Dark nudge
Also known as: Sludge nudge
A nudge engineered to serve the architect's interest at the expense of the person being nudged.
What it means
A dark nudge is a choice-architecture technique deployed to exploit rather than help — steering people toward decisions that benefit the designer (a firm or institution) while harming the chooser's own interests. It uses the same levers as benign nudges (defaults, framing, salience, social proof, scarcity cues) but inverts the welfare test that distinguishes legitimate nudging from manipulation. Dark nudges overlap with dark patterns (their digital-interface form) and with sludge (harmful friction), and the family is sometimes grouped as the unethical end of the behavioral toolkit. The diagnostic question is whose interest the steer serves and whether the person would object if they saw it clearly. It matters because the techniques of behavioral science are ethically neutral instruments: the same default that rescues a saver can trap a subscriber, so naming dark nudges sharpens the line practitioners and regulators must police.
What makes a nudge dark
The welfare test the definition names has two parts that are easy to run together. One is about outcome: whose interest the steer serves. The other, developed by Sunstein, is about process: an influence is manipulative when it works by bypassing a person's capacity for reflective, deliberative choice rather than engaging it. The two usually travel together, but the process test is what makes a technique feel wrong even when the money at stake is small. A pre-ticked insurance box does not persuade the buyer that the cover is worth having; it profits from the fact that she never forms a judgement at all. The steer is hidden by design, and would collapse the moment the chooser saw it plainly and asked whether she actually agreed.
What the evidence shows
The manipulability is measured, not assumed. In two large census-weighted experiments, Luguri and Strahilevitz walked participants through a sign-up flow: 11 percent of a control group accepted a dubious identity-protection service, rising to 26 percent under mild dark patterns and 42 percent under aggressive ones. A single confusingly worded question, a double negative, lifted acceptance from 19 to 33 percent. Tripling the service's price barely moved take-up, which tells you the choice was being driven by the interface rather than by value. A separate automated crawl by Mathur and colleagues found roughly 1,800 dark-pattern instances across 11,000 shopping sites, concentrated among the most popular retailers. The tactics are neither rare nor confined to the fringe.
Where it shows up
Beyond checkout pages, the same inversion appears wherever an institution both shapes the information and profits from the choice. Petticrew and colleagues documented it in alcohol-industry health bodies, whose materials buried breast-cancer risk deep in long menus, paired warnings with cheerful drinking imagery, and tweeted that people 'commonly give up' on Dry January, steering readers away from the very message the pages claimed to carry. Subscription services concentrate the friction at cancellation: one tap to join, a phone call to leave. Gambling interfaces dress losses up as wins with celebratory sound and leave autoplay switched on by default. What unites the cases is that the harmful design sits exactly where the chooser's interest and the architect's diverge.
Where the concept strains
The label is sharper in principle than in application. Persuasion and manipulation lie on a continuum, and reasonable people disagree about where a scarcity cue stops being useful information and becomes manufactured pressure; the same 'two rooms left' banner is honest if true and a dark nudge if invented. Intent is hard to prove and easy to deny, which is why regulators lean on observable features, hidden costs, obstructed exits, tricked consent, rather than on motive. The evidence also complicates the moral picture: aggressive tactics provoke a backlash that mild ones escape, so the most manipulative designs are not always the most profitable, and the quiet ones may do the most cumulative harm precisely because almost no one objects to them.
Examples
Pre-ticking a costly add-on and burying the opt-out in small print is a dark nudge — a default exploited to extract money the customer never meant to spend.
A booking site flashes 'only two rooms left, eleven people viewing this' to hurry the decision. The scarcity is manufactured, and it serves the seller's margin, not the traveller's judgement.
A free trial you can start with one tap but can only cancel by phoning during office hours — friction placed precisely where it profits the company and costs the customer.
A cookie banner makes 'Accept all' one bright tap, while refusing means expanding a menu and switching twenty vendors off by hand. The friction is aimed only at the privacy-protecting choice.
A newsletter pop-up offers two buttons: 'Yes, I want to save money' and 'No, I like paying full price.' The loaded wording shames the decline instead of informing it.
First described in Behavioral ethics discourse (2010s).
Key references
- Luguri, J., & Strahilevitz, L. J. (2021). Shining a Light on Dark Patterns. Journal of Legal Analysis, 13(1), 43-109. doi.org/10.1093/jla/laaa006
- Petticrew, M., Maani, N., Pettigrew, L., Rutter, H., & Van Schalkwyk, M. C. (2020). Dark Nudges and Sludge in Big Alcohol: Behavioral Economics, Cognitive Biases, and Alcohol Industry Corporate Social Responsibility. The Milbank Quarterly, 98(4), 1290-1328. doi.org/10.1111/1468-0009.12475
- Mathur, A., Acar, G., Friedman, M. J., Lucherini, E., Mayer, J., Chetty, M., & Narayanan, A. (2019). Dark Patterns at Scale: Findings from a Crawl of 11K Shopping Websites. Proceedings of the ACM on Human-Computer Interaction, 3(CSCW), Article 81. doi.org/10.1145/3359183
- Thaler, R. H. (2018). Nudge, not sludge. Science, 361(6401), 431. doi.org/10.1126/science.aau9241
- Sunstein, C. R. (2016). Fifty Shades of Manipulation. Journal of Marketing Behavior, 1(3-4), 213-244. doi.org/10.1561/107.00000014