Three key drivers of savings are life-cycle, precautionary, and bequest motives. What is their relative quantitative importance? We revisit this question focusing on the role of preferences and institutions. We address the challenge of disentangling the effects of different saving motives on one's decisions by considering many aspects of people's behavior both before and after retirement. We illustrate why this approach is informative about the underlying preference parameters, and hence allows us to uncover the relative strength of different motives. Our decomposition exercises reveal that bequest motive is the key driver of savings starting from the middle-age and long before retirement. We also find that life-cycle motive and precautionary motive due to medical expense shocks play a minor role. The former result is due the crowding out effect of Social Security. The latter is due to the combined effect of health insurance and the means-tested transfers.