The conventional wisdom is that one should live in retirement with an income of about 70-80% of pre-retirement income. It is less than 100% because the tax rate is lower, because Medicare takes care of some expenses, and because employment related expenses disappear. But these 70-80% are really a rule of thumb that should clearly be differentiated in various ways.
John Karl Scholtz and Ananth Seshadri do this using life cycle models. The easy parts: if you previously had a particularly high income, a lower share can do easily, in part because taxes drop much more. Same if you had many children as they are, hopefully, out of the house by retirement. The more complex part: one should draw a life-cycle model where the goal is to equalize the discounted expected marginal utility of consumption across time, where health, incomes and lifetimes all are uncertain.
Scholtz and Seshadri find replacement rates that differ widely, with a median of 68% (or 57% considering the five highest years of income) and range of 47% to 90%, depending on the household category. For example, married couples need a higher replacement rate because of the higher expected lifetime of the survivor compared to a single person. Also, changes in taxation rates are very important for determining the marginal utility of consumption, especially if tax rates will be increasing instead of decreasing like in recent years. Finally, shocks to earnings are very persistent, thus whether they happen at the start or the end of a career can change dramatically the replacement rate.
Might be useful to note the 68% figure reflects average pre-retirement income over all working years rather than final year's income. It's not an intuitive concept since few people would track that figure. I wasn't sure why the high-5 percentage was lower than the base figure until i follwed the link to the paper.
ReplyDelete