Econometric Modelling of Risk Adverse Behaviours of Entrepreneurs in the Provision of House Fittings in China Rita Yi Man Li, (Hong Kong Shue Yan University, Hong Kong) Abstract Entrepreneurs have always born the risk of running their business. They reap a profit in return for their risk taking and work. Housing developers are no different. In many countries, such as Australia, the United Kingdom and the United States, they interpret the tastes of the buyers and provide the dwellings they develop with basic fittings such as floor and wall coverings, bathroom fittings and kitchen cupboards. In mainland China, however, in most of the developments, units or houses are sold without floor or wall coverings, kitchen or bathroom fittings. What is the motive behind this choice? This paper analyses the factors affecting housing developers‘ decisions to provide fittings based on 1701 housing developments in Hangzhou, Chongqing and Hangzhou using a Probit model. The results show that developers build a higher proportion of bare units in mainland China when: 1) there is shortage of housing; 2) land costs are high so that the comparative costs of providing fittings become relatively low. Keywords: Imperfect information, Risk, Entrepreneur, Fittings, Probit model, Housing. Introduction In many places, such as Australia, the United Kingdom and the United States, housing developers provide floor and wall coverings, window frames, cupboards, and electrical fittings as standard equipment. In some cases, developers also provide heated floors and wine storage. There is, however, a completely different norm in mainland China. For instance, the majority of developers in Beijing and Nanjing sell bare dwellings to home buyers. There are no floor coverings, kitchen cupboards or basic bathroom fittings (Li, 2009). Table 1 shows the percentage of bare flats in Nanjing. Among all the residential developments available for sale in 2004 - 2007 in Nanjing, 90% were bare units. Table 2 displays a similar phenomenon. It shows the percentage of bare flats in Beijing from 1997 - 2008. Of all the first hand residential units sold in Beijing districts, more than 70% were sold as bare units. Similarly, bare units can also be found in Shanghai and Hangzhou. This paper aims at finding out the major drivers behind these housing entrepreneurs to build bare residential units in these cities. Hypotheses Purchasing bare units without any fittings in mainland China implies that buyers need to spend time in identifying relevant parties to decorate their housing units before they move in. Why do busy people, often trapped in crammed schedules accept such a time consuming building activity? There has been very limited study of this phenomenon. This paper aims to remedy this and test potential reasons behind this by testing theories of brand name, risk adverse behaviours and information costs etc. Specifically, the author proposes four hypotheses on the emergence of bare flats: 1. Developers tend to build housing units with fittings in high land price areas. 2. The larger the proportion of low income residents the higher the proportion of bare flats. 3. Developers with well renowned reputations (brand names) build more furnished flats. 4. The greater the shortage of residential units, the higher the proportion of bare flats. Australasian Journal of Construction Economics and Building Li, R Y M (2012) ‘Econometric modelling of risk adverse behaviours of entrepreneurs in the provision of house fittings in China’, Australasian Journal of Construction Economics and Building, 12 (1) 72-82 73 District Total residential developments available for occupation in 2004 - 2007 1 Percentage of bare flats Bai Xia 32 92 Da Han 4 50 Gao Chun 3 100 Gu Lou 65 94 Jian 71 93 Jiang Zhu 133 98 Li Shui 11 91 Lu He 12 92 Pu Kou 60 97 Qi Xia 50 100 Qian Huai 42 100 Xia Guan 24 92 Xuan Wu 35 92 Table 1 Percentage of bare residential units in Nanjing (Li, 2010) District Residential projects which build bare flats Total Percentage of bare flats Chao Yong 618 853 72 Chong Ping 160 182 88 Chong Wen 62 74 84 Da Xing 140 165 85 Dong Cheng 59 79 75 Fang Shan 83 90 92 Feng Toi 260 305 85 Hai Dian 379 470 81 Huai Rou 37 44 84 Mi Yun 43 52 83 Shi Jing Shan 48 54 89 Shun Yi 102 126 81 Tong Zhou 194 219 89 Xi Cheng 62 82 76 Xuan Wu 88 114 77 Yan Qing 17 24 71 Others 92 110 84 Table 2 Percentage of bare flats in Beijing from 1997 - 2008 (Li, 2009) 1 Each of the residential development refers to one residential project. Some of the projects consist of up to, or more than 1000 units. The districts refer to small areas inside the city. This applies also to Table 2. Australasian Journal of Construction Economics and Building Li, R Y M (2012) ‘Econometric modelling of risk adverse behaviours of entrepreneurs in the provision of house fittings in China’, Australasian Journal of Construction Economics and Building, 12 (1) 72-82 74 The first proposition: Developers tend to build housing units with fittings in areas with high land prices. While fittings in high land price areas only make up a small proportion in construction costs, fittings in low land price areas constitute a relatively large share. The costs of providing fittings which do not suit the taste of customers are relatively lower in areas with high land prices compared to the costs of providing wrong fittings in areas with low land prices as home purchasers might choose not to buy the flats when developers provide unsuitable fittings. As developers are risk adverse, it is natural that they provide fewer fittings in those areas to avoid the relatively high risks: P(f)/LH < P(f)/ LL where P (f) refers to the price of the fittings, LH refers to high land price area and LL refers to low land price area. The second proposition: The larger the proportion of low income residents, the higher the proportion of bare flats. Developers in affluent countries such as United Kingdom, Singapore, United States offer more fittings as compared to those in developing countries, such as China, Ghana, Indonesia etc. Buying favourite fittings implies that the buyers are investing the time and money in the act of searching information. A man who earns $4 per hour has a low discretionary income but may be happy to search by himself rather than pay the contractor to do all on behalf of him, while a man who earns $ 400 finds it not worthwhile to do so. Searching costs him $400 per hour which is more than the cost of letting the contractor do it. In view of this, high income areas theoretically should have a smaller portion of bare flats. Developers tend to build more well-furnished flats with fittings to suit the needs of customers. The third proposition: Developers with well developed reputations build more well- furnished flats and vice versa. Bare dwellings may avoid losses but do not help the developers to maximize gains from value added. Hence, well regarded developers may tend to build more housing units with good fittings to enhance their reputation for quality. For others, it may be better to provide no fittings as the provision of poor fittings may set back reputations which have taken years to build up. The fourth proposition: The supply is so limited that the huge demand lowers developers’ motivations to provide well-equipped units. Previous literature shows that shortages may lead to use of unqualified staff and poor materials. Compared to many overseas countries, China experiences a shortage of housing by private developers. With substantial excess demand, developers do not need any gimmicks to compete for potential buyers. Provisions of kitchen and bathroom fittings are unnecessary. Literature Review Assumptions, Functions and Factors affecting the Supply of Entrepreneurs Another important role played by entrepreneurs is risk-taking. Entrepreneurs are risk-takers (Blanchflower and Oswald, 1998) who accept the firms‘ risks (Carland et al., 1984) when running a business. Australasian Journal of Construction Economics and Building Li, R Y M (2012) ‘Econometric modelling of risk adverse behaviours of entrepreneurs in the provision of house fittings in China’, Australasian Journal of Construction Economics and Building, 12 (1) 72-82 75 Information, Imperfect Information and Asymmetric Information One major problem that the entrepreneurs face is imperfect information. Imperfect information affects market power and prices and leads to a dis-equilibrium where firms can charge monopoly prices (Granlund and Rudholm, 2011). Hence asymmetric information is a market failure which undermines the efficiency of product transactions. It is not unusual that consumers are uninformed about risks which affect their ability to choose terms which reflect their preferences correctly. Firms often exploit this ignorance by degrading contract quality intentionally and will then have little incentive to offer better deals as these will not increase sales (Bechern, 2008). Risk-averse Human Behaviour Because of imperfect information, individuals and entrepreneurs make their decisions under risk. Risk has been identified as the potential for threat, damage, injury, or other loss (Zou et al., 2007, Jin and Doloi, 2008). It may also be conceptualized as variance in outcome in any project (Das and Teng, 2001, Fellner and Maciejovsky, 2007). Risk can be managed by catastrophe planning, easing, insurance, control, identification, quantification and shifting it to other agents (Zou et al., 2007, Jin and Doloi, 2008). People‘s attitude to risk affect their behaviour and is important in decision making (Richard, 1975, Xiao and Yang, 2008). People‘s attitude towards risks is not consistent. Some people are more risk adverse than the others. Buyers cannot obtain information on housing quality from the previous owners when they buy new housing. Many of them only know they have bought a poor housing when they open the doors of their units (Gwin and Ong, 2000). Therefore, home buyers often rely on developers‘ reputation and previous residential project to make decisions so as to minimise their risk in home purchases. Loss-averse Human Behaviour Apart from risk aversion, many entrepreneurs are loss-averse which refers to a situation when decision-makers are more sensitive to losses than to gains. This phenomenon represents a discontinuity in their utility function (Berkelaar et al., 2004), graphically expressed as an abrupt change in the slope of the utility function at the reference point (Wang et al., 2009) which distinguishes gains from losses It also signifies that the utility function is steeper for losses than for gains, i.e. the disutility that one experiences in losing money is greater than the utility associated with gaining the same amount. Loss aversion has become an important tool to explain all sorts of phenomena which are not explained by traditional theory, such as the endowment effect (Tovar, 2009). Branded Product A brand name is a firm‘s most important asset when it comes to evaluating companies against each another (Laforet, 2011). Entrepreneurs spent a lot of effort on boosting the name of their companies. Branded products have valuable merits to consumers as well as sellers (Table 3) as they convey quality information to consumer. As early as 1960s, the American Marketing Association defined ―brand‖ as: ―a name, term, sign symbol or design, or a combination of these, intended to identify the goods or services of one seller or group of sellers and to differentiate them from those of competitors‖ (Zilg, 2011, p.284). With asymmetric information in China‘s housing industry, buyers use brand names to assess product quality. Reputation is an effective signal which provides quality information because it is firm-specific – gradually built up from the quality of the projects built by a developer in the past. In case where developers have a history of working in a community, home buyers can obtain information on quality of the units by observing previous projects. Australasian Journal of Construction Economics and Building Li, R Y M (2012) ‘Econometric modelling of risk adverse behaviours of entrepreneurs in the provision of house fittings in China’, Australasian Journal of Construction Economics and Building, 12 (1) 72-82 76 Merits of brand name Examples Allows firms to escape from the confines of generic prices Jensen and Drozdenko (2008); LeBel and Cooke (2008); Rotfeld (2004); Vukasovič (2009); Souiden and Pons (2009) Easier adoption of a new product Rotfeld (2004), Vukasovič (2009) Price premium Corkindale and Belder (1985); ; Olson (2008); Chen (2007) ; Li et al. (2009) disagree) Enhance perceptions of product or service quality Corkindale and Belder (1985); Horppu et al. (2008); Shannon and Mandhachitara (2008); Vukasovič (2009); Facilitate promotional effectiveness Corkindale and Belder (1985) The ―personality‖ of brand product add value to the consumer Charters (2011) Increase market share Vukasovič (2009) Decrease risk Vukasovič (2009); Matzler et al. (2008) Barrier to entry for competitors Omar et al. (2009) Provides information to the consumer Baltas and Saridakis (2009); Pechtl (2008); Pitta and Franzak (2008); Souiden and Pons (2009) Product differentiation Zilg (2011) To attract a group of loyal customers Not likely for housing Burnett and Bruce (2007); Horppu et al. (2008) Table 3 Merits of brand name Shortage and Sellers’ Goods Quality Market forces not only affect prices, they also determine suppliers‘ incentives to provide ‗extra‘ value which aims at being perceived as quality improvement. In case of shortage, suppliers do not need to compete with other sellers. As a result, they may minimize production costs by providing only limited accessories to their buyers (Hawthorne and Birrell, 2002, Ingersoll and Smith, 2003). Housing investment in the planned economy was limited before 1976. The government preferred to spend money on investments other than housing (Wang and Murie, 1996). Private housing investment was allowed after the death of Mao and rapid growth in population and subsequent alterations in national urban policy have brought about a rapid urbanization in China. Registered urban population has been increasing at a rate of 4% annually since 1980 (Wu, 1999). Residential unit supply never met demand. Housing shortages problems had became serious by the end of the Cultural Revolution in 1976 when there was only 3 m 2 of floor space per person on average. Despite the large quantity of housing that had been built since 1978, 4 million urban households still live with 4 m 2 per family member in 1994 and 400 000 had an average of 2.5 m m 2 living space per person (Wang and Murie, 1996). In view of the above, does it mean that the popularity of bare units in China is the results of a shortage of housing, risk adverse behavior by developers or something else? Research Method – Probit Model To find out the reasons behind the sale of bare units in China, the Probit model will be used. Covariates are measured from errors in regression model estimates. In binary dependent variables, the error term can also be dichotomous in nature, hence normal distribution assumption of error term in ordinary least square is not essential. Therefore, OLS regression which is suitable for continuous dependent variables may not be the best choice in case of dichotomous dependent variables to estimate an equation (Gujarati, 2006). Australasian Journal of Construction Economics and Building Li, R Y M (2012) ‘Econometric modelling of risk adverse behaviours of entrepreneurs in the provision of house fittings in China’, Australasian Journal of Construction Economics and Building, 12 (1) 72-82 77 In a binary response model, interest lies primary in the response probability Yt* = Xt P(y=1|x) = P(y1|x1,x2,…,xk) Consider an equation with binary response (0,1) : P(y=1|x)= Y(ð0 + ð 1 x1+ ð 2 x2+…+ ð k xk) = Y(ð 0+x ð) Where x denotes a complete set of explanatory variables, Y is a function which takes on values between 0 and 1, that is 0