Choice Architecture

What It is and how it can influence your decisions.

This post incorporates and expands on content from one I previously published on Psychology Today.

Image by Gerd Altmann from Pixabay

Think about the last time you signed up for a web service—maybe an online store like Amazon, a banking site like Chase, or a digital subscription like The New York Times. At some point during the process, you were likely asked whether you wanted to receive emails from the company or its partners.

On some sites, the default option had already been selected, granting permission unless you actively unchecked the box (opt-out). On others, the default was set to reject promotional emails unless you explicitly checked the box (opt-in).

Either way, you had the freedom to choose¹, but the website designer had already structured the decision environment by pre-determining the starting point (i.e., the default choice). If you didn’t override the default—whether due to choice, inattention, or uncertainty—you effectively accepted it. This is a simple example of choice architecture in action: the way choices are structured can shape how decisions unfold (Thaler & Sunstein, 2021).

The choice architect decides which options are presented, how they are framed, and in which order they are presented to a decision maker (often us). The architecture itself doesn’t dictate behavior, per se, but it can subtly influence which option we select, sometimes without us realizing it. Choice architecture is not inherently prescriptive—it simply refers to the way options are presented for a decision maker. However, when design choices are made with the explicit goal of steering people toward a particular decision, that moves beyond general choice architecture and into nudging².

In this post, I’ll explore how default settings, option framing, and decision complexity shape choices, the extent to which choice architecture can be neutral, and the ethical considerations that arise when choice architects design environments that influence decision making.

The Default Option Can Influence Decisions

If you run a website and require people to opt out of receiving notifications, you’re likely to see more people accept notifications than if the default is to not receive them. If all employees are automatically enrolled in the company’s retirement savings plan unless they opt out, you’re likely to see more people enroll than if they are required to opt into the plan.

These are simple examples of choice architecture that involve the default decision choice. In such environments, when two or more options are offered, one can be chosen as the pre-selected option. Unless an alternative choice is made, the pre-selected option is automatically chosen.

Defaults influence people’s choices because they tend go with the path of least resistance. Opting out requires effort—whether that means unchecking a box, making a phone call, or filling out a form—and many people simply stick with whatever is pre-selected, either because they have no strong preference or because they failed to notice that they had to consciously decide to reject the default.

Research shows default settings can influence behavior—sometimes more than we realize. And these defaults are rarely random. Choice architects structure decisions the way they do for many reasons, including logistical simplicity, efficiency, or user convenience. In many cases, defaults are intentionally set to guide behavior in a specific direction (Mills & Sætra, 2021).

Consider a few well-documented examples:

• Organ donation rates tends to be higher in countries where donation is the default option rather than requiring explicit opt-in (Shepard et al., 2014), though this conclusion has been called into question (Molina-Pérez et al., 2022).

• Retailers and digital platforms structure purchasing decisions in ways that encourage customers to accept pre-selected options (e.g., auto-renewing subscriptions or default add-ons at checkout; Ericson, 2020).

• Workplaces use automatic enrollment in retirement savings plans to boost participation rates without requiring employees to take action (Cribb & Emmerson, 2021).

Well-designed defaults can make decision making easier, especially when the default aligns with what most people would want anyway. But not all defaults serve the decision maker’s best interest. Some are designed to benefit the choice architect, regardless of their alignment with decision maker interests. Consider:

• Subscription traps, where free trials automatically convert into paid plans unless you cancel.

• Online shopping checkouts, where optional add-ons (like insurance or donations) are pre-selected.

• Privacy settings, where data-sharing is often enabled by default rather than requiring explicit consent.

These cases illustrate a key ethical issue in choice architecture: defaults don’t just guide decisions—they reflect the priorities of the choice architect who sets them. Sometimes there’s overlap between the interests of the choice architect and decision maker, but that isn’t always (or perhaps even often) the case.

In their defense, Thaler and Sunstein (2021) argue that defaults should be set in a way that aligns with the decision maker’s best interests—a concept known as libertarian paternalism. The idea is that people should be nudged toward beneficial choices while still having the freedom to choose otherwise. But this assumes the choice architect knows what’s best for everyone—a premise that isn’t always true.

For example, in 2019, some California cities switched to 100% renewable energy as the default utility plan (Roth, 2019). While this decision was framed as a pro-environment move and thus in everyone’s best interests, it also raised utility costs by 9%. Many residents, particularly those struggling financially, might have preferred a cheaper option, but because switching required effort, many simply stuck with the default³.

This highlights a fundamental problem: Defaults don’t just shape decisions—they embed the values of the people who determine what those defaults are. But defaults aren’t the only way choice architects influence behavior. Even when people actively make a choice, the way the choice is presented—its framing—can significantly shape how people perceive their options.

How Information Is Framed Can Influence Decisions

Choice architecture doesn’t just shape what options are available—it also shapes how those options are presented. The same decision can feel different depending on how it’s framed.

For example, consider a surgeon who tells a patient that a procedure has a 90% survival rate. That sounds reassuring. But if the same surgeon instead says there’s a 10% chance of death, the procedure suddenly feels riskier—even though the numbers haven’t changed.

This is an example of the framing effect (Kahneman & Tversky, 1979), where people’s choices shift depending on whether an option is presented in terms of gains or losses (and a key part of prospect theory). People are more likely to accept risk when avoiding a loss but tend to play it safe when securing a gain. And since choice architects control how options are framed, they can influence decisions without changing the actual choices available.

Framing effects appear in a wide range of decisions, including:

• Healthcare choices – Patients are more likely to accept a treatment when told it has a 90% success rate than when they hear there’s a 10% failure rate (Tversky & Kahneman, 1981).

• Financial decisions – Investors take fewer risks when potential losses are emphasized rather than potential gains (Nakasato et al., 2006).

• Consumer behavior – “Buy one, get one free” often feels better than “50% off when you buy two items,” even though the discount is the same (Yang & Mattila, 2020).

• Public policy & messaging – Framing environmental policies in terms of health benefits and pollution prevention (rather than restrictions on businesses) increases public support (Mossler et al., 2017).

Because framing can shape whether people see an option as safe or risky, beneficial or costly, choice architects can use it to steer decisions without removing any options⁴.

Framing is not inherently problematic—it can be used to help people make better-informed decisions. For example, graphic warning labels on cigarette packs make health risks more salient, and calorie counts on menus help diners understand nutritional trade-offs before ordering.

But framing can also manipulate choices by trying to exploit cognitive biases, such as companies emphasizing “free trials” while downplaying the auto-renewal fees that kick in later, or websites warning you about “losing access to premium features” rather than simply confirming a subscription cancellation. This highlights a fundamental ethical concern: framing doesn’t just shape decisions—it also reflects the priorities of those designing the choices. While it can help decision makers, it can just as easily be used to manipulate them.

But even when choice architects act with the best of intentions, their interventions can still backfire. In some cases, the problem isn’t how choices are framed—it’s that there are simply too many choices.

How Too Many Choices Can Backfire

Choice architecture can also be used in the context of how many options are presented in the first place. We tend to think that more options lead to better decisions, but in reality, too many options can create choice overload, leading to decision fatigue, inaction, or even regret (Buturak & Evren, 2015; Park & Jang, 2013).

In an ideal world, people would carefully weigh all their options, consider trade-offs, and make the best possible choice. But real-world decisions don’t work that way (as I’ve discussed previously). When faced with too many options, people often struggle to compare them effectively. In such cases they may make worse decisions or avoid making one at all.

Consider, for example, the jam study (Iyengar & Lepper, 2000), which found that when shoppers were presented with six different flavors of jam, they were 10x more likely to make a purchase than those presented with 24 flavors. While more options attracted more attention, fewer options led to higher sales—because it was easier for customers to decide.

This same pattern shows up in other domains too. For example:

• Retirement plans – People are less likely to enroll in 401(k) plans when they’re given too many investment options (Iyengar et al., 2004).

• Healthcare decisions – Patients facing a long list of insurance options often feel overwhelmed and default to suboptimal plans or avoid making a selection entirely (Biener & Zou, 2024).

• Consumer purchases – Having more options initially increases purchasing likelihood but eventually leads to a drop-off in decision making due to overload (Shah & Wolford, 2007).

Even though we assume that having more options gives us greater control, too many options often leave us feeling dissatisfied, anxious, or uncertain.

Since choice overload is a real problem, choice architects often simplify decisions by limiting options or structuring them in a way that reduces complexity. Some effective approaches include:

• Categorization – Breaking large sets of options into smaller, more manageable groups (e.g., organizing restaurant menus into distinct sections like "seafood," "pasta," and "steak").

• Defaults & Recommendations – Providing pre-selected options or highlighting "best picks" to guide people toward simpler, more confident decisions. (Think of Amazon’s “Amazon’s Choice” label or pre-selected health insurance plans for employees.)

• Progressive Disclosure – Showing only a few options at first and letting users opt in to see more (e.g., “Show More” buttons in online shopping filters).

These strategies help prevent choice overload, but doing so also gives choice architects a lot of power over what we see, what we consider, and ultimately, what we choose. For example, retirement plan options may appear to be in employees’ best interests but structured in ways that benefit the plan provider. Similarly, when search engines prioritize certain products, they may reduce complexity for users, but they also have the power to subtly manipulate purchasing behavior. And so, ultimately, while limiting the number of options may indeed be in the decision maker’s best interests, ultimately, the choice architect often decides which options are available to or prioritized for the decision maker.

The Ethics of Choice Architecture: Who Really Benefits?

This brings us back to the broader ethical concerns of choice architecture—it’s not just about helping people navigate decisions but also about who controls the process and what interests they serve.

Choice architecture is often framed as a way to make decision making easier or guide people toward better decisions. But as we’ve seen, the same mechanisms that can be used to help people can also be used to manipulate them. Defaults can encourage saving for retirement, but they can also trap consumers in auto-renewing subscriptions. Framing can help clarify trade-offs, but it can also exploit cognitive biases to steer decisions in a predetermined direction. Reducing complexity can prevent choice overload, but it can also limit meaningful options in ways that benefit companies more than consumers.

At the heart of the issue is the question of agency—do these interventions genuinely support better decision making, or do they subtly remove autonomy under the guise of convenience?

One of the biggest ethical concerns is transparency. Many people don’t even realize how much their choices are shaped by unseen design decisions. If a default is set in a particular way, shouldn’t people at least know why? If a company frames a decision to favor a certain outcome, shouldn’t that be disclosed?

Without transparency, choice architecture becomes a hidden force—one that people are subject to without their knowledge. This is where trust becomes relevant. If people believe that choice structures align with their interests, they’re more likely to accept them. But if they suspect they’re being manipulated for corporate or political gain, trust erodes.

It’s also important to recognize that while choice architecture exists and can subtly influence decision making, its effects are not uniform. Some people may highly susceptible to choice architecture manipulations, and choice architecture may influence decision making more in some contexts than others. A default or framing strategy that works for one group or in one context may have little to no effect on another group or in another context. This variability makes it even more important to consider how these interventions are designed and who they ultimately serve.

Perhaps the biggest challenge in discussing the ethics of choice architecture is that there is no such thing as a neutral choice environment. Every decision must be structured somehow, and even inaction is a choice. The question is often whether there is intent behind the choice architecture—as opposed to mere randomness, which can exist in some cases.

And if there is intent, the responsibility falls on choice architects—whether policymakers, employers, or businesses—to ensure that decision environments are designed ethically, transparently, and in ways that genuinely serve the people who are ultimately making the decisions. Otherwise, choice architecture isn’t about helping people make better decisions—it’s about exploiting people’s decision-making tendencies so they make choices that benefit the choice architect.

1. There may have also been a third default on some sites, where no option was selected but you were required to either approve or deny the receipt of communication.

2. Which I’ll write about for next time.

3. Or potentially didn’t realize how much more the switch was going to cost them.

4. It can also determine which information is presented in the first place. And while, intuitively, we might argue that more information is always better, as Smets (2025) argued, that is not necessarily the case.