Adjusting for circumstance, and what the adjustment costs

Every departure from a uniform rule buys something and spends something.

Suppose the household agrees that the person who cannot reach the top shelf gets the middle shelf instead, and everyone else shuffles up. Something real has been gained: the shelves are now usable by the people who have them. Something real has also been spent. There is now a rule that requires knowing about people, someone must decide when a similar claim in future is comparable, and the next person to ask will be assessed rather than simply told the rule. That is not a catastrophe. It is a cost, and it should be counted rather than assumed away.

Adjustments also carry a signalling cost that is easy to miss. The person who receives one may not want to be the reason a general rule was altered, and may prefer an unusable shelf to a standing reminder. This is why the best adjustments are usually the ones that change the arrangement for everybody rather than granting an exception to somebody: a step stool in the corner does the same work as the shelf reallocation without making anyone the subject of a decision. Design that removes the obstacle beats design that grants relief from it.

The honest summary is that adjustment and uniformity are both partial. A perfectly uniform rule will sometimes confirm a difference it did not create. A rule full of adjustments will be unpredictable, will reward those most comfortable making a case, and will place a lot of weight on whoever decides. Anyone who tells you their preferred approach has no downside has not looked hard, and the shelf is a small enough example to check that claim on before you try it on anything that matters more.