Mega-event effects on the housing market: Evidence from the Beijing 2008 Olympic Games

Cities xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Cities journal homepage: www.elsevier.com/locate/cities

Mega-event effects on the housing market: Evidence from the Beijing 2008 Olympic Games Mei Wang, Helen X.H. Bao⁎ Department of Land Economy, University of Cambridge, CB39EP, UK

A R T I C L E I N F O

A B S T R A C T

Keywords: Welfare economics Public goods Use and non-use value Housing preferences Planning and regeneration

Mega-event regeneration involves extensive government funds and public participation; thus, this study emphasises the importance of verifying if these financial and human investments can be justified by the net effects of mega-event regeneration. Accordingly, the contingent valuation method is used to establish a framework to quantify the welfare effects of event regeneration from the economic, social and environmental perspectives. We proposed a theoretical framework that enables the ranking of various event regeneration effects based on public welfare improvement. This holistic approach takes into account changes in economic, environmental, and social housing conditions due to mega-event simultaneously. This leads to more reliable estimation of mega-event effects on housing market. Our empirical findings indicate that, overall, accessible public transport, a sense of feeling good, air quality, relieved traffic congestion and green space are the top five welfare enhancers. Nevertheless, residents from different housing sectors or geographic regions value mega-event effects differently. Our results can assist the government to efficiently allocate limited public resources by looking after public needs. A better understanding of the heterogeneity of event regeneration effects on different housing sectors and geographic locations will also help governments to tailor public policies based on various social groups.

JEL classification: H41 H53 I31 Q51 R21 R58

1. Introduction Staging mega-events has emerged as a significant contribution to the public policy of cities that seek economic growth, urban development and city branding. Often, the attractions that surround events are linked to a reimaging process and strategies of urban regeneration (Bianchini & Schwengel, 1991; Roche, 1994; Loftman & Spirou, 1996). The hosting of mega-events is often justified in terms of direct or indirect long-term economic, social and environmental consequences (Mules & Faulkner, 1996). Under the guise of mega-event preparation, urban regeneration has been increasingly regarded as a panacea by city governments to solve urban problems, accelerate physical change and ensure social cohesion. Event regeneration policies often exert multi-dimentional and longterm effects on the host city. The effect on the housing sector is particularly notable given the heavy involvement of public institutions and the broad range of policy targets involved. As key components of physical regeneration, infrastructure development (Ritchie & Lyons, 1990; Mihalik & Simonette, 1998; Gratton et al., 2005; Wei & Yu, 2006; Atkinson, Mourato, & Szymanski, 2008; Walton, Longo, & Dawson, 2008) and event facility construction (Konstantaki & Wickens, 2010; Mihalik & Simonette, 1998; Ritchie & Lyons, 1990) can significantly

⁎

improve housing conditions, thereby enhancing resident welfare. Environmental quality, which is often measured by green public space and fresh air, is also important in influencing resident living conditions and well-being (Chalkley & Essex, 1999; Deccio & Baloglu, 2002). By contrast, the effect from event-associated social regeneration on resident housing conditions is often indirect and subtle. For example, improved national and community safety, which results from combined public security measures and civic order patrols, can improve housing conditions (Kim et al., 2006; Kim & Petrick, 2005; Ritchie, Shipway, & Cleeve, 2009). A people-oriented policy that promotes public participation in decision-making processes for community development effectively enhances a sense of belonging and civic pride in residents (Chalkley & Essex, 1999; Raco, 2004; Smith, 2012). Despite the positive effects of mega-event regeneration, its adverse consequences are eliciting concern. Infrastructure development and facility construction are the vital components of physical regeneration and often increase surrounding property prices and the living cost of the host city (Malfas, Theodoraki, & Houlihan, 2004). Environmental degradation, such as construction waste, air pollution and noise, may occur (Deccio & Baloglu, 2002). The core theme of social regeneration, that is, promoting public participation and involvement, has resulted in the multitudes that flock for the event and often lead to the crowded use

Corresponding author. E-mail address: [email protected] (H.X.H. Bao).

http://dx.doi.org/10.1016/j.cities.2017.07.014 Received 22 November 2016; Received in revised form 13 July 2017; Accepted 15 July 2017 0264-2751/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Wang, M., Cities (2017), http://dx.doi.org/10.1016/j.cities.2017.07.014

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

government needs to design public policies that can balance the social interest between the rich and the poor; redirecting government spending from specialised facilities to transport network and services may improve the overall social welfare; and Beijing residents are willing to support initiatives that can accomplish sustainable environmental improvements. Our study can serve as valuable and relevant references for similar studies in developing countries. The rest of this paper is organised as follows. The methodology section presents the theoretical underpinning and the general form of our theoretical model. The next section provides the institutional background of this study by defining and discussing the complete effects of the Olympic Games and the areas affected by the mega-event regeneration. The following section introduces the survey design and data collection process, followed by empirical evidences and discussions. The final section concludes with policy recommendations.

of public resources, traffic congestion and potential threat of increasing crime rate, as well as hostility from the locals (Atkinson et al., 2008; Davies, 2005; Konstantaki & Wickens, 2010; Mihalik & Simonette, 1998; Walton et al., 2008). Staging mega-events can stimulate extensive economic, social and environmental effects on the host city that can significantly influence local resident welfare either positively or negatively. However, the intersection between event effects and resident welfare change has not been quantified and analysed adequately. The majority of existing event studies are characterised as case-specific and qualitative research for example Zhang and Zhao (2009) because the effects of event regeneration tend to be ‘intangible’ and cannot be directly measured in the marketplace. Hence, quantifying their value is not an easy undertaking. The contingent valuation method (CVM) has gained increasing popularity on the valuation of non-market goods. CVM has been routinely applied in surveys to evaluate the well-being effect of non-market goods, such as improvements in transport infrastructure and environment quality (See, for example, Dolan and Metcalfe, 2008; Atkinson et al., 2008; Hui, 1999; Adamowicz et al., 1994). This approach has received increased attention among academics in evaluating the intangible effect of mega-events. Existing CVM research conclusions are mixed. For example, Johnson, Groothuis, Whitehead, and J. C. (2001) suggested that government subsidy for a sports project cannot generate sufficient valuable public goods to justify the stadium costs based on the aggregated willingness to pay (WTP) estimates. However, an increasing number of scholars have obtained empirical evidence to support government-sponsored investment in hosting mega-events (Atkinson et al., 2008; Walton et al., 2008; Süssmuth, Heyne, & Maennig, 2010; Wicker, Prinz, & Hanau, 2011; Humphreys et al., 2011). Estimates of WTP vary substantially across different studies even for the same mega-event (Atkinson et al., 2008Walton et al., 2008). The variation can largely be attributed to the different public goods investigated. Therefore, a complete list of various public goods generated by the event should be included to establish a holistic and reliable account of the event effects. Failure to conduct a holistic analysis will lead to a misunderstanding (i.e. either under- or over-estimation) of the effects and ultimately cause misallocation of scarce public resources. This may discourage public enthusiasm on and participation in mega events, and subsequently result in a reduction in general social welfare. Most of the CVM studies on event regeneration effects concentrate on an isolated aspect of the event effects (Humphreys et al., 2011; Wicker et al., 2011; and Heyne, Maennig, & Süssmuth, 2007), thereby failing to measure holistically the welfare value of mega-event regeneration. Scholars who have attempted to provide a holistic assessment for the event effects often failed to reliably quantify individual event effects considered in monetary terms (Atkinson et al., 2008; Walton et al., 2008). To bridge the gap in the literature, we set up a theoretical model that quantifies the welfare effects of mega-events from the economic, environmental and social aspects. The framework can be used by to better understand stakeholder preference towards hosting mega-events, and to allocate public resources efficiently and effectively. Another important contribution of our research is the addition of empirical evidence from China to the literature. Most studies are derived from developed countries (Atkinson et al., 2008; Walton et al., 2008; Süssmuth et al., 2010; Wicker et al., 2011; Humphreys et al., 2011). Thus, assuming that the conclusions from these studies can be generalised globally is potentially problematic due to the different economic development and cultural backgrounds among countries. Therefore, verifying whether the CVM broadly used in the West can also be applied to Mainland China, one of the rapidly growing emerging markets in the world, is of empirical importance for studies of megaevent effect in developing countries. Based on our empirical evidence, we provide policy recommendations regarding mega-event regeneration in China. For example, the divergent perceptions of the Olympic effects among private and public homeowners suggest that the

2. Methodology Our theoretical model stems from Hui (1999), where WTP is a function of resident's characteristics, property traits and changes in housing conditions. Specifically, the improvement of resident's wellbeing is determined by five groups of variables: the economic, environmental and social effects of event regeneration on the housing conditions; demographics, and property traits. Ordinary least squares (OLS) method is routinely used to disaggregate and quantify welfare changes from individual event effects on the housing market (i.e. a hedonic approach). We follow this practice by using OLS to establish the relationship between well-being improvement and the above-mentioned variables, as outlined below.

Well − being improvement = f (Δheco , Δhenvir , Δhsoc , S , P ),

(1)

where Δheco, Δhenvir, and Δhsoc are the changes in economic, environmental, and social housing conditions due to mega-event respectively, S is a matrix of household demographic characteristics, and P is a matrix of property attributes. Δheco refers to the change in economic housing conditions that influences resident welfare in the host city. It generally includes measurements of event facility and stadia, travel cost, transport infrastructure, housing price appreciation and income growth. Variables that are often used to evaluate the welfare effects of changes in environmental housing conditions Δhenvir include green space ratio, cultural landmarks, parks, and air quality. The welfare effect of change in social housing conditions, Δhsoc, can be determined by measuring the feel-good factor, neighbourhood security, traffic congestion and overcrowdedness. If several or all of the coefficient estimates of variables in each category are statistically significant, the proposition that residents are willing to pay for the change in economic, social and environmental housing condition is empirically confirmed. The hosting of mega-events may constitute positive change for several households and negative change for others. For example, eventled infrastructure development and environmental upgrades can substantially improve housing conditions and benefit residents in the surrounding area. By contrast, compulsory land acquisition and housing demolition for event-related developments and social regeneration can cause involuntary housing relocation, thereby making low-income residents' lives considerably difficult (Malfas et al., 2004). Olds (1998) also finds that accelerated urban restructuring is necessary to support major events in most cases; in such cases, the socially disadvantaged suffer disproportionately. The construction of sports facilities and landmark architecture can be used as new leisure facilities after the event and improve the urban landscape. These world-standard facilities may not be accessible by the deprived households, which typically are public housing occupiers, because of the geographic distance and affordability constraints. Hosting mega-events stimulates discrepant effects on resident welfare and cause further disparity and imbalance between different social groups (Wang, Bao, & Lin, 2015). Furthermore, 2

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

the physical environment and the functionality of different regions based on their existing infrastructure, economic condition and social structure. The goal of CRB is to reinvent Beijing's image as a cultural metropolis with a dynamic economy and liveable space. The City Core is a densely populated area with well-developed infrastructure. It is also renowned for its economic prosperity and political significance. Thus, the objective of the CRB programme in this area was to improve existing infrastructure while preserving the traditional neighbourhood and social structure. This programme is in contrast to the regeneration target and process implemented in the host regions, that is, the Haidian and Chaoyang districts. The Olympic Core District was allocated on the border between the Haidian and Chaoyang districts. These districts are relatively less populated and less economically active than the two City Core districts (Wang et al., 2015). Consequently, the CRB programme for these districts was mainly manifested in the construction of new event venues, transportation networks and landscaping. An investment of CNY 36 billion (approximately USD 5.9 billion) was made for the upgrade of public transportation infrastructures, including the construction of the Sixth Ring Road, ten expressways, seven subway lines and the expansion of the Beijing Capital International Airport. The CRB programme also emphasised the social and cultural development in this area by constructing dozens of new museums, libraries and public spaces. A total of 25 historical sites were also identified as preservation zones. In this study, Olympic-affected areas are defined as the districts that hosted event venues or received significant attention in the CRB programme. These areas include six urban districts (i.e. Dongcheng, Xicheng, Chaoyang, Haidian, Fengtai and Shijingshan) and one suburban district (Shunyi). Fig. 1 shows that most of the Olympic venues are located in Chaoyang, Haidian and Shijingshan. Shunyi mainly hosted water sports venues, but the expansion of the international airport and the construction of the new subway line to the airport substantially improved accessibility to this district. Similarly, Fengtai also benefited from convenient accessibility from the subway expansion. Although the two City Core districts (i.e. Xicheng and Dongcheng) did not host many event venues, the CRB programme allocated a substantial amount of funding to improve the natural, social and cultural environment in these districts. Subsequently, these districts were selected as the study areas. In addition to the seven districts, Changping, Tongzhou, Daxing and Fangshan were included as remote areas to capture the heterogeneity of the event regeneration effects between the different housing sectors and geographic regions.1 Thereafter, we conducted survey interviews in these districts to investigate if and how Olympic regeneration affected resident welfare by changing housing conditions, as well as how the event regeneration effects varied in the different housing sectors and regions.

the variance of resident demographic characteristics and property conditions between private and public housing sectors also affect how they perceive the mega-event effects and ultimately determine their willingness to pay for the event (Wang et al., 2015). These means the coefficient estimates of all four groups of variables in Eq. (1) may be different between residents from the public and private sector. Thus, a comparative study between the private and public sectors is necessary to understand the distinctive effect of housing quality change on different social groups. Specifically, Eq. (1) will be estimated by using data from the public and private sectors separately. Equally important, mega-event regeneration effects can vary greatly depending on a site's geographic distance to the main event site. For example, Atkinson et al. (2008) compared the intangible effects of the 2012 London Olympic Games on resident welfare in London, Manchester and Glasgow. They concluded that London experienced the highest welfare enhancement. Hence, geographic location plays a significant role in studying the event effects and resident welfare change. This calls for separate models for areas that are in close proximity to the event sites and other regions. Consequently, Eq. (1) will be estimated with subsamples from areas that are close to event sites and that are further away from event sites separately. 3. Institutional background As a significant recent global event, the 2008 Beijing Olympic Games is used to empirically verify the theoretical models proposed in the previous section. China is one of the few developing countries that were awarded the right to stage a large-scale mega-event, such as the Olympic Games. The successful bid for this highly competitive megaevent showcased China's increasing economic power and political significance on the world stage. Nevertheless, China's infrastructure provision, environmental protection and urban planning in 2001 remained in an infancy stage compared with those of many developed countries. Unsurprisingly, the national and local authorities in China and Beijing tailored a series of measures for Olympic regeneration that drew extensively from public funds to construct new facilities, upgrade infrastructure and improve environment quality to meet the requirements of the International Olympic Committee (IOC). This approach is different from that in the West, where countries tended to rely on existing infrastructure and, where necessary, built facilities that required a substantially lower public budget than China. Since Beijing successfully bid for the Summer Olympic Games in 2001, over CNY 300 billion (approximately USD 48.9 billion) was allocated to event preparation between 2002 and 2008. The Olympic Core District (the dark grey area, block 4 in Fig. 1) was designated as a main function area and a recreational centre, where new sporting venues and a National Park were to be connected by 62 roads and 4 flyovers. The construction of sports facilities and infrastructure upgrades were not limited to the Olympic Core District but spread throughout Haidian and Chaoyang districts, as indicated by the red dots along the subway lines in Fig. 1. Although most of the Olympic facilities are located in only two of the 16 districts in Beijing (i.e. Haidian and Chaoyang districts), the effect of the Games reached beyond these areas. In addition to the direct investment in event venues, the government spent an additional USD 40 billion on infrastructure, USD 26 billion of which went to transportation and the rest was used to improve utility network, water and sewage systems and urban environment.1 The unprecedented investment in infrastructure, particularly in the transportation network, improved the accessibility to the Olympic Core District, the city centre and the Beijing Capital International Airport. For example, four new subway lines were developed throughout the City Core area (i.e. Xicheng and Dongcheng districts); a new line was built to connect the international airport to the rest of the city (See Fig. 1). The master plan of this ambitious project is summarised in the City Regeneration and Beautification (CRB) programme. The CRB programme was designed to improve

4. Empirical implementation We collected and constructed our data set via survey interviews because of the shortage of official statistics on resident perception towards the Olympic effects and the resultant utility change in China. The survey was conducted in 2009, one year after the Games, when the event effects had taken effect and the residents' enthusiasm remained strong. In-person interviews directly investigate the locals' welfare change by asking their WTP for the Olympic regeneration and concomitant housing improvement. Thus, the risk of selection bias can be reduced. To strike a balance between sample representativeness and survey cost, we adopted a stratified sampling procedure as follows. We used the 12 districts as strata and performed systematic sampling to 1 To facilitate a comparison between the private and public housing sectors, we include the New Development regions (i.e., Changping, Shunyi, Tongzhou, Daxing, and Fangshan districts as shown in Figs. 2) where a vast majority of public housing projects are concentrated.

3

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

Fig. 1. Map of Olympic venues and underground (2002 vs. 2008). Fig. 2. Survey Location and Functional Regions of Beijing.

amenities. In this sense, each development can be viewed as a small but well-established community. The number of residents in each residential development varies from hundreds to thousands. Moreover, the social and economic background in these housing developments is often representative of its neighbourhood. A total of 2500 surveys were conducted in collaboration with the

randomly select the residential developments from each strata/district. This process resulted in 16 residential developments as our survey locations (marked by blue magnifiers in Fig. 2). Thereafter, we randomly selected the households in each survey location to conduct the interviews. In China, residential properties are typically developed as clusters of high-rise buildings, surrounded by an extensive range of 4

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

find that hypothetical willingness to pay is no different from actual willingness to pay when adjusted by respondent certainty. We used the experiment of Johannesson, Liljas, and Johansson (1998) as basis to minimise the hypothetical bias problem, in which the respondents were asked to simulate a real payment situation and select the lowest (highest) boundary that will ‘definitely’ pay (not pay). Following the practices of Blumenschein et al. (1998, 2001) and Atkinson et al. (2008), the lowest boundary of WTP that the respondent indicated as ‘definitely’ will pay is regarded as a ‘conservative option’; thus, this answer is used as the dependent variable in Eq. (1) to analyse the relationship between the Olympic effects and resident welfare changes that are manifested in their WTP for the event regeneration. The last part of the questionnaire collects information for individual respondents who are the household representatives. The questions cover the demographic characteristics of income range, education level and employment status. The empirical model is built based on the assumption that the respondent WTP is a function of the Olympic effects on the economic, social and environmental housing conditions and other factors. The other factors are mainly composed of household characteristics and property attributes. The empirical model is specified as follows. All variables in the equation are defined in Table 1.

State-owned Assets Supervision and Administration Commission of the State Council (SASAC) from October 2009 to April 2010 in Beijing. The surveys were conducted in 10 districts; the core urban area of Beijing is demarcated as the Beijing Urban Planning Committee. Overall, 1909 valid questionnaires were gathered. The response rate is 76.3%. The questionnaire is divided into four parts. First, the respondents were asked to reveal their perception of the Olympic effects, which were divided into economic, social and environmental categories. A total of 13 questions were designed in response to the three categories of the Olympic effect analysed. Initially, the respondents were asked if the Olympic Games had any effect on the housing conditions in Beijing. This question comprises two parts: if the respondent confirmed the Olympic effects, then he or she is led to the next question; otherwise, the respondent is considered to have experienced no welfare change and the interview is completed. The following question analysed the extent to which the respondents perceived the event effect on a specific housing characteristic (e.g. transport infrastructure, sense of belongingness to the neighbourhood). This effect was measured on a scale from 0 to 10, with 10 indicating the highest level of improvement. In addition, the Olympic effect on the following three housing attributes were assessed monetarily (in CNY): daily commuting costs, housing price appreciation and income growth. Based on the respondents' evaluation of the event effects, the respondents were asked if they were willing to pay for the 2008 Summer Olympic Games in Beijing. This question comprises two parts: if the respondent expressed a positive WTP for the event, then he or she was led to the next question; otherwise, he or she was regarded as having experienced no welfare change, thereby completing the interview. Subsequently, the respondents were asked to determine the total amount of their WTP for the mega-event. Specifically, they were told that the amount of WTP that they stated would be collected from them by the Beijing Municipal Commission of Housing and Urban-Rural Development (BJJS) as an additional estate management fee.2 The coercive payment is designed to remind the respondents about their budget constraint, thereby encouraging them to take the questionnaire seriously and mitigate hypothetical bias. The hypothetical bias problem is often associated with the use of CVM, which is defined as the tendency for hypothetical willingness to pay or overestimate real willingness to pay (Blumenschein, Johannesson, Blomquist, Liljas, & O'Conor, 1998; Cummings et al., 1995 and 1997). Hypothetical bias occurs when CVM respondents state that they will pay for a good when in fact they will not, or they will actually pay less when placed in a similar purchase decision. Hypothetical bias is usually attributed to the presence of passive use values and a lack of familiarity with paying for policies that provide passive use value. However, CVM deemphasises the significance of this kind of bias, as in the case of Mitchell and Carson (1989) because of ‘the lack of document influence’. Carson (2011) also argues that the bias can be reduced and minimised by careful survey design. Later on, Cummings and Taylor (1999) find empirical evidence to support Carson's (2011) argument that the divergence between hypothetical and real willingness to pay is eliminated by the cheap talk script. In a “cheap talk” script, the interviewer defines hypothetical bias to respondents, explain why it may occur, and ask respondents to behave as if they are in a real payment situation. Furthermore, Blumenschein et al. (1998) and Blumenschein, Johannesson, Yokoyama, and Freeman (2001) offered another solution for the bias by asking the interviewees to indicate their answer in two categories: one is ‘definitely sure’ about payment, and the other category is ‘probably sure’. They consider only those respondents who indicate they are ‘definitely sure’ about payment and

WTP = β0 + β1 Olympic + β2 City + β3 Edu + β4 Income + β5 Hukou + β6 Age + β7 Time + β8 FeelGood + β9 Security + β10 Crowd + β11 Congestion + β12 Green + β13 Air + β14 Park + β15 Amenity + β16 Transport + β17 Travel + β18 Facility + β19 Appreciation + β20 Growth Table 2 shows that the descriptive statistics of the respondent WTP provides preliminary evidence confirming that the majority of the respondents are willing to pay for the mega-event regeneration. Regional divergence is more apparent than the disparities between the private and public housing sectors. That is, the number of respondents who were willing to pay for the event regeneration in the event affected area is twice the number of respondents in the outskirts. Accordingly, the difference in percentages of respondents in the public and private sectors who were willing to pay for the event is insignificant. Meanwhile, the disparity of the respondent WTP between the two geographic regions is larger than that between the different housing sectors. 5. Empirical findings and discussions Prior to exploring the empirical outcomes, we must be aware of two problems that are embedded in OLS. Firstly, the potential imprecision is rooted in the measurement of the explanatory variables. For example, people may feel uncomfortable when asked about their earnings. To address this concern, we only ask the interval of their income in the survey and use the mean value to represent their economic condition. Secondly, OLS is generally biased and inconsistent because of the endogenous problem when a correlation occurs between the explanatory variable and error term. The presence of the endogeneity problem can considerably induce bias in the coefficient of the independent variables, thereby leading to unreliable result estimation. To eliminate (or at least mitigate) the problem of endogeneity of one or more explanatory variables, Instrumental Variable (IV) method is routinely used in the literature. In this study, variable Income (i.e., household income) is likely to suffer from endogenous problem because some missing variables such as education level and work experience of the respondents are correlated with both Income and the WTP. To correct the endogeneity problem, household entertainment is selected as the IV for household income based on previous studies (Clotfelter & Cook, 1990; Scott and Garen 1994; Farrell,

2 No council tax system is used in China; the payment vehicle is used as an additional residential estate management fee. Given that BJJS is supervised by SASAC, which is entitled to adjust and manage the rate of estate management fee, such an institutional setting is believed to be realistic and credible to the respondents.

5

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

Table 1 Variable Definition. Category

Definition

Variable name

Willingness to pay

Step 1. The attitude towards paying for the Olympic Games. Step 2. The monetary value of utility gains from the Olympic Games =1 if the respondent resides in the Olympic areas, namely, Haidian and Chaoyang =1 if the respondent is accommodated in the Inner City, namely, the western and eastern districts Year of housing purchased by the respondent =1 if the respondent received primary education. Equivalent to elementary level =2 if the respondent received education equivalent to GCSE level or middle school =3 if the respondent received education equivalent to A-level or high school =4 if the respondent received college/university education =5 if the respondent received education equivalent to Master's level =6 if the respondent received education equivalent to PhD level or above Total monthly income earned by the family members (k) Age range of the respondent =1 if the respondent holds Beijing Hukou; otherwise, 0 =1 if the staging of the Olympic Games increased the resident's feeling good about living in their community, 0 otherwise Rated between 0 and 10; 0 indicates no effect; 1 indicates a significant deterioration of public safety; 10 indicates a substantial decrease in crime rate =1 if over-crowded, 0 otherwise =1 if traffic congestion worsened because of the Olympic Games =1 if private green space in the neighbourhood increased because of the mega-event =1 if the air quality in Beijing improved because of the Olympics, 0 otherwise. Rated between 0 and 10. 0 indicates no effect; 1 represents a significant downgrade of recreational facility, 10 means a significant improvement of event-related amenity and cultural landmarks Rated between 0 and 10. 0 indicates no effect; 1 represents a substantial reduction of public green space (i.e. neighbourhood parks), 10 indicates a considerable improvement of public green space because of the Games =1 if the accessibility of public transport facility and services improved because of the Olympics, 0 otherwise Change in monetary cost on daily commuting indicated by the respondent in CNY/month Change in housing price per square metre under the Olympic effect indicated by the respondent in thousand CNY Rated between 0 and 10; 0 indicates no effect; 1 represent significant downgrade of event venues and facilities; 10 indicates significant improvement of the event facility under the Olympic effects =1 if the respondent agreed that the household income is increased because of the Olympic Games

WTP

Property Attributes (P)

Olympic area Inner city

Demographic characteristics (D)

Purchase time Edu

Family Income Age Hukoua Feel-good

Social effect (S)

Environmental impact (EN)

Neighbourhood security Over-crowded Traffic congestion Green space ratio Air quality Cultural landmark (amenity) Park

Economic effect (EC)

Public transport Travel cost Housing price appreciation Olympic facility Income growth

Olympic area Inner city Time Edu

Income Age Hukou Feel good Security Crowd Congestion Green Air Amenity

Park

Transport Travel Appreciation Facility Growth

a Hukou is the household registration system in China. Individuals are broadly categorised as a “rural” or “urban” based on their Hukou registration, and Hukou system is used to controls the movement of people between urban and rural areas. The number of individuals that are allowed to move from urban to rural areas (i.e., migrants) or into certain urban areas (e.g., Beijing City) is often tightly controlled. People who live outside their registered Hukou area are not qualify for housing benefit, free health care and education in their resident area.

a correction of the endogeneity problem. Thus, the selected instrument variable (i.e., means of household entertainment) is proven to be a valid measurement that can produce more accurate and reliable results than OLS (Wooldridge, 2012). To further verify the validity of IV, we follow Kung and Bai (2010)'s example of repeating the same steps to perform OLS but add IV to the estimation as the exogenous regressor to test whether it has any direct effect on the dependent variable. Only when the exogenous regressor has no direct effect on WTP can they be regarded as plausible instruments. This test can be regarded as an easy-to-interpret version of the over-identification test (Kung & Bai, 2010). The results are presented in column (3) in Table 3. The IV is insignificant as an exogenous regressor, thereby suggesting and the IV (i.e., means of household entertainment) is valid. According to Table 3, most of the 13 housing attributes that were used to determine the WTP for Olympic Games turned out to be significant, with the exception of income growth and amenity (which are subsequently removed from the full model). We found strong evidence to support the proposition that the respondents are willing to pay for the economic, social and environmental effects of the Olympic Games on their housing. Although previous studies tend to focus on one specific aspect (e.g. social) of the event effects, our theoretical model is designed to provide a holistic account of the mega-event effects via economic, social and environmental perspectives. Thus, the welfare effects of a mega-event can be measured more accurately compared with previous research. In

Table 2 Descriptive Statistics for WTP.

Full sample Private sector Public sector Affected regions Remote regions

Median

Mean

Min (count %)

Max (count %)

Std. dev

200 225 200 500 50

306.3 342.0 270.0 527.2 120.9

0 0 0 0 0

2000 (0.5%) 2000 (0.5%) 1200 (1.67%) 2000 (0.5%) 800 (3.75%)

329.4 385.2 255.7 344.8 156.6

(26.4%) (29.6%) (24.4%) (5.9%) (44.8%)

Morgenroth, & Walker, 1999; Blalock, Just, & Simon, 2007). As household income increases, residents should be able to afford additional means of entertainment; but its correlation with age will be much smaller. This satisfies the requirement to be a good IV. Columns (1) and (2) in Table 3 show the comparison of OLS and Two-Stage least squares (2SLS)3 models, in which the coefficient of household income (instrumented variable) has increased from 53.882 in OLS model to 63.070 in 2SLS model. The increase in the coefficient estimate of the instrumented variable (i.e., household income) suggests

3 Two-Stage least squares (2SLS) regression analysis is an extension of the OLS method. It is used to address endogenous issues in standard OLS analysis by using instrumental variables. In the first stage of regression, each endogenous regressor is regressed on its instrumental variable(s) and all exogenous regressors. In the second stage, the regression model is estimated with each endogenous regressor being replaced with its corresponding predicted values from the first stage.

6

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

Table 3 Determinants of WTP (instrumental variable approach vs. OLS). Variable

Inner city Olympic area Public ownership Time Income Entertainment Hukou Feel good Security Crowd Congestion Green Air Park Transport Travel Appreciation Facility

(1) Full model with IV

(2) Full model without IV

(3) IV validity check

Coef.

Std. Err.

P-value

Coef.

Std. Err.

P-value

Coef.

Std. Err.

P-value

71.120 78.115 − 38.321 − 16.722 63.070 – 125.806 110.849 11.997 − 46.846 − 137.32 81.159 145.018 9.307 160.856 − 0.346 13.948 7.529

13.777 12.997 18.590 2.776 7.073 – 16.361 10.559 2.132 10.184 17.055 10.495 15.183 2.091 12.761 0.043 3.813 1.549

< 0.01 < 0.01 0.039 < 0.01 < 0.01 – < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

78.176 86.008 − 42.727 − 14.304 53.882 – 118.523 114.314 12.511 − 47.479 − 128.73 83.442 145.659 9.713 168.788 − 0.350 15.166 7.541

13.062 12.012 18.453 2.306 3.880 – 15.743 10.366 2.115 10.221 16.203 10.437 15.245 2.083 11.743 0.043 3.748 1.556

< 0.01 < 0.01 0.021 < 0.01 < 0.01 – < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

77.604 84.656 − 43.169 − 14.118 49.978 7.005 118.749 113.989 12.441 − 47.606 − 126.941 83.204 146.033 9.741 168.268 − 0.352 15.213 7.504

13.061 12.038 18.446 2.308 4.629 4.536 15.735 10.363 2.115 10.217 16.236 10.433 15.239 2.082 11.742 0.043 3.746 1.555

< 0.01 < 0.01 0.019 < 0.01 < 0.01 0.123 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

Table 4 Comparisons with models that investigated event regeneration effects in isolation. Variables

Constant Feel good Crowd Congestion Security Green Park Air Transport Travel Appreciation Growth Facility Income Education Inner Olympic area Public Ownership Time Age Hukou R-squared No of obs.

Full model

Model 1 (Atkinson)

Model 2 (Walton)

Model 3 (Andersson)

Coef.

P-value

Coef.

P-value

Coef.

P-value

Coef.

P-value

− 268.595 110.849 − 46.846 − 137.318 11.997 81.159 9.307 145.018 160.856 − 0.346 13.948

< 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

− 252.760 185.741 − 79.282 − 239.921 13.615 126.294 12.456

< 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

− 128.231 256.303

< 0.01 < 0.01

−156.889 216.102 −68.742 −283.502

< 0.01 < 0.01 < 0.01 < 0.01

23.137

0.591

7.529 63.070

< 0.01 < 0.01

8.762 89.050 − 10.259

< 0.01 < 0.01 0.207

8.805 98.282 − 25.251

< 0.01 < 0.01 0.006

102.129

< 0.01

71.120 78.115 − 38.321 − 16.722

< 0.01 < 0.01 0.039 < 0.01 − 6.229

0.280

125.806 0.742 1394

< 0.01 0.536 1394

0.41 1394

0.468 1394

public housing sectors are helpful. By considering the variations in the Olympic regeneration policies applied to different geographic regions, separate models were built to determine the regional heterogeneity between the Olympics-affected area and the rest. Table 5 shows that the economic, environmental and social effects of the Olympic Games influenced resident utility in substantially diverse means between the private and public housing sectors.4 Firstly, the respondent welfare changed at different magnitudes in the two housing sectors. Specifically, private homeowners gained substantially

Table 4, we estimated our model by including factors considered in Atkinson et al. (2008), Walton et al. (2008) and Andersson, Rustad, and Solberg (2004). These models are labelled as Models 1, 2, and 3 respectively. The outputs of our model (labelled as Full Model in Table 4) are also included to facilitate comparison. The results show that the coefficients estimates are considerably biased, and R-squares are significantly low if event regeneration effects are studied in isolation. This finding offers strong support to the reliability and validity of our theoretical model. The full model in Table 4 provides a general indication that a significant association is determined between the resident WTP and the Olympic effects on the economic, environmental and social housing conditions. By contrast, although many local residents benefited from event regeneration and developments, a few economically and socially disadvantaged households often suffered disproportionally. To obtain a holistic and objective understanding of the mega-event effects, a comparative study of the resident welfare change between the private and

4 As pointed out by one of the referees, we do not have objective measures of the differences in housing quality between private and public sector groups as baseline or after the event. Although we include some objective measures of housing conditions in our model (i.e., Olympic Area, Inner City, and Time), this cannot guarantee that omitted variable bias is completely removed. The large discrepancy in welfare effect between public and private sectors may include the difference in housing conditions, and are likely to be over-estimated.

7

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

Table 5 Regional Variations of Mega-event Effects. Variable

(1) Public

Inner Olympic area Public Ownership Time Income Age Hukou Feel good Security Crowd Congestion Green Air Park Transport Travel Appreciation Facility R2 No. of obs.

(2) Private

(3) Affected

(4) Remote

Coef.

P value

Coef.

P value

Coef.

P value

Coef.

P value

57.658 84.407 – − 18.184 49.850 13.750 – 141.314 4.561 – − 78.698 80.740 – 23.242 167.557 − 0.207 − 19.293 3.948 0.755 715

< 0.01 < 0.01 – < 0.01 < 0.01 0.008 – < 0.01 0.027 – < 0.01 < 0.01 – 0.027 < 0.01 0.004 < 0.01 0.009

43.717 79.576 – −15.715 59.179 – 90.466 84.841 21.192 −70.861 −169.299 78.128 179.922 8.613 139.776 −0.267 58.283 11.991 0.775 679

0.050 < 0.01 – 0.035 < 0.01 – < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

– – – − 19.285 75.763 – 165.147 113.253 17.443 − 94.128 − 171.823 82.116 144.348 13.838 135.463 − 0.380 16.890 12.260 0.694 821

– – – < 0.01 < 0.01 – < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 0.002 < 0.01

– – − 76.776 − 12.501 37.383 – 85.527 105.137 6.256 31.190 − 65.708 113.745 113.013 – 177.238 − 0.131 – – 0.612 573

– – < 0.01 < 0.01 < 0.01 – < 0.01 < 0.01 0.004 0.004 0.001 < 0.01 < 0.01 – < 0.01 0.028 – –

WTP (i.e., Table 6 column 1). Table 6 columns (2) and (3) also confirm the variance between the two housing sectors. For example, AIR, APPRECIATION, TRANSPORT, CONGESTION and SECURITY are ranked as the most important factors in determining private homeowners' WTP. For the public sector, the rank is substantially different: TRANSPORT, FEEL GOOD, GREEN, APPRECIATION and CONGESTION. Table 6 columns (4) and (5) offer additional supporting evidence for the regional variation. The table compares the determinants of the resident WTP between different regions based on relative significance. In the Olympics-affected area, TRANSPORT, FEEL GOOD, AIR, TRAVEL and PARK were considered the most decisive factors, whereas TRANSPORT, GREEN, FEEL GOOD, AIR and CONGESTION were the top five variables that received the priority in the other areas.

more utilities than did their public counterparts from FACILITY and SECURITY, respectively. However, private homeowners suffered as much as double the welfare loss than did their public counterparts from CONGESTION. By contrast, public home-occupiers gained up to triple the welfare improvement from PARK compared to private flat dwellers. Secondly, the respondent welfare altered in opposite ways under the Olympics effects. That is, APPRECIATION served as a beneficial consequence for private flat buyers but unfavourable for the public residents. Residents in different geographic regions were expected to experience different welfare changes because infrastructure and social composition vary across regions, and that the regeneration goals and the associated financial investment made between the two regions are considerably contrasting. Table 5 columns (3) and (4) show that the Olympics regeneration has more to do with the event-affected area than the outskirt regions because the mega-event influenced the former to a deeper extent and a wider scale than the latter. This finding is consistent with our expectations. Despite notable variances in resident backgrounds and the inherent uniqueness of the two regions, the respondents' utility is determined to be influenced by the economic, environmental and social effects of the Olympic Games in similar aspects but to a different extent. In summary, Table 5 reveals notable variations in mega-event regeneration effects between different housing sectors and geographic regions. To obtain further evidence to support this notion, we rank the 13 housing variables based on their standardised coefficient estimates to facilitate the comparison for the relative importance of the Olympic effects on individual housing variable. The results of all five models (i.e., Full, Public, Private, Affected, and Remote) are given in Table 6. Overall, local residents perceived TRANSPORT, FEEL GOOD, AIR, CONGESTION and GREEN as the most important outcomes of the Olympic regeneration that significantly influenced the respondents'

6. Conclusions and recommendations In this paper, we proposed a theoretical framework that enables the ranking of various event regeneration effects based on public welfare improvement. This holistic approach takes into account simultaneously changes in economic, environmental, and social housing conditions due to mega-event regeneration. This leads to more reliable estimation of mega-event effects on housing market. We demonstrated that our model provided more reliable estimates of mega-event effects than those obtained based on Atkinson et al. (2008), Walton et al. (2008) and Andersson et al. (2004). This is particularly important for policymakers, who need to consider the overall welfare of all stakeholders when designing and implementing public policies. Our theoretical framework can assist the government to efficiently allocate limited public resources by looking after public needs. The econometric method adopted is in the form of hedonic price modelling, which not only identified the changing housing factors on welfare measurements but also quantified their significance and extent. The endogeneity problem is also considered and mitigated via instrument variables. Our empirical findings confirm that the welfare of the locals was influenced by state-sponsored event regeneration in different extent and aspects; however, a general welfare improvement is identified in the form of positive WTP. This result is consistent with the conclusion drawn from prior literature (Andersson et al., 2004; Heyne et al., 2007; Atkinson et al., 2008; Walton et al., 2008; Wicker et al., 2011; Humphreys et al., 2011). Other factors, such as the respondent demographic characteristics and their property conditions, also influence the respondent welfare change.

Table 6 Rank of the relative importance of housing attributes in WTP. Rank

(1) Full

(2) Public

(3) Private

(4) Affected

(5) Remote

1 2 3 4 5

Transport Feel good Air Congestion Green

Transport Feel good Green Appreciation Congestion

Air Appreciation Transport Congestion Security

Transport Feel good Air Travel Park

Transport Green Feel good Air Congestion

8

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

Our findings suggest that respondents with different property ownerships experienced various welfare changes. Private homeowners gained more utility from the Olympic Games than did their counterparts in the public sector, which was reflected in a high percentage of positive welfare change and a large WTP. The private homeowners enjoyed a significantly higher level of welfare gain than did their public counterparts from the Olympics facility, air quality and public security. Housing price appreciation had a value-added effect on the private home-buyers but significantly undermined the welfare of the public counterparts. These alarming findings have significant policy implications. In particular, governments should make special provision for caring for the economically vulnerable and impoverished who are likely to suffer on account of the escalation of housing prices. The divergent perceptions of the Olympic effects among private and public homeowners requires specialised solutions to maximise the welfare gain of affluent people and minimise the adverse effects on the deprived, as well as to balance the social interest between the rich and the poor. To maximise social welfare, public resources and funds should be allocated based on the stakeholders' preference. For example, we determined that Beijing residents are willing to pay for the staging of the event because of its positive effects on their housing conditions (i.e. an overall social welfare improvement). The top five priorities were accessibility to public transport, sense of feel-good, air quality, traffic congestion and green space in the neighbourhood. The locals revealed diverse preferences towards changing economic, environmental and social housing conditions, among which transport-related variables played considerable importance. The residents' strong preference towards accessible transport facility and increasing social concerns over traffic congestion require prioritisation of public transport developments in terms of network and services. The Chinese government invested USD 48.9 billion on the Olympic venues and related projects but only invested USD 26 billion on transport. The redirection of government spending from specialised facilities to transport network and services may improve the overall social welfare because these facilities tend to benefit a segment of the population (who live in proximity). By contrast, expenditure on transport facilities can improve accessibility to the city centre and workplace, as well as to moderate traffic congestion. The empirical results offer important policy implications for the government to efficiently allocate scarce public resources by meeting the demands of the public. The respondents evidently expressed an increasing interest in environmental protection and quality of life. If residents perceive environmental improvements offer merely short-term benefit, they tend to pay less or not at all. However, our results imply pressing demand from the public for sustainable environmental improvements. In preparation for the Olympics, a series of policies were implemented to improve environmental standards, notably the relocation of heavily pollutant state-owned enterprises to the neighbouring province of Hebei, traffic control on pollutant vehicles, and suspension of manufacturing production during the event (Cui & Pan, 2009). These measures are effective to enable the environmental standard to meet the IOC's requirement under a tight time-constraint but difficult to sustain because of extensive operational costs and potential risks for unemployment. Therefore, some alternative methods are demanded. For example, industrial restructuring may be a plausible option that facilitates the transformation of the city from heavily relying on industrial sectors to a service-oriented economy, thereby effectively reducing the use of resources, such as oil and coal and, consequently, air pollution. Of course, this process could be complicated and time-consuming. Improved public safety is an important welfare enhancer that can be attributed to a strengthened Public Patrol Force, as well as the ‘forced relocation’ of low-income native residents and migrant workers5 from

the regenerated area to the periphery regions. These measures may be temporarily effective but can be problematic in the long-term. Expelling migrant workers will distort the labour market (e.g. the construction and food-supply industries), thereby severely damaging social equality and creating regional discrimination and social havoc. To minimise such risks of social turbulence caused by inequality, the government is advised to remove restrictions on non-local residences during jobhunting and increase the supply of low-rent housing to ensure a decent quality of life for migrant workers. This study also revealed the heterogeneity of the event regeneration effects on the different housing sectors and geographic locations. For example, public housing occupants elicited stronger preferences towards improved accessibility to public transport than did their private counterparts. This situation sends a clear signal that public home dwellers are in substantial need for public transport services than private homeowners because the former are less likely to afford using private vehicles than the latter. A similar story is true in less-developed outskirt areas in which residents placed a high premium on accessible public transport, thereby indicating a deficiency of infrastructure provision in the urban edge and an urgent need for public transport development. Furthermore, the respondents in the outskirts disclosed an insignificant effect from the event facilities, housing price appreciation and parks. By contrast, people in the event-affected area benefited from substantial utility improvements from these changing housing conditions. Unevenly distributed Olympics effects deter outskirt residents from welfare improvements and discourage a sense of feel-good. Unbalanced Olympic regeneration and developments may stimulate social resentment between regions and housing sectors. To ensure balanced regional development and social cohesion, urban regeneration and development policies should be tailored based on public needs. Specifically, substantial public spending should be allocated on transit development, particularly in the remote and public housing concentrated areas, thereby alleviating intensified traffic congestion. Environmental upgrades, such as neighbourhood green space and communal parks, should be carried out in less affluent areas, where more public housing projects are developed. The variation between private and public housing sectors and geographic regions also call for regeneration policies that can direct valuable resources to where they are needed the most. Finally, a complementary use of ex-ante and ex-post methods would be rewarding for future research because they can provide a comparison for the ‘expected’ and ‘real’ effects of mega-event regeneration in relation to the respondent utility. The current results provide significant references for local authorities to assess the quality of their work in implementing event regeneration and yield important policy implications for future improvements in regeneration planning and enforcement measures. Furthermore, a combination of ex-ante and ex-post studies is another strategy to mitigate the hypothetical bias. The ex-ante approach can be used to address hypothetical bias at the survey design stage, whereas the ex-post approach can further address hypothetical bias with follow-up questions on the hypothetical willingness to pay questions. Given that we conducted a pilot study to improve survey design and mitigating hypothetical bias, the results of the formal survey are generally consistent with the findings of the pilot study, thereby developing our confidence in the representativeness and reliability of our formal survey findings. References Andersson, T. D., Rustad, A., & Solberg, H. A. (2004). Local residents’ monetary evaluation of sports events. Managing Leisure, 9(3), 145–158. Atkinson, G., Mourato, S., & Szymanski, S. (2008). Are we willing to pay enough to ‘back the bid’?: Valuing the intangible impacts of London's bid to host the 2012 summer

5 These migrant workers are characterized as those who have neither stable income, permanent accommodation and social network in Beijing; thus, they are considered a

(footnote continued) highly ‘threatening’ factor for social cohesion and security.

9

Cities xxx (xxxx) xxx–xxx

M. Wang, H.X.H. Bao

generated by a major league sports team-the CVM approach. Journal of Sports Economics, 2(1), 6–21. Kim, S. S., & Petrick, J. F. (2005). Residents' perceptions on impacts of the FIFA 2002 world cup: The case of Seoul as a host city. Tourism Management, 26(1), 25–38. Konstantaki, M., & Wickens, E. (2010). Residents' perceptions of environmental and security issues at the 2012 London Olympic Games. Journal of Sport and Tourism, 15(4), 337–357. Kung, J. K. S., & Bai, Y. (2011). Induced institutional change or transaction costs? The economic logic of land reallocations in Chinese agriculture. Journal of Development Studies, 47(10), 1510–1528. Loftman, P., & Spirou, C. S. (1996). Sports stadiums and urban regeneration: the British and United States experience. Paper presented in the conference: Tourism and Culture: Towards the 21st Century, Durham. Malfas, M., Theodoraki, E., & Houlihan, B. (2004). Impacts of the Olympic Games as mega events. ICE Proceeding: Municipal Engineer. Mihalik, B., & Simonette, L. (1998). Resident perceptions of the 1996 summer Olympic Games: Year II. Festival Management and Event Tourism, 5(1), 9–19. Mitchell, R. C., & Carson, R. T. (1989). Using surveys to value public goods: The contingent valuation method. Washington D.S.: Resources for the Future. Mules, T., & Faulkner, B. (1996). An economic perspective on major events. Tourism Economics, 12(2), 107–117. Olds, K. (1998). Urban mega-events, evictions and housing rights: The Canadian case. Current Issues in Tourism, 1(1), 2–46. Ritchie, B. W., Shipway, R., & Cleeve, B. (2009). Resident perceptions of mega-sporting events: A non-host city perspective of the 2012 London Olympic Games. Journal of Sport and Tourism, 14(2–3), 143–167. Ritchie, J. R. B., & Lyons, M. M. (1990). Olmpulse VI: A post-event assessment of resident reaction to the XV Olympic winter games. Journal of Travel Research, 28(3), 14–23. Roche, M. (1994). Mega-events and urban policy. Annals of Tourism Research, 21(1), 1–19. Smith, A. (2012). Events and urban regeneration: The strategic use of events to revitalise cities. London: Rouledge. Süssmuth, B., Heyne, M., & Maennig, W. (2010). Induced civic pride and integration. Oxford Bulletin of Economics and Statistics, 72(2), 9035–9049. Walton, H., Longo, A., & Dawson, P. (2008). A contingent valuation of the 2012 London Olympic Games a regional perspective. Journal of Sports Economics, 9(3), 304–317. Wang, M., Bao, H. X. H., & Lin, P. (2015). Behavioural insights into housing relocation decisions: The effects of Beijing Olympics. Habitat International, 47, 20–28. Wei, Y. D., & Yu, D. (2006). State policy and the globalization of Beijing: Emerging themes. Habitat International, 30(3), 377–395. Wicker, P., Prinz, J., & Hanau, T. (2011). Estimating the value of national sporting success. Sport Management Review, 15(2), 200–210. Wooldridge, J. M. (2012). Introductory econometrics: A modern approach (7th edition). South-Western College Publication. Zhang, L., & Zhao, S. X. (2009). City branding and the Olympic effect: A case study of Beijing. Cities, 26, 245–254.

Olympic Games. Urban Studies, 45(2), 419–444. Bianchini, F., & Schwengel, H. (1991). Reimagining the city. In J. Comer, & S. Harvey (Eds.), Enterprise and Heritage: Crosscurrents of National Culture (pp. 214–234). London: Routledge. Blalock, G., Just, D. R., & Simon, D. H. (2007). Hitting the jackpot or hitting the skids: Entertainment, poverty, and the demand for state lotteries. American Journal of Economics and Sociology, 66(3), 545–570. Blumenschein, K., Johannesson, M., Blomquist, G. C., Liljas, B., & O'Conor, R. M. (1998). Experimental results on expressed certainty and hypothetical bias in contingent valuation. Southern Economic Journal, 65(1), 169–177. Blumenschein, K., Johannesson, M., Yokoyama, K. K., & Freeman, P. R. (2001). Hypothetical versus real willingness to pay in the health care sector: Results from a field experiment. Journal of Health Economics, 20(3), 441–457. Carson, R. T. (2011). Contingent valuation: A comprehensive bibliography and history. Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Chalkley, B., & Essex, S. (1999). Urban development trough hosting international events: A history of the Olympic Games. Planning Perspectives, 14(4), 369–394. Clotfelter, C. T., & Cook, P. J. (1990). On the economics of state lotteries. The Journal of Economic Perspectives, 4(4), 105–119. Cui, S., & Pan, P. (2009). Beijing's economic society under the Olympic impacts. Beijing Bureau of Statistics. Cummings, R. G., Elliott, S., Harrison, G. W., & Murphy, J. (1997). Are hypothetical referenda incentive compatible? Journal of Political Economy, 105(4), 609–621. Cummings, R. G., Harrison, G. W., & Rutström, E. E. (1995). Homegrown values and hypothetical surveys: Is the dichotomous choice approach incentive-compatible? American Economic Review, 85(1), 260–266. Cummings, R. G., & Taylor, L. O. (1999). Unbiased value estimates for environmental goods: A cheap talk Design for the Contingent Valuation Method. American Economic Review, 89(3), 649–665. Davies, L. E. (2005). Not in my back yard! Sports stadia location and the property market. Area, 37(3), 268–276. Deccio, C., & Baloglu, S. (2002). Nonhost community resident reactions to the winter Olympics: The spillover impacts. Journal of Travel Research, 41(1), 46–56. Farrell, L., Morgenroth, E., & Walker, I. (1999). A time series analysis of U.K. lottery sales: Long and short run price elasticities. Oxford Bulletin of Economics and Statistics, 61(4), 513–526. Heyne, M. Maennig, W. and Süssmuth, B (2007) ‘Mega-sporting events as experience goods.’ International Association of Sports Economists. (Working Paper Series), Paper No. 07–06. Hui, E. C. M. (1999). Willingness to pay for better housing in Hong Kong: Theory and evidence (of dwelling space). Urban Studies, 36(2), 289–304. Johannesson, M., Liljas, B., & Johansson, P.-O. (1998). An experimental comparison of dichotomous choice contingent valuation questions and real purchase decisions. Applied Economics, 30(5), 643–647. Johnson, B., Groothuis, K., Whitehead, P. A., & J. C. (2001). The value of public goods

10