skip to main content
10.1145/3517745.3561450acmconferencesArticle/Chapter ViewAbstractPublication PagesimcConference Proceedingsconference-collections
research-article
Public Access

Measurement and analysis of implied identity in ad delivery optimization

Published: 25 October 2022 Publication History

Abstract

Online services such as Facebook and Google serve as a popular way by which users today are exposed to products, services, viewpoints, and opportunities. These services implement advertising platforms that enable precise targeting of platform users, and they optimize the delivery of ads to the subset of the targeted users predicted to be most receptive. Unfortunately, recent work has shown that such delivery can---often without the advertisers' knowledge---show ads to biased sets of users based only on the content of the ad. Such concerns are particularly acute for ads that contain pictures of people (e.g., job ads showing workers), as advertisers often select images to carefully convey their goals and values (e.g., to promote diversity in hiring). However, it remains unknown how ad delivery algorithms react to---and make delivery decisions based on---demographic features of people represented in such ad images. Here, we examine how one major advertising platform (Facebook) delivers ads that include pictures of people of varying ages, genders, and races. We develop techniques to isolate the effect of these demographic variables, using a combination of both stock photos and realistic synthetically-generated images of people. We find dramatic skews in who ultimately sees ads solely based on the demographics of the person in the ad. Ads are often delivered disproportionately to users similar to those pictured: images of Black people are shown more to Black users, and the age of the person pictured correlates positively with the age of the users to whom it is shown. But, this is not universal, and more complex effects emerge: older women see more images of children, while images of younger women are shown disproportionately to men aged 55 and older. These findings bring up novel technical, legal, and policy questions and underscore the need to better understand how platforms deliver ads today.

Supplementary Material

M4V File (315.m4v)
Presentation video

References

[1]
12 CFR § 202.4 (b) - Discouragement. https://www.law.cornell.edu/cfr/text/12/202.4.
[2]
24 CFR § 100.75 - Discriminatory Advertisements, Statements And Notices. https://www.law.cornell.edu/cfr/text/24/100.75.
[3]
Exhibit A - Programmatic Relief. https://nationalfairhousing.org/wp-content/uploads/2019/03/FINAL-Exhibit-A-3-18.pdf.
[4]
Facebook Plans Crackdown On Ad Targeting By Email Without Consent. https://techcrunch.com/2018/03/31/custom-audiences-certification/.
[5]
Title VII of the Civil Rights Act of 1964. https://www.law.cornell.edu/wex/title_vii.
[6]
47 USC § 230 - Protection For Private Blocking And Screening Of Offensive Material. https://www.law.cornell.edu/uscode/text/47/230.
[7]
About Advertising Objectives. https://www.facebook.com/business/help/517257078367892.
[8]
About Audiences For Special Ad Categories. https://www.facebook.com/business/help/2220749868045706.
[9]
About bid and budget pacing. https://www.facebook.com/business/help/571961726580148?id=2196356200683573.
[10]
About Customer Match. https://support.google.com/adwords/answer/6379332?hl=en.
[11]
Laura Alessandretti, Ulf Aslak, and Sune Lehmann. The scales of human mobility. Nature, 587(7834):402--407, 2020.
[12]
Muhammad Ali, Angelica Goetzen, Alan Mislove, Elissa Redmiles, and Piotr Sapiezynski. All things unequal: Measuring disparity of potentially harmful ads on facebook. In 6th Workshop on Technology and Consumer Protection (ConPro'22). IEEE, 2022.
[13]
Muhammad Ali, Piotr Sapiezynski, Miranda Bogen, Aleksandra Korolova, Alan Mislove, and Aaron Rieke. Discrimination through optimization: How facebook's ad delivery can lead to biased outcomes. Proceedings of the ACM on Human-Computer Interaction, 3(CSCW):1--30, 2019.
[14]
Muhammad Ali, Piotr Sapiezynski, Aleksandra Korolova, Alan Mislove, and Aaron Rieke. Ad delivery algorithms: The hidden arbiters of political messaging. In Proceedings of the 14th ACM International Conference on Web Search and Data Mining, pages 13--21, 2021.
[15]
Jisun An and Haewoon Kwak. Gender and racial diversity in commercial brands' advertising images on social media. In International Conference on Social Informatics, pages 79--94. Springer, 2019.
[16]
Jisun An and Ingmar Weber. Diversity in online advertising: a case study of 69 brands on social media. In International Conference on Social Informatics, pages 38--53. Springer, 2018.
[17]
Julia Angwin and Terry Parris Jr. Facebook lets advertisers exclude users by race. ProPublica, 2016. https://www.propublica.org/article/facebook-lets-advertisers-exclude-users-by-race.
[18]
Derek R Avery, Morela Hernandez, and Michelle R Hebl. Who's watching the race? racial salience in recruitment advertising 1. Journal of Applied Social Psychology, 34(1):146--161, 2004.
[19]
Sandra L Bem and Daryl J Bem. Does sex-biased job advertising "aid and abet" sex discrimination? 1. Journal of Applied Social Psychology, 3(1):6--18, 1973.
[20]
Lawrence Bowen and Jill Schmid. Minority presence and portrayal in mainstream magazine advertising: An update. Journalism & Mass Communication Quarterly, 74(1):134--146, 1997.
[21]
Scott Coltrane and Melinda Messineo. The perpetuation of subtle prejudice: Race and gender imagery in 1990s television advertising. Sex roles, 42(5):363--389, 2000.
[22]
Custom Audiences From Your Customer List. https://www.facebook.com/business/help/606443329504150.
[23]
Amit Datta, Anupam Datta, Jael Makagon, Deirdre K Mulligan, and Michael Carl Tschantz. Discrimination in online advertising: A multidisciplinary inquiry. In Conference on Fairness, Accountability and Transparency, pages 20--34. PMLR, 2018.
[24]
Amit Datta, Michael Carl Tschantz, and Anupam Datta. Automated experiments on ad privacy settings. Proceedings on privacy enhancing technologies, 2015(1):92--112, 2015.
[25]
Scott Detrow. What did cambridge analytica do during the 2016 election? NPR, 2018. https://www.npr.org/2018/03/20/595338116/what-did-cambridge-analytica-do-during-the-2016-election.
[26]
Emily Dreyfuss. Facebook changes its ad tech to stop discrimination. WIRED, 2019. https://www.wired.com/story/facebook-advertising-discrimination-settlement/.
[27]
Facebook: About The Delivery System: Ad Auctions. https://www.facebook.com/business/help/430291176997542.
[28]
Facebook Ad Targeting Options. https://www.facebook.com/business/ads/ad-targeting.
[29]
Facebook: Designated Market Areas for Ad Targeting. https://www.facebook.com/business/help/1501907550136620.
[30]
Facebook Makes Moves To Stop Discriminatory Ad Targeting. http://www.socialmediatoday.com/social-business/facebook-makes-moves-stop-discriminatory-ad-targeting.
[31]
Florida Voter Extract Disk File Layout. https://files.floridados.gov/media/704671/voter-extract-disk-file-layout.pdf.
[32]
Danielle Gaucher, Justin Friesen, and Aaron C Kay. Evidence that gendered wording in job advertisements exists and sustains gender inequality. Journal of personality and social psychology, 101(1):109, 2011.
[33]
Phillipa Gill, Vijay Erramilli, Augustin Chaintreau, Balachander Krishnamurthy, Konstantina Papagiannaki, and Pablo Rodriguez. Best paper-follow the money: understanding economics of online aggregation and advertising. In Proceedings of the 2013 Internet Measurement Conference (IMC'13), pages 141--148, 2013.
[34]
GitHub StyleGAN 2 Repository. https://github.com/NVlabs/stylegan2.
[35]
Jessica J Good, Julie A Woodzicka, and Lylan C Wingfield. The effects of gender stereotypic and counter-stereotypic textbook images on science performance. The Journal of Social Psychology, 150(2):132--147, 2010.
[36]
Good questions, real answers: How does facebook use machine learning to deliver ads? https://www.facebook.com/business/news/good-questions-real-answers-how-does-facebook-use-machine-learning-to-deliver-ads/.
[37]
Google: About Audience Targeting. https://support.google.com/google-ads/answer/2497941?hl=en.
[38]
Joshua Green and Sasha Issenberg. Inside The Trump Bunker, With Days To Go, 2016. https://www.bloomberg.com/news/articles/2016-10-27/inside-the-trump-bunker-with-12-days-to-go.
[39]
The Guardian. Cambridge Analytica: how did it turn clicks into votes? https://www.theguardian.com/news/2018/may/06/cambridge-analytica-how-turn-clicks-into-votes-christopher-wylie.
[40]
Help: Choosing A Special Ad Category. https://www.facebook.com/business/help/298000447747885.
[41]
Scott Highhouse, Sandra L Stierwalt, Peter Bachiochi, Allison E Elder, and Gwenith Fisher. Effects of advertised human resource management practices on attraction of african american applicants. Personnel Psychology, 52(2):425--442, 1999.
[42]
Basileal Imana, Aleksandra Korolova, and John Heidemann. Auditing for discrimination in algorithms delivering job ads. In Proceedings of the Web Conference 2021, pages 3767--3778, 2021.
[43]
Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and improving the image quality of stylegan. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2020.
[44]
Ava Kofman and Ariana Tobin. Facebook ads can still discriminate against women and older workers, despite a civil rights settlement. ProPublica, 2019. https://www.propublica.org/article/facebook-ads-can-still-discriminate-against-women-and-older-workers-despite-a-civil-rights-settlement.
[45]
Aleksandra Korolova. Facebook's Illusion Of Control Over Location-related Ad Targeting. Medium, December 2018. https://medium.com/@korolova/facebooks-illusion-of-control-over-location-related-ad-targeting-de7f865aee78.
[46]
Linkedin: Account Targeting. https://business.linkedin.com/marketing-solutions/ad-targeting/account-targeting.
[47]
Patrick F McKay and Derek R Avery. Warning! diversity recruitment could backfire. Journal of Management Inquiry, 14(4):330--336, 2005.
[48]
Madhumita Murgia. Time to turn off Facebook's digital fire hose. Financial Times, 2021.
[49]
Dmitry Nikitko. Learn direction in latent space. https://github.com/Puzer/stylegan-encoder/blob/master/Learn_direction_in_latent_space.ipynb.
[50]
North Carolina Laws, Chapter 163: Elections and Election Laws. https://www.ncleg.gov/Laws/GeneralStatuteSections/Chapter163.
[51]
North Carolina Voter Extract Disk File Layout. https://s3.amazonaws.com/dl.ncsbe.gov/data/layout_ncvoter.txt.
[52]
Panagiotis Papadopoulos, Nicolas Kourtellis, Pablo Rodriguez, and Nikolaos Laoutaris. If you are not paying for it, you are the product: How much do advertisers pay to reach you? In Proceedings of the 2017 Internet Measurement Conference, pages 142--156, 2017.
[53]
Lesley A Perkins, Kecia M Thomas, and Gail A Taylor. Advertising and recruitment: Marketing to minorities. Psychology & Marketing, 17(3):235--255, 2000.
[54]
Pinterest: Audience Targeting. https://help.pinterest.com/en/business/article/audience-targeting.
[55]
Angelisa C Plane, Elissa M Redmiles, Michelle L Mazurek, and Michael Carl Tschantz. Exploring user perceptions of discrimination in online targeted advertising. In 26th USENIX Security Symposium (USENIX Security 17), pages 935--951, 2017.
[56]
Filipe N. Ribeiro, Koustuv Saha, Mahmoudreza Babaei, Lucas Henrique, Johnnatan Messias, Fabrício Benevenuto Oana Goga, Krishna P. Gummadi, and Elissa M. Redmiles. On Microtargeting Socially Divisive Ads: A Case Study Of Russia-linked Ad Campaigns On Facebook. In Conference on Fairness, Accountability, and Transparency, pages 140--149, Atlanta, Georgia, USA, January 2019. ACM.
[57]
Richard Rothstein. The color of law: A forgotten history of how our government segregated America. Liveright Publishing, 2017.
[58]
Piotr Sapiezynski, Avijit Ghosh, Levi Kaplan, Alan Mislove, and Aaron Rieke. Algorithms that "Don't See Color": Measuring Biases in Lookalike and Special Ad Audiences. In Proceedings of the AAAI/ACM Conference on Artificial Intelligence, Ethics, and Society (AIES'22), Oxford, United Kingdom, Aug 2022.
[59]
Sefik Ilkin Serengil and Alper Ozpinar. Lightface: A hybrid deep face recognition framework. In 2020 Innovations in Intelligent Systems and Applications Conference (ASYU), pages 23--27. IEEE, 2020.
[60]
Shutterstock Stock Photos. https://www.shutterstock.com/.
[61]
Craig Silverman. How A Massive Facebook Scam Siphoned Millions Of Dollars From Unsuspecting Boomers. https://www.buzzfeednews.com/article/craigsilverman/facebook-subscription-trap-free-trial-scam-ads-inc.
[62]
Craig Silverman and Ryan Mac. Facebook Profits As Users Are Ripped Off By Scam Ads. https://www.buzzfeednews.com/article/craigsilverman/facebook-ad-scams-revenue-china-tiktok-vietnam.
[63]
Scott Spencer. An update on our political ads policy, 2019. https://blog.google/technology/ads/update-our-political-ads-policy/.
[64]
Thomas H Stevenson. A six-decade study of the portrayal of african americans in business print media: Trailing, mirroring, or shaping social change? Journal of Current Issues & Research in Advertising, 29(1):1--14, 2007.
[65]
The 2022 Florida Statues, Title IX: Electors and Elections, Chapter 97: Qualification and Registration of Electors. http://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&Search_String=&URL=0000-0099/0097/Sections/0097.0585.html.
[66]
Giridhari Venkatadri, Elena Lucherini, Piotr Sapiezyński, and Alan Mislove. Investigating sources of PII used in Facebook's targeted advertising. In Proceedings of the Privacy Enhancing Technologies Symposium (PETS'19), Stockholm, Sweden, Jul 2019.
[67]
Ellen L. Weintraub. Don't Abolish Political Ads On Social Media. Stop Microtargeting., November 2019. https://www.washingtonpost.com/opinions/2019/11/01/dont-abolish-political-ads-social-media-stop-microtargeting/.

Cited By

View all
  • (2024)AI Caramba!Leveraging AI for Effective Digital Relationship Marketing10.4018/979-8-3693-5340-0.ch011(309-352)Online publication date: 18-Oct-2024
  • (2024)#BigTech @Minors: Social Media Algorithms Personalize Minors’ Content After a Single Session, but Not for Their ProtectionSSRN Electronic Journal10.2139/ssrn.4674573Online publication date: 2024
  • (2024)Why Am I Seeing This Ad? The affordances and limits of automated user-level explanation in Meta’s advertising systemNew Media & Society10.1177/1461444824125179626:9(5130-5149)Online publication date: 30-Aug-2024
  • Show More Cited By

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Conferences
IMC '22: Proceedings of the 22nd ACM Internet Measurement Conference
October 2022
796 pages
ISBN:9781450392594
DOI:10.1145/3517745
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

Sponsors

In-Cooperation

  • USENIX Assoc: USENIX Assoc

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 25 October 2022

Permissions

Request permissions for this article.

Check for updates

Badges

  • Best Paper

Qualifiers

  • Research-article

Funding Sources

Conference

IMC '22
IMC '22: ACM Internet Measurement Conference
October 25 - 27, 2022
Nice, France

Acceptance Rates

Overall Acceptance Rate 277 of 1,083 submissions, 26%

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • Downloads (Last 12 months)343
  • Downloads (Last 6 weeks)41
Reflects downloads up to 20 Feb 2025

Other Metrics

Citations

Cited By

View all
  • (2024)AI Caramba!Leveraging AI for Effective Digital Relationship Marketing10.4018/979-8-3693-5340-0.ch011(309-352)Online publication date: 18-Oct-2024
  • (2024)#BigTech @Minors: Social Media Algorithms Personalize Minors’ Content After a Single Session, but Not for Their ProtectionSSRN Electronic Journal10.2139/ssrn.4674573Online publication date: 2024
  • (2024)Why Am I Seeing This Ad? The affordances and limits of automated user-level explanation in Meta’s advertising systemNew Media & Society10.1177/1461444824125179626:9(5130-5149)Online publication date: 30-Aug-2024
  • (2024)On the Use of Proxies in Political Ad TargetingProceedings of the ACM on Human-Computer Interaction10.1145/36869178:CSCW2(1-31)Online publication date: 8-Nov-2024
  • (2024)Beyond the Guidelines: Assessing Meta's Political Ad Moderation in the EUProceedings of the 2024 ACM on Internet Measurement Conference10.1145/3646547.3689020(480-487)Online publication date: 4-Nov-2024
  • (2024)Auditing for Racial Discrimination in the Delivery of Education AdsProceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency10.1145/3630106.3659041(2348-2361)Online publication date: 3-Jun-2024
  • (2023)A Normative Framework for Assessing the Information Curation Algorithms of the InternetPerspectives on Psychological Science10.1177/17456916231186779Online publication date: 27-Nov-2023
  • (2023)An Epistemic Lens on Algorithmic FairnessProceedings of the 3rd ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization10.1145/3617694.3623248(1-10)Online publication date: 30-Oct-2023
  • (2023)Sociotechnical Audits: Broadening the Algorithm Auditing Lens to Investigate Targeted AdvertisingProceedings of the ACM on Human-Computer Interaction10.1145/36102097:CSCW2(1-37)Online publication date: 4-Oct-2023
  • (2023)Discrimination through Image Selection by Job Advertisers on FacebookProceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency10.1145/3593013.3594115(1772-1788)Online publication date: 12-Jun-2023

View Options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

Login options

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media