If you look at life satisfaction or happiness research, you'll see that these ratings are affected by big events for about a year, then they return to baseline. You see it in cancer patients -- big drop in life satisfaction, then return to baseline after 6-12 months. You see it in lottery winners -- big increase in life satisfaction, then return to baseline in about a year. You see it in getting married or getting divorced -- big changes in life satisfaction, then return to baseline after about a year or two.
What that tells us is that the external event doesn't have a whole lot to do with happiness or life satisfaction over the long haul, only over the short haul. Over the long haul, what matters is the individual and the way he/she is, thinks, processes, behaves, etc.
Assuming a year or so has gone by, most people will return to their general level of life satisfaction or happiness after any big life change, including retirement. My guess is that the people who're very satisfied with their lives in retirement were very satisfied with their lives prior to retirement, too, and people dissatisfied with their lives in retirement were dissatisfied with their lives prior to retirement.
I know that's a generalization, and there will be individual exceptions. But as a general rule, it holds. It's not so much about the event "retirement" as it is about the person, who he/she is (including a bunch of stuff outside our control), how they process things, how they behave, etc.
