Who Designed That Default? You Did.

When Microsoft set "reply all" as the default behavior for email responses in Outlook years ago, they probably ran some calculation about which default produced fewer complaints. The consequence of that choice is not neutral. When millions of people click reply without thinking, they reply to everyone on the thread. Most of them would have chosen reply if they had been paying full attention. Most of them were not paying full attention. The default resolved their inattention in a particular direction.

That is choice architecture. Richard Thaler and Cass Sunstein wrote "Nudge" in 2008 and made the concept accessible to a wider audience than behavioral economics had previously reached. Their core argument is that the design of any choice environment, the layout, the defaults, the ordering, the framing, shapes the decisions people make without limiting their options. People remain free to choose otherwise. Most do not. The architecture does the work.

Thaler and Sunstein used pension enrollment as one of their cleaner examples. When employees have to actively opt into a pension scheme, enrollment rates tend to be lower than when enrollment is automatic and employees have to opt out. The same logic applies to the contribution rate. If the default is set to, say, 6 percent of salary, most employees stay at 6 percent. Some had intended to contribute more. Some had intended to contribute less. But inertia and the cognitive cost of changing a default mean that the default itself carries enormous behavioral weight. The employer or plan designer who set that default made a consequential decision even if they were not thinking of it that way.

IS designers make these decisions constantly. Every enterprise system ships with defaults for notification frequency, privacy settings, sharing permissions, data retention periods, and dozens of other choices that users technically can adjust and practically never do. A collaboration tool that defaults to public channels rather than private ones is making a choice about organizational transparency. An analytics dashboard that defaults to showing last-quarter data is making a choice about which time horizon managers use when evaluating performance. A mobile app that defaults to sending location data is making a choice about data collection. The user can change any of these. The user usually does not.

What I find genuinely uncomfortable about this is that IS designers are making these choices whether or not they think of themselves as doing so. There is no neutral default. Choosing a setting means someone gets nudged in a particular direction. Not choosing, and leaving whatever the system default is, means the same thing. The decision to not think carefully about defaults is itself a design decision, and it has consequences for user behavior at scale.

The ethics of this get complicated fast. Thaler and Sunstein were generally talking about nudges in the direction of outcomes that serve the person being nudged: saving more for retirement, eating healthier options in a cafeteria, enrolling in organ donation. These nudges have defenders and critics, but the intent is at least oriented toward the user's own stated interests. The mechanism is identical when the nudge serves the platform's interests rather than the user's.

A subscription service that makes cancellation require navigating five screens while sign-up takes one click is using the same tools. A cookie consent banner that makes "accept all" a bright green button while "manage preferences" requires reading through a gray menu is using the same tools. An e-commerce site that defaults to the more expensive shipping option during checkout is using the same tools. The behavioral science behind these patterns is identical to the pension enrollment research. The difference is whose interests the nudge serves.

There is a category of these practices, increasingly called dark patterns, where interface design exploits cognitive biases in ways that are arguably deceptive rather than helpful. Regulators in Europe have started paying attention to this. The EU's General Data Protection Regulation and the Digital Services Act both address, in different ways, design practices that manipulate user choices. I am not going to cite specific articles here because the regulatory landscape is shifting and the specifics are contested, but the direction is clear: regulators are beginning to treat design choices as consequential in a legal sense, not just an ethical one.

The IS angle on this that I keep thinking about is the gap between what IS designers are trained to do and what they are actually doing when they make these choices. Software engineering education focuses on functionality, performance, and correctness. Interaction design education focuses on usability and user goals. Neither tradition has a well-developed vocabulary for thinking about behavioral consequences at population scale. When I design a default, I am making a choice that will affect not one user in a usability test but thousands or millions of users in real conditions. The tools I need to reason about that are drawn from behavioral economics and public policy, not from system design.

Thaler and Sunstein argued for what they called "libertarian paternalism": preserving freedom of choice while arranging the environment to make the better choice easier. That is a defensible principle in the context of helping people save more for retirement. It is a much more contested principle when applied to helping a platform extract more revenue. Both are nudges. Both use defaults. The ethical question is not about the mechanism but about who benefits.

I do not think IS designers are generally trying to manipulate users. I do think most IS designers are not thinking carefully about defaults as behavioral choices with population-level consequences. The field has the concepts to do better. Behavioral IS is a recognized research area. The work from Thaler, Sunstein, and the behavioral economics tradition is well documented and accessible. The gap is mostly between what the research knows and what gets applied in practice when someone is shipping a product.