Friction
The small costs and hassles that quietly throttle behavior.
What it means
Friction refers to any effort, delay, or hassle that sits between a person and an action, and a central insight is that even small frictions can reduce how often people complete a behavior by amounts out of proportion to their objective cost. The mechanism is that minor barriers interact with present bias, limited attention, and inertia: a hurdle now can loom larger than a meaningful benefit later, so people abandon or postpone the action. This makes friction a double-edged lever, since removing it is often a cheap way to raise a desired behavior while adding it can deter unwanted ones, a principle related to Lewin's channel factors. Not all friction is bad: a little 'good friction' or a cooling-off step can protect people from impulsive choices, whereas friction that obstructs a person's own interest becomes sludge. It matters across product design, public services, and policy, though field trials show gains from shaving a click or a form field are often smaller than laboratory results suggest.
Why it happens
The puzzle is not that barriers deter action but that trivial ones deter it so much. Three ordinary features of human judgment combine to magnify a small cost. The first is present bias: an effort or delay is paid now, in full, while the benefit it stands in front of arrives later and is discounted, so a two-minute form can outweigh a reward worth far more. The second is limited attention. Every extra step is a fresh moment at which attention can wander, a question can arise, or a competing demand can intervene, and each of those moments carries a chance of the task being dropped. Stack several steps in sequence and the chances of abandonment compound. The third is inertia, the tendency to stay on whatever path requires no decision, which means the option reached by doing nothing has a standing advantage over the one that requires an act of will. Seen this way, friction is where good intentions leak. People who genuinely mean to enrol, apply, switch, or cancel often fail to, not because they reconsidered but because a small obstacle arrived at a moment when motivation was low and attention was elsewhere. The barrier does not have to be large to intercept a fragile intention; it only has to be there.
The original demonstration
The idea has a clear lineage. Kurt Lewin, working in the 1940s, argued that behavior is governed less by the strength of a person's attitudes than by small features of the immediate situation, which he called channel factors: minor details that open or close the path to an act. A person poised to do something can be tipped either way by a facilitating cue or a small obstruction. The cleanest early test came from Howard Leventhal, Robert Singer, and Susan Jones in 1965. Studying tetanus vaccination among university students, they crossed the strength of a fear-arousing health message with whether students also received a concrete plan of action: a map marking the campus health service and a prompt to look over their weekly schedule and settle on a time to go. The message's intensity moved reported intentions but barely moved behavior. The action plan moved behavior sharply, raising the share who actually got vaccinated from roughly 3 percent to about 28 percent, and it did so at every level of fear. A small situational scaffold, not the emotional force of the appeal, decided who acted. Lee Ross and Richard Nisbett later drew on such findings to popularize channel factors as a general lesson: tiny arrangements of the situation routinely outweigh dispositions in predicting what people do.
What the evidence shows
The strongest modern evidence concerns two structural forms of friction: defaults and administrative burden. On defaults, Eric Johnson and Daniel Goldstein showed in 2003 that countries operating presumed-consent (opt-out) organ-donation systems recorded effective consent rates above 90 percent, against well under a third in otherwise comparable opt-in countries, with a controlled experiment reproducing the same gap. The only difference was which choice the paperwork treated as automatic. On administrative burden, Eric Bettinger and colleagues ran a field experiment in which tax preparers helped low-income families complete the United States college-aid form by pre-filling it from their returns; the completion help raised college enrolment among graduating seniors, while giving matched families the same aid information without the completion help did essentially nothing. Saurabh Bhargava and Dayanand Manoli found, in an experiment with tens of thousands of eligible non-claimants of a tax credit, that a resent and simplified notice which made the benefit more salient produced substantial additional claiming, though messages aimed at lowering the perceived stigma and hassle of applying did not. The honest complication is that friction-based nudges shrink markedly once they leave the laboratory. Stefano DellaVigna and Elizabeth Linos assembled 126 field trials run by two large government nudge units, covering some 23 million people, and found the average intervention moved behavior by 1.4 percentage points, against 8.7 percentage points in the published academic literature. About seventy percent of that six-fold gap was traced to selective publication and the low statistical power of the smaller academic studies. The meta-analytic picture is contested in the same way. A 2022 synthesis by Stephanie Mertens and colleagues reported a medium overall effect of choice-architecture interventions, but a reanalysis of the same data by Maximilian Maier and colleagues found that once publication bias was corrected the overall effect was no longer distinguishable from zero, with defaults among the few categories whose evidence survived. The pattern that emerges is coherent: the frictions that reliably move behavior are the structural ones baked into a default or a required step, while lighter informational and reminder nudges are real but modest and easily overstated.
Using it in practice
Two moves follow from the same lever. To raise a wanted behavior, subtract friction: pre-fill forms from data already held, cut steps and required fields, drop mandatory attachments, and let the desired option be the one reached by doing nothing. To slow an unwanted or high-stakes one, add friction: a confirmation step, a mandatory pause, a cooling-off period, or a small screening ordeal. The formal version of the first move is the sludge audit proposed by Cass Sunstein, in which an organization maps every step a person must pass through and tallies the frictions, then removes those that serve no legitimate purpose. The same audit exposes the second move's abuse, since firms routinely engineer asymmetry, making sign-up a single tap while cancellation runs through a maze. The governing distinction is not how large the barrier is but whose interest it serves. Friction that protects the chooser's own goals, such as a waiting period before an irreversible decision, is good friction; friction inserted to obstruct people for someone else's benefit is sludge. In practice the highest-yield question is rarely how to persuade someone more forcefully, but which step in the process is quietly shedding people who already meant to act.
Limits and caveats
Friction is an umbrella term, not a single measurable quantity, and treating it as one invites overreach. Its bite depends heavily on motivation: behaviors people badly want are robust to considerable hassle, while marginal ones collapse under a trivial step, so the same added click deters almost no one in one setting and half the traffic in another. Effects are therefore heterogeneous and, as the field evidence shows, typically smaller and less certain than a striking laboratory demonstration implies. The good-versus-bad distinction is a value judgment about whose ends a barrier serves, not a property of the barrier itself, and reasonable people disagree about where a protective cooling-off period shades into paternalistic obstruction. Measurement is awkward too, because friction is usually inferred from the difference in completion rates between two otherwise identical processes, an inference that other simultaneous changes can contaminate. And removal is not always benign: a frictionless path can carry people into choices they later regret, which is precisely why some friction is worth keeping. The defensible claim is narrower than the slogan. Structural frictions move behavior dependably; the rest is real but modest and context-bound.
Examples
Requiring one extra click or field at checkout measurably drops completion rates.
A subscription service lets a customer sign up with one tap but routes cancellation through five screens and a phone call during limited hours. The asymmetry is deliberate sludge, and it inflates retention without changing anything about the service itself.
A benefits office pre-fills an application from records it already holds and shortens the form from twelve pages to one. Take-up among eligible households rises even though the eligibility rules and the size of the benefit are unchanged.
A retailer inserts a brief review-your-order pause and a confirmation step before unusually large purchases. Impulse buying and later returns both fall, a case of friction added on purpose to protect the buyer rather than to trap them.
A clinic switches vaccination appointments from something patients must book to a pre-assigned slot they can decline. Attendance climbs, and the exact wording of the reminder matters far less than the fact that declining now requires the extra step.
First described in Behavioral design principle; see Lewin's channel factors.
Key references
- Leventhal, H., Singer, R., & Jones, S. (1965). Effects of fear and specificity of recommendation upon attitudes and behavior. Journal of Personality and Social Psychology, 2(1), 20-29. doi.org/10.1037/h0022089
- Johnson, E. J., & Goldstein, D. (2003). Do defaults save lives? Science, 302(5649), 1338-1339. doi.org/10.1126/science.1091721
- Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2012). The role of application assistance and information in college decisions: Results from the H&R Block FAFSA experiment. The Quarterly Journal of Economics, 127(3), 1205-1242. doi.org/10.1093/qje/qjs017
- Bhargava, S., & Manoli, D. (2015). Psychological frictions and the incomplete take-up of social benefits: Evidence from an IRS field experiment. American Economic Review, 105(11), 3489-3529. doi.org/10.1257/aer.20121493
- Sunstein, C. R. (2022). Sludge audits. Behavioural Public Policy, 6(4), 654-673. doi.org/10.1017/bpp.2019.32
- DellaVigna, S., & Linos, E. (2022). RCTs to scale: Comprehensive evidence from two nudge units. Econometrica, 90(1), 81-116. doi.org/10.3982/ECTA18709
- Mertens, S., Herberz, M., Hahnel, U. J. J., & Brosch, T. (2022). The effectiveness of nudging: A meta-analysis of choice architecture interventions across behavioral domains. Proceedings of the National Academy of Sciences, 119(1), e2107346118. doi.org/10.1073/pnas.2107346118
- Maier, M., Bartos, F., Stanley, T. D., Shanks, D. R., Harris, A. J. L., & Wagenmakers, E.-J. (2022). No evidence for nudging after adjusting for publication bias. Proceedings of the National Academy of Sciences, 119(31), e2200300119. doi.org/10.1073/pnas.2200300119
Where this comes up
- More Than Motivation: Understanding Executive Functioning Skills Using Behavioral ScienceIncreasing your Executive Functioning Skills can help create a more balanced and sustainable life.