||It has been identified in Nie and Wong (2012) that substantial rental value losses can be induced by excessive land exploitation in urban villages in the mainland of China. It has also been suggested that incomplete and unclearly delineated property rights, transaction cost, and consequently the absence of effective building regulations have played an important role leading to such tragic results. Through a comparison of 96 pairs of cases collected from three major Chinese cities (Shenzhen, Guangzhou, and Xi'an), it has been found that compare to non-village estates, an average of 72% of urban villages have been overbuilt, which leads to a problem of 'more construction but less rental values'. The overdevelopment problem can be viewed as, according to institution theories, a production change (a distortion of resource combination), given the constraint of defective institutional environment. It is natural then a different timing of contract change may follow, since certain conditions for renewal projects can be met differently. Throughout the lifetime of a building community, natural rent dissipation happens due to outdated design, changing market condition, or simply building aging. As a result, a building community has an economic life (a type of economic selection), which may be shorter than its physical life. An empirical test conducted in this study on 16 demolished non-village estates in Shenzhen, 14 in Guangzhou, and 10 in Xi'an in the recent years shows that the average economic lifetimes for non-village residential estates in the three cities are respectively 19.8, 22.5, and 28.9 years. However, urban villages have shortened life expectancies. An empirical study of 1054 online samples in Shenzhen, 138 samples in Guangzhou, and 134 sample in Xi'an shows a paradox that the average building ages in urban villages in the above three cities are quite new, respectively only 12.1, 12.2, and 13.8 years, but many such villages are already demolished or planned for demolition soon. The data for non-village buildings were collected through a scan of the official renewal documents on the governments' websites. The data for urban villages were collected online from real estate websites, on which landlords post their 'for rent' information. Such a paradox can be explained in connection to the previous pattern identified in Nie and Wong (2012), that in comparison to non-village estates, urban villages suffer a problem of more severe rent dissipation induced by defective initial institutional settings. The problem can make the critical condition for renewal due earlier, thereby shortening the life expectancies for urban villages. Besides, the income transfer effect caused by the state-villager interaction and the possible cost difference for negotiation can also bring forward renewal schedules. The combination of the three reasons then leads to earlier renewals.