An empirical study of lucky residential transaction price in Hong Kong

Abstract

This study examines the clustering of the rightmost non-zero digit at the number “8” in housing transaction prices in a Chinese-dominated society from a new institutional economics perspective. The second-hand transaction records from the 35 largest high-rise housing estates (with over 200,000 observations) in Hong Kong show that the frequency of number 8 as the rightmost non-zero digit is substantially higher than other numbers. Since the number 8 is considered to be a lucky number in Chinese culture, such a high frequency of its occurrence in the rightmost non-zero digit (hereafter referred to as the "Lucky Transaction Price" or LTP), is consistent with previous studies that suggest economic behaviour is influenced by culture (Henrich, 2000, Henrich et al., 2005, Fernández, 2011, Oosterbeek et al., 2004, Beugelsdijk and Maseland, 2010) This study conjectures that the Chinese cultural preference for the number 8 is sometimes utilized as a “sweetener” to facilitate transactions in the housing market, where transaction cost is high (the "Sweetener Conjecture"). One source of transaction cost in the housing market is the problem of information asymmetry, where the seller knows the quality of the housing unit more than the buyer. Since the number 8 is valuable to some market participants, settling on an LTP can help offset the high transaction cost due to information asymmetry. This implies that LTP is more (or less) likely to be observed in housing units with more (or less) information asymmetry problems. The empirical evidence from the 35 high-rise housing estates is consistent with this implication. First, LTP is more likely to be observed in older housing units, which have more serious information asymmetry problems since these units are more likely to have defects that cannot be easily spotted by visual inspection. This problem cannot be resolved by appointing a professional building condition inspector to do a presale inspection of the housing units in multi-storey buildings since the inspected unit may also be affected by defects in adjacent units or common areas. Second, larger housing units are also more likely to be transacted at LTP since there are more places for the potential latent defects to occur for larger units. Third, LTP is less likely to be observed in larger housing estates since the common areas and facilities of these estates are usually better maintained and managed by professional property management companies. There is a scale effect in property management and maintenance as the cost of which can be shared by more unit owners. This can help reduce some of the information asymmetry problems. Fourth, proximity to an MTR exit, an important attribute in Hong Kong, can reduce the probability of observing LTP. Since this attribute can be easily observed by both buyers and sellers, which implies less information asymmetry problem and thus less reliance on using LTP as a sweetener to facilitate transactions. Furthermore, another source of transaction cost in the housing market is information cost, which is higher when the market is thin. The empirical result shows that LTP is more frequently observed when transaction volume (excluding those transacted at LTP) is low, which is consistent with the Sweetener Conjecture. In addition, LTP is positively correlated with units on lucky floor levels. This may suggest that buyers who favour lucky floor level units have a higher valuation of the luckiness associated with the number 8, making LTP a better facilitator of transactions for these buyers. However, more research is needed to confirm this explanation. One contribution of this study is the innovative application of new institutional economics analysis within the context of the economics of culture, grounded in behavioural economics. The author has "merged" these two streams of economic approaches that have very different origins and fundamental assumptions to explain human behaviour in the housing market. From the estimated hedonic price models, this study shows that the value created by LTP is not evenly distributed between the buyers and sellers. The results show that the prices of units transacted at LTP are slightly higher, suggesting that the sellers capture most of the value created by LTP. This is reasonable as the LTP is associated with the unit which is in possessed by the buyer after the transaction. As such, the luckiness associated with LTP should psychologically reside with the buyer after the transaction, while the seller only “feels” the luckiness during the transaction. Therefore, the buyer is more willing to pay for the "sweetener" than the seller. This finding has important practical implications in the negotiation process, i.e., the sellers can increase their reservation prices to the nearest LTP to test the buyers' willingness to pay a slightly higher price. Another important implication is that real estate agents can suggest LTP to buyers and sellers to increase the probability or a transaction.