Engineer the default and skip the persuasion.
Johnson and Goldstein's 2003 paper in Science is the strongest single demonstration in this codex, because the effect size is not marginal.
Countries with explicit-consent organ donation — where you must opt in — showed effective consent rates in the single digits to around a quarter. Germany: about twelve percent. Denmark: about four. Countries with presumed consent, where you must opt out, showed rates above ninety-nine percent. Austria: 99.98. Belgium: 98.
Austria and Germany are neighbors with comparable cultures and comparable views about donation. The difference is the form.
Three mechanisms, and they compound.
Effort. Any action costs something, and a default requires none, so satisficing does the rest.
Implied recommendation. A pre-selected option is read as what the institution advises, and in most contexts that inference is reasonable.
And ownership. Once a default is mentally accepted, changing it is processed as a loss, which is chapter 17's endowment effect operating on a checkbox.
Thaler and Sunstein generalized this as choice architecture, and their argument is that there is no neutral option: every form has a layout, every list has an order, every setting has a starting value, and whoever writes it influences the outcome whether they intend to or not. The choice is not whether to have an architecture but which one.
The operational instruction is accordingly not to persuade. It is to build the form. Pre-ticked boxes, opt-out subscriptions, auto-renewal, default privacy settings, the pre-selected shipping upgrade, the recommended plan positioned in the middle.
The boundary is consent, and it is where the field has moved legislatively. A default that reflects what most people want, and that is easy to change, is good design. A default that transfers money or data, that the user did not notice, and that is difficult to reverse is the roach motel of chapter 263, and it is now specifically prohibited in the EU's Digital Services Act and litigated by the FTC.
The test is asymmetry. If getting in takes one click and getting out takes six screens and a phone call, the design is not expressing a recommendation.
The counter is chapter 293's mirror and it is mechanical: read the pre-ticked box, every time, and treat any asymmetry between entry and exit as the tell it is.
No defense chapter in the outline. A counter written for this guide.