[notahacker]

Of course a significant contributory factor behind the "hedonic treadmill" is that subjective indications of happiness as survey responses are essentially a proxy for the person's actual emotional state (which of course is largely unmeasurable mental activity). That is of course assuming that happiness actually is a tendency towards certain mental states. People's recollection of how they actually felt during a prior time period is imperfect, and subjective expectations for their maximum possible level of mental satisfaction are subject to revision. Its possible for survey participants to sincerely believe their overall satisfaction is still six out of ten on some crude scale whilst every single measurable or unmeasurable aspect of their mental well being has actually moved in the right direction (fewer anxiety symptoms, higher dopamine levels, less brain activity devoted to sources of irritation, higher confidence etc.)

Of course its also quite possible for participants to forget prior survey responses resulting in participants assigning same cardinal score to their happiness when surveyed in two different time periods, yet being able to unequivocally agree that their happiness (or indeed satisfaction with work) is a fair bit higher in period B than period A.