MorselFlow

The choice-architecture note

Food choices: willpower, or the default?

Trying to make the best possible choice every time is tiring. What happens when the option already in front of us changes?

In three randomized field experiments at conferences, registration forms changed lunch from a non-vegetarian default to a vegetarian default. Participants remained free to switch. Vegetarian selection changed sharply in all three experiments.

Change the default, and the selected meal changed too.

non-vegetarian defaultvegetarian default
Lunch choices in conference registration forms. Participants in both conditions remained free to change away from the default.Source: Journal of Public Health, “Nudging healthy and sustainable food choices...”

Not choosing is also a form of choice.

A default can feel recommended, normal or simply easier. Even when switching remains possible, the option already selected carries weight.

That does not mean people are thoughtless. On a day full of small decisions, avoiding the cost of comparing every option is understandable.

A jump from 2% to 87% is not magic for every cafeteria.

These experiments took place in conference registration, lunch was free, and the participants were not a representative population. Price, taste, religion, allergies and availability can all change the outcome.

Still, the study makes one alternative to “try harder” unusually visible: redesign what is easiest to choose.

Instead of trying harder every time, place the choice you want closer to the start.

MorselFlow’s default is to begin with a photo.

A meal log that asks for every nutrient first creates a heavy starting point. MorselFlow begins with a photo, suggests candidates, and leaves only the parts that need review or correction.

Do not test willpower; design a lighter default. The idea can apply to both food choice and food records.

This article explains research and statistics for a general audience. It is not medical diagnosis, treatment or individualized nutrition advice. Results depend on the participants and conditions and do not guarantee the same outcome for everyone.