How Search Engines Reinforce Who We See as “Political”

Table of links Table of links ABSTRACT ABSTRACT ABSTRACT INTRODUCTION INTRODUCTION INTRODUCTION SEARCH ENGINES AND WOMEN’S DESCRIPTIVE REPRESENTATION SEARCH ENGINES AND WOMEN’S DESCRIPTIVE REPRESENTATION SEARCH ENGINES AND WOMEN’S DESCRIPTIVE REPRESENTATION MEASURING THE EXTENT OF ALGORITHMIC REPRESENTATION MEASURING THE EXTENT OF ALGORITHMIC REPRESENTATION MEASURING THE EXTENT OF ALGORITHMIC REPRESENTATION GENERAL DISCUSSION, CONCLUSION AND REFERENCES GENERAL DISCUSSION, CONCLUSION GENERAL DISCUSSION, CONCLUSION AND AND REFERENCES REFERENCES GENERAL DISCUSSION, CONCLUSION AND REFERENCES GENERAL DISCUSSION, CONCLUSION AND REFERENCES What is the role of search engines in upholding views of politics as a masculine and white domain? In this article, we proposed a framework of algorithmic representation of minoritized groups that theorizes this question as a chain of four intertwined types of bias, which we then empirically tested across four empirical studies. Overall, we find consistent support of all expectations. First, search engines algorithmically underrepresent women in their output in absolute terms. Second, the extent of women’s algorithmic underrepresentation in Google image searches generally tracks with their collective descriptive representation in 84 legislative bodies across 56 countries. Third, exposure to gender, race, and intersectional bias in search engine outputs consistently distort perceptions of the political reality by decreasing the estimated presence of politicians belonging to these groups among individual search engine users. Finally, these misperceptions act as a casual conduit through which algorithmic biases not only diminish the perceived chances of winning an election for minority candidates but also result in voters feeling that their voices matter less. Our findings have several implications for the study of representation in politics. First, our account of algorithmic representation introduces search engines as an important piece in the puzzle of persistent underrepresentation of minoritized groups in contemporary politics. The findings illustrate how widely used digital tools shape—and are shaped by—existing structural inequalities; they raise concerns about the future impacts of these tools, in particular considering the rapid adoption of AI-driven systems in different societal sectors. This combination of structural and algorithmic explanations of political bias goes with and beyond more established candidate- and voter-centered approaches (see also Dolan and Hansen 2018). On the supply-side, Juenke and Shah (2016) showed that voters can rarely “choose amongst a menu of racially or politically diverse candidates” (84). Our findings suggest that search engines may compound this dearth of minoritized candidates by rendering those who are running less visible through algorithmic filtering (see Thomsen and King 2020; O’Brien 2015). Moreover, googling politics and not finding any role models who share your gender or skin color might stymy nascent political ambition in potential future candidates (Bos et al. 2022; Fox and Lawless 2011; Kanthak and Woon 2015). On the demand-side, recent studies indicate that overt forms of voter bias are disappearing (Rohrbach et al. 2023; Schwarz and Coppock 2022; Teele et al. 2018). By conceptually broadening the notion of strategic discrimination (Bateson 2020; Green et al. 2022), we show that search engine outputs can act as the basis for voters’ assessments of politicians’ electability and thus present a more subtle and indirect form of voter bias. Empirically, our mediation analyses consistently support the notion of strategic discrimination as a causal driver of race- and gender-based prejudice in contemporary politics. Second, we have theoretically and empirically outlined the circular logic of algorithmic bias in politics (see Savaget et al. 2019). Algorithms power AI models which are trained on data reflecting existing structural inequalities; in turn, algorithmic (mal)performance amplifies social and political realities through inherent biases in their filtering (Burrell and Fourcade 2021). All findings from our experimental conditions attest to this vicious cycle in which search engines underrepresent candidates from minoritized groups and thereby diminish their perceived electoral chances. However, this pattern reverses into a virtuous cycle in the conditions where we artificially removed bias from search engine output. Algorithmic overrepresentation of minoritized groups boosts voters’ estimations of their political presence in government. This finding extends existing work on public perceptions of political udnerrepresentation (Stauffer 2021; Dolan and Hansen 2018; Dolan and Sanbonmatsu 2009) by positioning search engines as sources of public impression formation. In line with previous research (Stauffer 2021), the evidence linking perceived inclusion of minority groups to increased feelings of external efficacy is especially noteworthy. In times of fatiguing democratic support around the world (Diamond 2015; Gavras et al. 2022; Graham and Svolik 2020), algorithmic representation could play a vital role in fostering public support of political institutions. Third, our empirical evidence contributes to ongoing public debates and cross-disciplinary research on algorithmic fairness (Kalluri 2020; Weinberg 2022; Wong 2020). Our findings are integral for increasing societal awareness about the discriminatory potential of AI-driven systems in the context of political participation (Savaget et al. 2019; Stier et al. 2022). By delineating the consequences of exposure to algorithmic representation, we showcase that the concept of "bias in, bias out" extends beyond a catchphrase; the functioning of proprietary AI-driven systems, such the ones used by Google, should be more actively incorporated into broader discussions on algorithmic governance and injustice (Birhane 2021). Accordingly, our findings provide an empirical basis for developing new regulation for preventing risks associated with the growing adoption of AI and is thus relevant for a broad range of stakeholders, including policymakers and industry, but also civil society and human right advocacy groups. This study comes with several limitations. The algorithm audits center on a single case of Google image search. Insightful and widely used as this specific case may be, it remains unclear how our framework of algorithmic representation transfers to other AI-curated digital spaces. Future research should extend the focus to social media platforms like TikTok that heavily rely on AI to structure content resulting in the amplification of undesirable political messages, including hate and disinformation (Weimann and Masri 2023). Another limitation arises from the use of computer vision. Though manual coding suggests good reliability, this approach is limited to a binary classification and disregards the complexity of gender identity (Scheuerman et al. 2019). For ethical and baseline-related reasons regarding the use of AI, our auditing studies also ignore the question of representation of race. Critical analysis and in-depth qualitative work could help establish a benchmark for how search engines (mis)represent non-white politicians. Much of our analysis captures dynamics on the level of political institutions and groups of candidates. A promising approach would be extending our framework to the level of individual candidates to study algorithmic representation in the context of politicians with concrete political histories and, for instance, partisan identities (Pradel 2021). Finally, our experimental evidence documents algorithmic underrepresentation decreases perceived electability. However, more work is needed to delineate downstream electoral consequences of such diminished electability, ideally using observational data from actual elections or longitudinal designs. Whereas the political landscape is only slowly becoming more inclusive towards women and non-white politicians, the algorithmic landscape has been evolving rapidly in all directions. This article captured how these two landscapes are interlocked to sustain white and masculine views of politics at present; tracking how these dynamics unfold in the future remains a major concern for the legitimacy of democratic structures and processes. REFERENCES REFERENCES Albiero, Vitor, Krishnapriya Ks, Kushal Vangara, Kai Zhang, Michael C King, and Kevin W Bowyer. 2020. “Analysis of gender inequality in face recognition accuracy”. In Proceedings of the WACV 2020, Snowmass, Colorado, pp. 81–89. American Academy of Arts & Sciences. 2019. “Time Spent Reading”. Technical report, American Academy of Arts & Sciences. Ashworth, Scott, Christopher R. Berry, and Ethan Bueno de Mesquita. 2024, January. “Modeling Theories of Women’s Underrepresentation in Elections”. American Journal of Political Science 68 (1): 289–303. Atkeson, Lonna Rae and Nancy Carrillo. 2007. “More is Better: The Influence of Collective Female Descriptive Representation on External Efficacy”. Politics & Gender 3 (1): 79–101. Bandy, Jack. 2021, April. “Problematic Machine Behavior: A Systematic Literature Review of Algorithm Audits”. Proceedings of the ACM on Human-Computer Interaction 5 (1): 1–34. Bast, Jennifer. 2024, January. “Managing the Image. The Visual Communication Strategy of European Right-Wing Populist Politicians on Instagram”. Journal of Political Marketing 23 (1): 1–25. Bateson, Regina. 2020. “Strategic discrimination”. Perspectives on Politics 18 (4): 1068–1087. Birhane, Abeba. 2021, February. “Algorithmic injustice: a relational ethics approach”. Patterns 2 (2): 1–9. Bos, Angela L, Jill S Greenlee, Mirya R Holman, Zoe M Oxley, and Celeste J Lay. 2022. “This One’s for the Boys: How Gendered Political Socialization Limits Girls’ Political Ambition and Interest”. American Political Science Review 116 (2): 484–501. Bratton, Kathleen A and Leonard P Ray. 2002. “Descriptive representation, policy outcomes, and municipal day-care coverage in Norway”. American Journal of Political Science 46 (2): 428–437. Type: Journal Article. Burrell, Jenna and Marion Fourcade. 2021, July. “The Society of Algorithms”. Annual Review of Sociology 47 (1): 213–237. Carpinella, Colleen and Nichole M. Bauer. 2021, March. “A visual analysis of gender stereotypes in campaign advertising”. Politics, Groups, and Identities 9 (2): 369–386. Cavazos, Jacqueline G, Jonathon P. Phillips, Carlos D. Castillo, and Alice O. O’Toole. 2021. “Accuracy Comparison Across Face Recognition Algorithms: Where Are We on Measuring Race Bias?”. IEEE Transactions on Biometrics, Behavior, and Identity Science 3 (1): 101–111. Corbett, Christianne, Jan G Voelkel, Marianne Cooper, and Robb Willer. 2022. “Pragmatic Bias Impedes Women’s Access to Political Leadership”. Proceedings of the National Academy of Sciences 119 (6): 1–11. Crabtree, Charles, Jae Yeon Kim, S. Michael Gaddis, John B. Holbein, Cameron Guage, and William W. Marx. 2023, March. “Validated names for experimental studies on race and ethnicity”. Scientific Data 10 (1): 130. Dassonneville, Ruth and Ian McAllister. 2018, April. “Gender, Political Knowledge, and Descriptive Representation: The Impact of Long-Term Socialization”. American Journal of Political Science 62 (2): 249–265. Diakopoulos, Nicholas, Daniel Trielli, Jennifer Stark, and Sean Mussenden. 2018. “I vote for—how search informs our choice of candidate”. In M. Moore and D. Tambini (Eds.), Digital Dominance: The Power of Google, Amazon, Facebook, and Apple, Volume 22, pp. 320–341. Oxford University Press. Diamond, Larry. 2015. “Facing up to the democratic recession”. Journal of Democracy 26 (1): 141–155. Diekman, Amanda B., Wind Goodfriend, and Stephanie Goodwin. 2004, February. “Dynamic Stereotypes of Power: Perceived Change and Stability in Gender Hierarchies”. Sex Roles 50 (3): 201–215. Dolan, Kathleen. 1998. “Voting for Women in the "Year of the Woman"”. American Journal of Political Science 42 (1): 272–293. Dolan, Kathleen and Kira Sanbonmatsu. 2009, May. “Gender Stereotypes and Attitudes Toward Gender Balance in Government”. American Politics Research 37 (3): 409–428. Dolan, Kathleen A. and Michael Hansen. 2018. “Blaming Women or Blaming the System? Public Perceptions of Women’s Underrepresentation in Elected Office”. Political Research Quarterly 71 (3): 668–680. Eagly, Alice H, Christa Nater, David I Miller, Michèle Kaufmann, and Sabine Sczesny. 2020. “Gender stereotypes have changed: A cross-temporal meta-analysis of US public opinion polls from 1946 to 2018”. American psychologist 75 (3): 301–315. Edelman Trust Institute. 2024. “2024 Edelman Trust Barometer”. Technical report, Edelman Trust Institute. Epstein, Robert and Ronald E Robertson. 2015. “The search engine manipulation effect (SEME) and its possible impact on the outcomes of elections”. Proceedings of the National Academy of Sciences 112 (33): 4512–4521. Fabris, Alessandro, Alberto Purpura, Gianmaria Silvello, and Gian Antonio Susto. 2020. “Gender stereotype reinforcement: Measuring the gender bias conveyed by ranking algorithms”. Information Processing & Management 57 (6): 1–24. Fox, Richard L and Jennifer L Lawless. 2011. “Gendered perceptions and political candidacies: A central barrier to women’s equality in electoral politics”. American Journal of Political Science 55 (1): 59–73. Friedman, Batya and Helen Nissenbaum. 1996. “Bias in computer systems”. ACM Transactions on Information Systems 14 (3): 330–347. Friesen, Amanda, Ryan Burge, and Kylee Britzman. 2021, February. “Digital Segregation: Gender, Occupation, and Access to Politics”. Social Science Computer Review 39 (1): 38–55. Gavras, Konstantin, Harald Schoen, and Alexander Wuttke. 2022. “Have Europeans Grown Tired of Democracy? New Evidence from Eighteen Consolidated Democracies, 1981–2018”. British Journal of Political Science 52 (1): 416–428. Gershon, Sarah A. 2013. “Media Coverage of Minority Congresswomen and Voter Evaluations: Evidence from an Online Experimental Study”. Political Research Quarterly 66 (3): 702–714. Gidengil, Elisabeth and Joanna Everitt. 2000, October. “Filtering the Female”. Women & Politics 21 (4): 105–131. Graham, Matthew H and Milan W Svolik. 2020. “Democracy in America? Partisanship, Polarization, and the Robustness of Support for Democracy in the United States”. American Political Science Review 114 (2): 392– 409. Green, Jon, Brian Schaffner, and Sam Luks. 2022, December. “Strategic Discrimination in the 2020 Democratic Primary”. Public Opinion Quarterly 86 (4): 886–898. Guilbeault, Douglas, Solène Delecourt, Tasker Hull, Bhargav Srinivasa Desikan, Mark Chu, and Ethan Nadler. 2024, February. “Online images amplify gender bias”. Nature 626 : 1049–1055. Haim, Mario, Florian Arendt, and Sebastian Scherr. 2017, February. “Abyss or Shelter? On the Relevance of Web Search Engines’ Search Results When People Google for Suicide”. Health Communication 32 (2): 253–258. Hessami, Zohal and Mariana Lopes da Fonseca. 2020. “Female political representation and substantive effects on policies: A literature review”. European Journal of Political Economy 63 (101896): 1–9. Huddy, Leonie and Nayda Terkildsen. 1993. “Gender stereotypes and the perception of male and female candidates”. American journal of political science 37 (1): 119–147. Inter-parliamentary Union. 2024. “Global and regional averages of women in national parliaments”. Report, Inter-parliamentary Union. Juenke, Eric Gonzalez and Paru Shah. 2016. “Demand and Supply: Racial and Ethnic Minority Candidates in White Districts”. The Journal of Race, Ethnicity, and Politics 1 (1): 60–90. Kahn, Kim. 1994. “Does gender make a difference? An experimental examination of sex stereotypes and press patterns in statewide campaigns”. American Journal of Political Science 38 (1): 162–195. Type: Journal Article. Kalluri, Pratyusha. 2020. “Don’t ask if artificial intelligence is good or fair, ask how it shifts power”. Nature 583 (7815): 169–169. Kanthak, Kristin and Jonathan Woon. 2015, July. “Women Don’t Run? Election Aversion and Candidate Entry”. American Journal of Political Science 59 (3): 595–612. Publisher: John Wiley & Sons, Ltd. Lau, Richard and David Redlawsk. 2001. “Advantages and Disadvantages of Cognitive Heuristics in Political Decision Making”. American Journal of Political Science 45 (4): 951–971. Lowande, Kenneth, Melinda Ritchie, and Erinn Lauterbach. 2019, July. “Descriptive and Substantive Representation in Congress: Evidence from 80,000 Congressional Inquiries”. American Journal of Political Science 63 (3): 644–659. Makhortykh, Mykola, Aleksandra Urman, and Roberto Ulloa. 2021. “Detecting Race and Gender Bias in Visual Representation of AI on Web Search Engines”. In L. Boratto, S. Faralli, M. Marras, and G. Stilo (Eds.), Advances in Bias and Fairness in Information Retrieval, Cham, pp. 36–50. Springer. McDermott, Monika L. 1997. “Voting cues in low-information elections: Candidate gender as a social information variable in contemporary United States elections”. American Journal of Political Science 41 (1): 270–283. Mendelberg, Tali, Christopher F Karpowitz, and Nicholas Goedert. 2014. “Does descriptive representation facilitate women’s distinctive voice? How gender composition and decision rules affect deliberation”. American Journal of Political Science 58 (2): 291–306. Type: Journal Article. Niven, David. 1998. “Party elites and women candidates: The shape of bias”. Women & Politics 19 (2): 57–80. Niven, David. 2004. “A Fair Test of Media Bias: Party, Race, and Gender in Coverage of the 1992 House Banking Scandal”. Polity 36 (4): 637–649. Noble, Safiya Umoja. 2018. Algorithms of Oppression. New York University Press. O’Brien, Diana Z. 2015. “Rising to the top: Gender, political performance, and party leadership in parliamentary democracies”. American Journal of Political Science 59 (4): 1022–1039. Type: Journal Article. Pew Research Center. 2023a. “118th congress has a record number of women”. Technical report, Pew Research Center. Pew Research Center. 2023b. “U.s. congress continues to grow in racial, ethnic diversity”. Technical report, Pew Research Center. Philpot, Tasha S and Hanes Walton. 2007. “One of our own: Black female candidates and the voters who support them”. American Journal of Political Science 51 (1): 49–62. Pradel, Franziska. 2021, July. “Biased Representation of Politicians in Google and Wikipedia Search? The Joint Effect of Party Identity, Gender Identity and Elections”. Political Communication 38 (4): 447–478. Riedl, Andreas, Tobias Rohrbach, and Christina Krakovsky. 2022, March. ““I Can’t Just Pull a Woman Out of a Hat”: A Mixed-Methods Study on Journalistic Drivers of Women’s Representation in Political News”. Journalism & Mass Communication Quarterly Online first : 1–24. Rohrbach, Tobias, Loes Aaldering, and Daphne J Van Der Pas. 2023. “Gender Differences and Similarities in News Media Effects on Political Candidate Evaluations: A Meta-Analysis”. Journal of Communication 73 (2): 101– 112. Savaget, Paulo, Tulio Chiarini, and Steve Evans. 2019, June. “Empowering political participation through artificial intelligence”. Science and Public Policy 46 (3): 369–380. Scheuerman, Morgan Klaus, Jacob M Paul, and Jed R Brubaker. 2019. “How computers see gender: An evaluation of gender classification in commercial facial analysis services”. Proceedings of the ACM on Human-Computer Interaction 3 (CSCW): 1–33. Schwarz, Susanne and Alexander Coppock. 2022. “What Have We Learned about Gender from Candidate Choice Experiments? A Meta-Analysis of Sixty-Seven Factorial Survey Experiments”. The Journal of Politics 84 (2): 655–668. Schwemmer, Carsten, Carly Knight, Emily D. Bello-Pardo, Stan Oklobdzĳa, Martĳn Schoonvelde, and Jeffrey W. Lockhart. 2020, January. “Diagnosing Gender Bias in Image Recognition Systems”. Socius 6 : 1–17. Shehata, Adam and Jesper Strömbäck. 2014. “Mediation of political realities: Media as crucial sources of information”. In F. Esser and J. Strömbäck (Eds.), Mediatization of politics, pp. 93–113. New York: Palgrave Macmillan. Stauffer, Katelyn E. 2021. “Public Perceptions of Women’s Inclusion and Feelings of Political Efficacy”. American Political Science Review 115 (4): 1226–1241. Stephens-Davidowitz, Seth. 2014, October. “The cost of racial animus on a black candidate: Evidence using Google search data”. Journal of Public Economics 118 : 26–40. Stier, Sebastian, Frank Mangold, Michael Scharkow, and Johannes Breuer. 2022. “Post Post-Broadcast Democracy? News Exposure in the Age of Online Intermediaries”. American Political Science Review 116 (2): 768– 774. Stokes-Brown, Atiya Kai and Kathleen Dolan. 2010, November. “Race, Gender, and Symbolic Representation: African American Female Candidates as Mobilizing Agents”. Journal of Elections, Public Opinion and Parties 20 (4): 473–494. Teele, Dawn L, Joshua Kalla, and Frances M. Rosenbluth. 2018. “The Ties That Double Bind: Social Roles and Women’s Underrepresentation in Politics”. American Political Science Review 112 (3): 525–541. Thomsen, Danielle M and Aaron S King. 2020. “Women’s Representation and the Gendered Pipeline to Power”. American Political Science Review 114 (4): 989–1000. Type: Journal Article. Trevisan, Filippo, Andrew Hoskins, Sarah Oates, and Dounia Mahlouly. 2018, January. “The Google voter: search engines and elections in the new media ecology”. Information, Communication & Society 21 (1): 111–128. Trielli, Daniel and Nicholas Diakopoulos. 2022, January. “Partisan search behavior and Google results in the 2018 U.S. midterm elections”. Information, Communication & Society 25 (1): 145–161. Trimble, Linda, Jennifer Curtin, Meagan Auer, Angelia Wagner, and V. K. G. Woodman. 2022, April. “Works Well with Others: Gendered Mediation of Collaborative Leadership Skills”. Journal of Women, Politics & Policy 43 (2): 245–255. Urman, Aleksandra and Mykola Makhortykh. 2022. ““Foreign beauties want to meet you”: The sexualization of women in Google’s organic and sponsored text search results”. New Media & Society Online First : 1–22. Urman, Aleksandra and Mykola Makhortykh. 2023. “You are how (and where) you search? comparative analysis of web search behavior using web tracking data”. Journal of computational social science 6 (2): 741–756. Urman, Aleksandra, Mykola Makhortykh, and Aniko Hannak. 2024. “Mapping the Field of Algorithm Auditing: A Systematic Literature Review Identifying Research Trends, Linguistic and Geographical Disparities”. arXiv preprint : 1–22. Urman, Aleksandra, Mykola Makhortykh, and Roberto Ulloa. 2022, October. “The Matter of Chance: Auditing Web Search Results Related to the 2020 U.S. Presidential Primary Elections Across Six Search Engines”. Social Science Computer Review 40 (5): 1323–1339. U.S. Census Bureau. 2023. “QuickFacts: United States”. van der Pas, Daphne, Loes Aaldering, and Angela L. Bos. 2023, August. “Looks Like a Leader: Measuring Evolution in Gendered Politician Stereotypes”. Political Behavior Online first : 1–23. Van Der Pas, Daphne J and Loes Aaldering. 2020. “Gender Differences in Political Media Coverage: A Meta-Analysis”. Journal of Communication 70 (1): 114–143. van Oosten, Sanne, Liza Mügge, and Daphne van der Pas. 2024. “Race/Ethnicity in Candidate Experiments: a Meta-Analysis and the Case for Shared Identification”. Acta Politica 59 (1): 19–41. Vlasceanu, Madalina and David M Amodio. 2022. “Propagation of societal gender inequality by internet search algorithms”. Proceedings of the National Academy of Sciences 119 (29): 1–8. Wallace, Julian. 2018, March. “Modelling Contemporary Gatekeeping”. Digital Journalism 6 (3): 274–293. Weimann, Gabriel and Natalie Masri. 2023, May. “Spreading Hate on TikTok”. Studies in Conflict & Terrorism 46 (5): 752–765. Weinberg, Lindsay. 2022. “Rethinking fairness: An interdisciplinary survey of critiques of hegemonic ml fairness approaches”. Journal of Artificial Intelligence Research 74 : 75–109. White, Ryen W and Eric Horvitz. 2015. “Belief dynamics and biases in web search”. ACM Transactions on Information Systems 33 (4): 1–46. Wong, Pak-Hang. 2020, June. “Democratizing Algorithmic Fairness”. Philosophy & Technology 33 (2): 225–244. Wängnerud, Lena. 2009, June. “Women in Parliaments: Descriptive and Substantive Representation”. Annual Review of Political Science 12 (1): 51–69. Zweig, Katharina A. 2017. “Watching the watchers: Epstein and Robertson’s „Search Engine Manipulation Effect“”. Authors": Tobias Rohrbach Mykola Makhortykh Maryna Sydorova Authors": Tobias Rohrbach Mykola Makhortykh Maryna Sydorova Tobias Rohrbach Tobias Rohrbach Mykola Makhortykh Mykola Makhortykh Maryna Sydorova Maryna Sydorova This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. available on arxiv available on arxiv