Lior Strahilevitz Shines a Light on Dark Patterns

By TAP Staff Blogger

Posted on May 27, 2021


Dark patterns are user interfaces whose designers knowingly confuse users, make it difficult for users to express their actual preferences, or manipulate users into taking certain actions. They typically exploit cognitive biases and prompt online consumers to purchase goods and services that they do not want or to reveal personal information they would prefer not to disclose. This article provides the first public evidence of the power of dark patterns.
- “Shining a Light on Dark Patterns” by Professor Lior Strahilevitz and Jamie Luguri


The term “dark patterns” refers to various tactics companies employ online and in gaming and smartphone apps to manipulate their customers for the companies’ benefit. It might be to upsell products, gather customer personal information, or keep a customer on a mailing list. Below are a few of the schemes that consumers must navigate online in order to achieve their desired objectives.

  • “Confirm-shaming” – the only way a customer can unsubscribe from a list or close a web page is the click the button that says something along the lines of, “You will lose valuable opportunities to save money [or learn the latest] if you choose to unsubscribe.”
  • “Obstruction” – a customer has to wade through options for unsubscribing, such as “receive emails once a week” or “receive fewer emails” and then the user will have to click through to a subsequent page to get the option to “receive no more emails”.
  • “Trick Questions” – prompts that make it hard for tech-savvy consumers to figure out how to accomplish their objective.
  • “Roach Motel” – the tactic of making it very easy to sign up for a service; however, to cancel the service, the customer has to make a phone call or send a letter.
  • “Sneak into the Cart” – an item appears in an online shopping cart that the customer did not select.
  • “Nagging” – presenting the consumer the choice to subscribe to a service with the only options being “yes” and “not now”.
  • “Bait and Switch” – the customer agrees to purchase an item and then is shown a barrage of ads for things the customer does not want.
  • “Falso or Misleading Messages” – messages that display information about demand for products or testimonials. (e.g., “Anna from Anchorage just purchased this jacket.” Or a random number generator being used to display “XX items just sold.”)

In their recent paper, “Shining a Light on Dark Patterns,” law professor Lior Strahilevitz and Jamie Luguri, both of the University of Chicago Law School, share findings from two large-scale experiments in which consumers were exposed to dark patterns. The authors found that “users exposed to mild dark patterns were more than twice as likely to sign up for a dubious service as those assigned to the control group, and users in the aggressive dark pattern condition were almost four times as likely to subscribe.”


Below are a few excerpts from “Shining a Light on Dark Patterns” by Lior Strahilevitz and Jamie Luguri:


Our bottom line is that dark patterns are strikingly effective in getting consumers to do what they would not do when confronted with more neutral user interfaces.


The Scale of E-Commerce


We said at the outset that dark patterns are different from other forms of dodgy business practices because of the scale of e-commerce. There may be poetic justice in the fact that this very scale presents an opportunity for creative legal regulators. It is exceedingly difficult to figure out whether a door-to-door salesperson’s least savory tactics significantly affected the chances of a purchase—was the verbal sleight of hand material or incidental? Who knows? But with e-commerce, firms run thousands of consumers through identical interfaces at a reasonable cost and see how small software tweaks might alter user behavior. Social scientists working in academia or for the government can do this too; we just haven’t done so before today. Now that scholars can test dark patterns, we can isolate causation in a way that’s heretofore been impossible in the brick-and-mortar world. Unlike brick-and-mortar manipulation, dark patterns are hiding in plain sight, operate on a massive scale, and are relatively easy to detect. Those facts strengthen the case further for the legal system to address their proliferation.


Are Dark Patterns Unlawful?


There are several plausible legal hooks that could be used to curtail the use of dark patterns by U.S. firms in e-commerce. First, the Federal Trade Commission Act restricts the use of unfair or deceptive practices in interstate trade, providing the Commission with a mandate to regulate and restrict such conduct. Second, state unfair competition laws include similar frameworks. Finally, there is a broad question about whether consumer consent that is procured in a process that employs highly effective dark patterns should be voidable, which would entitle consumers to various remedies available under contract law and which could open up liability for firms that engage in various activities (for example, engaging in surveillance or processing biometric information) without having first obtained appropriate consumer consent.


Concluding Thoughts


The problem we identify here, then, is both an old problem and a new one. Companies have long manipulated consumers through vivid images, clever turns of phrase, attractive spokesmodels, or pleasant odors and color schemes in stores. …


The online environment is different. It is perhaps only a difference of degree, but the degrees are very large. Through A-B testing, firms now have opportunities to refine and perfect dark patterns that their Mad Men-era counterparts could have never imagined. By running tens of thousands of consumers through interfaces that were identical in every respect but one, firms can determine exactly which interface, which text, which juxtapositions, and which graphics maximize revenues. What was once an art is now a science. As a result, consumers’ ability to defend themselves has degraded. The trend toward personalization could make it even easier to weaponize dark patterns against consumers.


Read the full article: “Shining a Light on Dark Patterns” by Lior Strahilevitz and Jamie Luguri (Journal of Legal Analysis, March 29, 2021).


More from Professor Strahilevitz on Dark Patterns


In April, the Federal Trade Commission (FTC) held an online workshop to examine what dark patterns are and how they affect consumers and the marketplace. During the “Bringing Dark Patterns to Light” workshop, Professor Strahilevitz presented findings from his paper with Jamie Luguri and he was a panelist on the session that addressed ‘Potential Strategies for Dealing with Dark Patterns’.


Below are a few takeaways from Professor Strahilevitz’ testimony at the FTC’s workshop on “Bringing Dark Patterns to Light”:


Key Takeaways from Study


As we think about normative takeaways, from our perspective, the data indicates that it's these mild dark patterns that are most insidious. They significantly increase acceptance rates for dubious service that we were offering, without substantially generating any consumer backlash. The less educated are more vulnerable to dark patterns. And when we think about standard economic models of consumer behavior, consumers are supposed to be really responsive to price increases when they decide whether to buy something. And yet what we find is that they're so much more responsive to dark patterns than they are to reductions in price, even if those reductions are on the order of $30 a month.


We don't claim to be the first social scientists who've ever used a randomization to run these kinds of studies to figure out the effectiveness of dark patterns. In fact, I suspect that a lot of social scientists working in-house for e-commerce companies have been running studies exactly like the ones that Jamie and I ran for years. We're just the first to publish our results and to share this data with the world. But we think it's precisely because social scientists working in-house for tech companies have done studies like ours on their own and seen how effective these dark patterns are, we think that explains the proliferation of dark patterns on various electronic platforms.


Consumer Learning


I do think consumers learn to a degree about dark patterns. And one of the results from our paper that I referenced earlier suggested that consumers at least have built up a defense mechanism against one type of dark pattern, which is the, if you don't act within 60 seconds this great deal is going to disappear. I think consumers over the years have come to associate that kind of sales practice with sort of dodgy, fly-by-night operators. And it may well be that, when consumers see that kind of blatant manipulative behavior, they react negatively, and it makes them less likely to purchase. Our data is consistent with that story.


On the other hand, consumer learning can actually make the problem worse rather than better. So I'll give you an example, which is nagging. A lot of us have gotten nagging prompts on our smartphones, with an app repeatedly asking us to share our location or repeatedly asking us to authorize push notifications. And if you say yes, it stops nagging you. And if you say maybe later, which is the only option you've got-- no, is not an option-- It'll ask you a week later and a week later and a week later, until you say yes. And consumers get worn down by this.


And eventually, they just say, well, they're going to get me eventually, so I may as well go ahead and agree to push notifications now. So sometimes then, consumers learn, but what they're learning is a learned helplessness. And that's really an example that cries out for aggressive government regulation. Because to characterize consumers as having consented after repeated nagging to do something that they don't want to do is to sort of make a mockery of the idea of consent.


Video and transcript of the FTC’s workshop is available at “Bringing Dark Patterns to Light”.


Read more:
Professor Strahilevitz and his paper with Jamie Luguri is discussed in the following articles: