not influence how valuable potential policy decision-makers in Armenia perceive their advice to be and how likely they are to adopt it. Therefore, it would seem that, at least when it comes to expert advice considerations, Armenian graduates really do not seem to care “who says so.” Other factors, such as the position that the expert supports, seem to be more important than the expert’s gender and national identity when considering expert advice.
35This study highlights that Armenia’s future policy decision-makers do not disregard expert advice entirely when making policy decisions. In fact, Armenian graduates generally “somewhat” value expertise, regardless of the gender and national identity of the source. However, this study was not able to determine whether these future decision-makers value expert advice more than non-expert advice. It could only conclude that they somewhat value advice branded as ‘expert. Moreover, this study cannot make general causal claims that Armenian graduates who value expertise will act on expert recommendations (or possibly change their views and decisions on policy issues).
36The ranking of means in this study (Figure 2) indicates a slight preference for local and diaspora Armenian experts compared to foreign experts. This perceived “favoritism” is understandable, however, since the subject of advice was wealth redistribution policy and tax structures in Armenia specifically. It seems logical that respondents would consider local and diaspora Armenian expertise on this subject to be more knowledgeable, and hence valuable, than that of foreign experts. At the same time, such a topic seemingly renders generalized conclusions regarding national partiality of expertise implausible. This analysis is further corroborated by the results of the two-sample t-tests, which indicate a lack of statistically significant differences in policy-makers’ perceptions of expertise value when comparing enders and national identity groups.
37While gender and national identity do not significantly influence how valuable Armenian graduates perceive policy advice or their willingness to act upon it, correlation testing showed that these potential policy-makers exhibit high levels of policy position consistency. Of the 16 statistically significant policy statement pairings, all highlight that Armenian graduates tend to group pro-redistribution and anti-redistribution policy statements separately and value statements corroborating one position over the other. In other words, if an Armenian graduate rated a pro-redistribution policy statement highly, it was extremely likely that they would also rate another pro-redistribution statement highly, regardless of the source of expertise. The same is true for the anti-redistribution policy position. The mean ranking corroborates this, highlighting the fact that pro-redistribution policy statements were generally perceived as more valuable than anti-redistribution statements, regardless of the expert source. These results suggest that the personal conviction and political/policy persuasion of the respondent are the most important factors influencing how respondents perceive expert advice and how likely they are to act on it. In other words, whether a policy statement aligns with the future policy-makers’ worldview is the ultimate deciding factor in whether expert advice is adopted.
38Previous studies (Carlin et al., 2021; Neal, 2014; Greve-Poulsen et al., 2021) corroborate the fact that expert source characteristics seemingly matter very little in whether expertise is adopted by decision-makers. If Armenia’s future policy-makers are heavily swayed by personal conviction when making policy decisions, policy literacy must be addressed in the Armenian context, with special emphasis on policy analysis methods. Subjective judgments are an integral (and unavoidable) part of the policy-making process. However, future policymakers would benefit greatly from grounding their policy decisions in broader models of public policy ethics. If the value of policy is to be judged on personal standards of “good” versus “bad”, it would be beneficial for Armenia’s future policy-makers to receive training in what is meant by the public good, private vs. public interest, and “just” state policies – themes which could be explored in a comprehensive ethics and public policy course. Policymakers need to understand why they believe what they believe, and if and why specific policy advice is “good” or “right” (regardless of its source).
39It must be acknowledged that there are some very clear limitations to this exploratory study. Most importantly, the sample size is far too small to generalize about policymaking in Armenia. Further research is necessary to replicate this study on a large-N sample in order to get a more complete picture of the state of policy decision-making in Armenia. Furthermore, this study includes only a small pool of future Armenian policy-makers and only those educated in the social sciences. The study could have also benefited from a more direct investigation of possible differences between female and male policy-makers. The sample in this study is heavily biased towards female participants and potential policy decision-makers.
40Apart from sampling constraints, the study could have tested additional expert source characteristics (such as expert’s age) and considered other ways to present gender and national identity differences more explicitly. For example, the survey relied entirely on written cues. Future research could incorporate visual cues, such as pictures of male and female experts, to more accurately reveal implicit bias when considering expert advice. Furthermore, the generalizability of this study could be improved by testing other policy issues. This study tests one particular policy issue - wealth distribution policy. This topic was chosen because it is perceived as less divisive. However, using other policy issues related to gender and national identity may reveal a more pronounced gendered or national valence. Finally, the study’s survey design could also ameliorate the visibility of expert indicators and devise cues that more effectively differentiate between the national identity of local and diaspora Armenians. For example, fictitious experts with the professional title prefix “Dr.” could be accompanied by a short narrative/bio listing their credentials and accomplishments to assure participants of their expertise. Similarly, a description of the expert’s national background could be used instead of relying solely on the differentiated spelling of the surname suffixes. These issues would therefore have to be addressed and reassessed in the research design stage if this study were to be replicated. As it is, the study should be considered exploratory with preliminary findings that lay the groundwork for more in-depth studies, at least as far as the Armenian context is concerned.