Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 1 Impact of Institutional Distance on Environmental and Social Practices in 2 Host Countries: Evidence from International Construction Companies 3 Meng Ye1, Weisheng Lu2, and Fan Xue3 4 5 Abstract 6 Construction businesses expanding internationally often need to devise corporate social responsibility 7 (CSR) as an indispensable component of their competitive strategies. Companies will customize their 8 CSR programs in line with host countries’ institutional environments, meanwhile, this customization 9 will be unavoidably influenced by the institutions at home countries. This research aims to explore 10 whether the institutional distance between home and host countries matters to CSR, in particular its 11 related environmental and social practices. Data regarding the CSR practices in host countries are 12 extracted from CSR/sustainability reports by using content analysis and text mining. Logistic regression 13 models are then applied to test the roles of institutional distance and host country contexts on the two 14 types of CSR practices. It is found that the institutional distance has no impact on environmental 15 practices in host countries, but the embedded contexts of host countries positively affect the practices. 16 It is also found that the institutional distance is positively correlated with social practices; however, the 17 positive relationship is less pronounced when the host country’s development level is higher. The 18 novelty of this research lies in considering both host countries’ contexts and the institutional distance. 19 The findings offer companies new insights on how to engage in environmental and social practices and 20 develop CSR strategies in international construction markets. 21 22 Keywords: Institutional distance, corporate social responsibility, environmental practice, social 23 practice, legitimacy, host country context 1 Ph.D. Associate Professor. 1. School of Economics and Management, Southwest Jiaotong University, Chengdu 610031, China; 2. Service Science and Innovation Key Laboratory of Sichuan Province, Chengdu 610031, China. Corresponding author. Email: mengye@swjtu.edu.cn 2 Ph.D. Professor. Dept. of Real Estate and Construction, Faculty of Architecture, the University of Hong Kong, Pokfulam, Hong Kong 3 Ph.D. Assistant Professor. Dept. of Real Estate and Construction, Faculty of Architecture, the University of Hong Kong, Pokfulam, Hong Kong 1 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 24 1. Introduction 25 In line with the Sustainable Development Goals (SDGs) increasingly promoted by the United Nations 26 (2015), the Paris Climate Agreement, and the worldwide agreement on Carbon Neutrality, companies 27 nowadays are expected to balance environmental, social, and economic sustainability (Mio et al. 2020). 28 Nevertheless, construction companies are often criticized for their adverse environmental and societal 29 effects (Lu et al. 2016). This is particularly true when construction companies are competing in the 30 international arena with different environmental, social, and economic conditions. Without fully 31 considering these different conditions, international construction companies (ICCs) may be faced with 32 criticism, loss, and even project failure. For example, the media reported a big concern on the potential 33 damage to the environment and communities from Chinese-backed hydropower projects along the 34 Mekong River (Gokkon 2018), and the construction of the dam in Myanmar has been even suspended 35 by the local government after nationwide protests due to their concerns on the loss of livelihoods, wide- 36 scale environmental damage, and destruction of cultural heritage sites (HumanRightWatch 2019). ICCs 37 are compelled to embrace environmental and social practices to respond to public concerns, build 38 reputations, and attain legitimacy in the long term (Bondy et al. 2012; Zheng et al. 2015). 39 40 According to institutional theory, environmental and social practices are shaped, mediated, and 41 channeled by the institutional environment faced by companies (Jackson and Rathert 2016). When 42 operating in international markets outside their home countries (i.e., “host countries”), ICCs are 43 sensitive to host-country attributes (Lee et al. 2016) because construction business is very much site- 44 specific and cannot be disconnected with a local social, economic, environmental, and legal context. 45 Local conditions and corresponding expectations from stakeholders and the general public in a 46 particular context are the most crucial factors for construction companies (Jackson and Rathert 2016) 47 to gain social acceptability and credibility in host countries (Scott 2008) and to survive and thrive in a 48 competitive market. 49 2 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 50 Whilst the importance of local conditions for a construction business is constantly emphasized, scholars 51 also stress the importance of capturing home-country institutional forces simultaneously. It is 52 understandable that the powerful influence of a head office, as well as culture and norms along with the 53 entire organization, will impact the overall strategies, capabilities, and core operations of the 54 organization (Kolk 2005). Hence, researchers in international business emphasize the importance of 55 institutional distance between a country pair to reflect the in-between tension of diverse institutional 56 environments (Kostova and Zaheer 1999). Kostova (1997) defines institutional distance as “the extent 57 of dissimilarity between regulatory, cognitive, and normative institutions of two countries”, which has 58 been used as a theoretical lens to examine corporate social responsibility (CSR) of multinational 59 enterprises. Campbell et al. (2012) state that a greater distance from home countries may on the one 60 hand strategically motivate companies to conduct host-country CSR to reduce the liability of 61 foreignness, while they may have less willing or reduced ability to engage in CSR on the other. 62 Jacqueminet (2017) argues that the relationship between the institutional distance and CSR relies on 63 the subsidiaries’ relative needs for internal versus external legitimacy. However, little research has 64 considered construction companies’ concerns on environmental and social practices, and whether 65 institutional distance still matters to their CSR practices in host countries. 66 67 This research aims to explore whether the institutional distance has impacts on the environmental and 68 social practices by focusing on the relationship in the international construction context. Firstly, the 69 construction industry provides a particular context, where construction business is largely embedded in 70 host countries. The host-country effects may be more significant due to the industry’s project-based and 71 field-oriented production characteristics. Secondly, the controversial feature of the construction 72 industry leads to two diverse considerations of the practices: environmental practices are conducted to 73 decrease the negative externality, compliance with regulative standards in specific countries; while 74 social practices are to increase positive externality by responding to moral pressures or legitimate 75 requirements arise from local communities (Bustamante 2011; Ye et al. 2020). This research seizes the 76 opportunity to address the different kinds of practices and figure out the diverse effects on 77 environmental and social practices rather than treating CSR as a whole. Thirdly, this research examines 3 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 78 specific practices in the environmental and social aspects in host countries. Rather than relying on the 79 database, which can only provide CSR performance for the entire company, data with respect to 80 environmental and social practices is extracted by using content analysis with the assistance of text 81 mining. 82 83 The remainder of this paper is structured as follows. Section 2 provides a theoretical background for 84 this research, based on which, hypotheses are developed and illustrated in a conceptual framework. 85 Section 3 introduces the sample and the process of extracting environmental and social practices from 86 CSR reports by using content analysis and text mining. It also elaborates the steps of using logistic 87 regression models. Section 4 presents the results of the tests. Section 5 discusses the results and Section 88 6 draws the conclusion. 89 90 2. Theoretical background and hypotheses 91 2.1 Environmental and social practices 92 Environmental practices are responses to environmental issues, which loom large in the CSR agenda 93 being addressed by construction companies (Jones et al. 2006; Lu et al. 2018). Construction activities 94 have significant adverse impacts on the environment, including dust and greenhouse gas emissions, 95 noise pollution, discharged water and waste, and land degradation (Lu et al. 2016). To this end, 96 construction companies emphasize environmental practices attempting to minimize their negative 97 impacts on the environment (Jiang and Wong 2016). For instance, ISO 14000 series environmental 98 management systems are applied by many construction companies to guarantee their environmental 99 practices. Some construction companies demonstrate their energy-intensive activities and operate in 100 strict regulatory environments. Environmental practices are always treated under international standards 101 such as UN Global Compact since environmental protection attracts global attention with a common 102 set of values, core strategies, and policies across the company (Bustamante 2011). 103 4 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 104 From a social perspective, the construction industry is a critical component of the labor market and 105 generates large numbers of jobs (Zhao et al. 2012). However, with regards to the controversial nature 106 of the construction industry, construction companies are supposed to contribute extra efforts to improve 107 social welfare (Jiang and Wong 2016). For example, they emphasize the commitments and 108 responsibilities they have to the communities in which they work (Jones et al. 2006). Ye et al. (2020) 109 also emphasize the on-site community engagement practices for the conduct of construction projects 110 because they need to get the operating license. Construction businesses are helpful to deal with social 111 issues, including the support for community construction (Jones et al. 2006), reconstruction for disaster 112 relief (Ye et al. 2018), and building schools and providing electricity (Ye et al. 2020). These social 113 issues are cultural grounds with culture-specific expectations (Bustamante 2011; Ye et al. 2020). Social 114 practices are seldom regulated but conducted with local CSR strategy responsive to the local context 115 and local stakeholders (Collinge 2020; Duran and Bajo 2014). 116 117 2.2 Institutional distance and business operations 118 Institutional distance, meaning “the extent of dissimilarity between host and home institution” (Kostova 119 1997), is derived from a strand of institutional theory where institutional environment influences firm 120 structure and behavior (DiMaggio and Powell 1983). Organizational theorists have suggested that 121 institutional environments comprise a variety of institutions, including regulations, educational systems, 122 norms and cultures, and so on (Kostova and Zaheer 1999). Operating in both home and host countries, 123 companies are faced with multiple institutional environments and thus with a divergence of institutional 124 environments. The main explanation of why institution matters here is regarding legitimacy- countries 125 have diverse institutions, and therefore, diverse ways of conducting certain functions to be viewed as 126 “legitimate” (Kostova and Zaheer 1999). 127 128 Institutional distance has been used to explain multinational enterprise behaviors, such as location 129 choice (Xu and Shenkar 2002), entry mode choice (Hernández and Nieto 2015), ownership strategy 130 (Eden and Miller 2004), and staffing strategies (Ando and Paik 2013) mainly with a basic argument 131 that institutional distance affects “the liability of foreignness operating in host countries” (Eden and 5 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 132 Miller 2004), which refers to the costs of doing business abroad (Quer et al. 2019) due to the lack of 133 familiarity with the local environment. This largely affects the location choice and modes of entry 134 strategies, and once they have entered the market, institutional distance impacts the “establishment of 135 organizational legitimacy in host countries” (Kostova and Zaheer 1999). In other words, the larger 136 degree of dissimilarity between the two countries referring to the larger institutional distance, the more 137 possibilities that companies’ activities are improper or impropriate in host countries, leading to greater 138 pressures on companies for local responsiveness (Prahalad and Doz 1987). 139 140 CSR, environmental and social practices particularly, has been conceptualized as a set of legitimation 141 strategies that multinational enterprises adopt to refine the role of business in society (Castelló and 142 Lozano 2011) and to respond to different forms of normative and coercive stakeholder influence 143 (Rathert 2016). Multinational enterprises at a greater distance from the home countries should therefore 144 be strategically motivated to engage in host-country environmental and social practices (Campbell et 145 al. 2012), but they may have less ability for these practices because they have fewer resources and the 146 adaptation to local norms is too expensive (Jacqueminet 2017). Meanwhile, the influence of institutional 147 distance is largely in line with CSR strategies drawn from the underlying global-local considerations 148 (Bustamante 2011) and types of isomorphism pressure (Ye et al. 2020). 149 150 2.3 Host country context, institutional distance, and environmental practices 151 In line with the global sustainable development goals and mitigating climate changes, countries around 152 the world have their own environmental concerns, regulations, or policies, which are formulated by 153 governments or customer green demands (Qiu et al. 2020). Some of these regulations are not specific 154 to the construction industry, but they inevitably regulate it owing to construction’s roles. By exploring 155 ICCs’ environmental practices, Ye et al. (2020) state that environmental practices are conducted under 156 compliance pressures with well-established regulations, rules, or laws. Institutional theory suggests 157 firms conform to prevailing societal rules and beliefs regarding environmental protection in the 158 operating context (Wang et al. 2019) to establish regulative legitimacy (Scott 2008) or normative 159 legitimacy (Suddaby et al. 2016). Environmental practices are thus highly conditioned by the 6 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 160 institutional profiles of the different countries where firms operate. Babiak and Trendafilova (2011) 161 echo that executives are motivated to address environmental issues to conform to external institutional 162 pressures, such as acquiescence to government directives or facilitating compliance with environmental 163 laws and regulations. 164 165 With regards to the specific environmental regulations in different host countries, Rasoulkhani et al. 166 (2020) report that the whole process of capital projects in infrastructure sectors would be significantly 167 affected. Chowdhury et al. (2020) try to propose a regulatory future-proofing process that includes 168 instructions and guidance for regulated entities in construction and infrastructure sectors to proactively 169 adapt to the outside environmental regulatory context. Lu et al. (2013) also report the reasons why 170 construction companies conduct environmental practices, one of which is the government regulation 171 motivations and the increasing pressure from large green companies on the supply chain. The context 172 where construction companies operate with higher environmental performance provides them with 173 either regulative or normative pressure to conduct environmental practices. Therefore, the first 174 hypothesis is that: 175 H1: The environmental practice is positively correlated with the host country context where companies 176 are embedded 177 178 Host country context sets a regulative baseline for international companies to conduct environmental 179 practices, especially for ICCs, whose construction business is largely embedded in the contexts. The 180 regulative baseline is strong enough that every company needs to follow. Aguilera-Caracuel et al. (2013) 181 state that considering environmental regulations as coercive, companies exhibit different levels of 182 environmental practices that correspond to each country’s legal requirements. There are two distinct 183 situations: 1) when the legal requirements are high in the host countries, ICCs intend to conduct 184 environmental practices to achieve regulative and normative legitimacy in the context regardless of the 185 existing institutional distance. For example, due to the high level of energy requirements in Germany, 186 ICCs operating there mandate a certain percentage of energy to be obtained from renewable or 187 alternative sources. When the compliance pressure in the host country is high, ICCs prefer to reach the 7 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 188 regulative baseline but are restricted to conduct more for best practices. 2) When the requirements on 189 environmental issues in the host country are lower than those in the home country, ICCs would also 190 demonstrate their home-country environmental practices in the host countries, for example, Spanish 191 construction companies would like to conduct environmental practices, such as energy-efficient 192 equipment and waste management in Latin American markets even though these countries are with 193 lower environmental requirements (Ye 2018). However, institutional distance makes no sense in these 194 two situations. So, there comes the second hypothesis that: 195 H2: The institutional distance between home and host countries does not matter to companies’ 196 environmental practices in the host countries 197 198 2.4 Host country context, institutional distance, and social practices 199 Organizations need to conform to or be consistent with established cognitive structures in the society 200 to pursue legitimacy (Kostova and Zaheer 1999). Jackson and Apostolakou (2010) find that firms in 201 high-impact industries would respond to the pressures by adopting more extensive practices. Ye et al. 202 (2020) describe social practices as contextualized practices that are characterized by a strong cultural 203 grounding or societal needs in the context of the international construction business. For example, 204 disaster-relief initiatives are always conducted as the priority in disaster-prone countries, such as 205 donations and initiatives related to the construction business, i.e. structural assessments and 206 infrastructure evaluation. 207 208 When in a country with a lower development level, the country may be faced with higher societal needs, 209 various poverty issues, healthy issues, or education problems, it is easier for companies to seek 210 legitimacy by social practices. For example, ICCs can initiate activities called “answer to hunger” or 211 “relief for the poor” in some African countries and deal with the clean water problem to respond to 212 poverty issues; provide free medical care or suitable medical infrastructure to respond to health issues; 213 and build schools and provide electricity to deal with basic education problems. While in a country with 214 a higher quality of life, social practices are less implemented for companies for the purpose of 8 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 215 establishing legitimacy with the most frequently mentioned social practices in the CSR reports being 216 communication with local communities. The third hypothesis is: 217 H3: The social practice is negatively correlated with the development level of the host country where 218 companies are embedded 219 220 When going abroad, a company often faces entry barriers (e.g. culture shock) in a host country market 221 (Eden and Miller 2004). Institutional distance between home and host countries undoubtedly enhances 222 the barriers. For example, it increases the liability of foreignness, which can be decomposed into 223 unfamiliarity, relational and discriminatory hazards (Eden and Miller 2004; Salomon and Wu 2012). In 224 other words, the institutional distance reduces the legitimacy of existing practices of companies from 225 home countries and increases demands to establish cognitive legitimacy (Meyer and Scott 1983). For 226 example, initiatives to deal with the basic education of children are not suitable for the developed 227 countries. 228 229 The institutionally distant host country increases the salience of a given social issue, whereby 230 stakeholders are seen as legitimate claimants to practices in the absence of regulation (Rathert 2016). 231 For construction companies, social issues stand out due to the industry’s controversial features. For 232 example, construction projects always affect the living environment of local communities with their 233 adverse impacts such as land occupation, noise and dust pollution, and even hazardous substances. 234 Distinguished from the environmental practices, which are conducted with the compliance pressure, 235 ICCs are expected to conduct more social practices to address the social issues and gain cognitive 236 legitimacy from stakeholders in the institutional distant host country. Meanwhile, companies’ 237 operations are placed in a stricter examination context than domestic companies in the host countries, 238 compared with which, construction companies are expected to undertake social practices to meet 239 societal needs, to build their reputation, and to seek cognitive legitimacy. There thus comes the fourth 240 hypothesis that: 241 H4: The social practice is positively correlated with the institutional distance between home and host 242 countries 9 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 243 244 Based on the literature review and hypothesis formation, a conceptual framework of this research is 245 developed, as shown in Fig.1. 246 <> 247 248 3. Research Methods 249 The research methods are devised to test the hypotheses as derived in the last section. They comprise 250 three interconnected steps: (1) Data preparation, (2) Defining variables and measures, and (3) 251 Hypothesis testing. 252 253 3.1 Data preparation 254 3.1.1 Sample and document collection 255 The sample of international construction companies (ICCs) is derived from two sources: one is the top 256 international contractor lists compiled by Engineering News Records (ENR), a construction, building, 257 and engineering-oriented magazine, and the other is the Sustainability Disclosure Database of Global 258 Reporting Initiative (GRI), which incorporates a detailed overview of CSR/sustainability reports of 259 ICCs published from 1999 to 2018. By excluding some of the companies due to a lack of 260 CSR/sustainability reports, 68 ICCs are chosen as the sample for the analysis. 261 262 CSR/sustainability reports of the sampled ICCs over seven years (2011-2017) are retrieved from their 263 websites or the GRI’s database. Either sustainability or CSR reports are collected as both disclose 264 similar contents regarding environmental and social practices. Not all ICCs disclose reports every year 265 and some 2017 CSR reports are not included as they are not yet released at the collection date (March 266 2018). In total, 369 CSR/sustainability reports are collected for the analysis (shown in Table 1). 267 <> 268 269 10 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 270 3.1.2 Environmental and social practices extraction 271 Environmental and social practices of ICCs in specific host countries (as dependent variables) were 272 extracted from CSR/sustainability reports by using a latent semantic analysis (LSA)-assisted content 273 analysis. LSA is defined as “a theory and method for extracting and representing the contextual-usage 274 meaning of words by statistical computations applied to a large corpus of text” (Landauer et al. 1998) 275 and has been applied to analyze construction documents (Mahfouz and Kandil 2010; Yalcinkaya and 276 Singh 2015). It can derive measures of the similarity of the word meanings from texts, assisting the 277 content analysis by allocating words (i.e., input) or texts into categories (i.e. output). Particularly, it can 278 automatically process a large volume of documents. Therefore, LSA is adopted in this study by 279 allocating the texts in the CSR/sustainability reports into topics of environmental and social practices. 280 The LSA-assisted content analysis process is presented in Fig. 2 and elaborated as follows. 281 <> 282 283 Firstly, topics related to environmental and social practices are identified based on the GRI’s (2014) 284 Construction and Real Estate Sector Supplement (“CRESS”) Guidelines. They are illustrated in Table 285 2 to help decode CSR reports to different topics as interested. 286 <> 287 288 Then, manual decoding is conducted to provide a reliable baseline for the subsequent text mining using 289 machine. A total of 50 CSR/sustainability reports from 8 ICCs were chosen. Texts were first coded 290 according to the country names using the qualitative data analysis software NVivo Pro 11. Coded texts 291 were then manually classified into the 9 topics shown in Table 2. For example, if the text mentioned 292 “biodiversity”, this text was classified into the topic of biodiversity conservation (EN2) marked with 293 “1”. CSR reports were regarded as a group over the past seven years to identify environmental and 294 social practices in host countries. Through this decoding method, the researchers care about which topic 295 has been demonstrated over the years in each country regardless of the specific year. In total, there were 296 139 groups of extracted texts describing the ICCs’ environmental and social practices in each country. 297 This manual decoding process was conducted by 2 coders to ensure the reliability of the results. 11 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 298 299 Next, a pilot LSA text mining is then conducted using the 50 CSR reports chosen from the 8 ICCs as a 300 sample. It is conducted by using genism (version 3.4, available at: https://github.com/RaRe- 301 Technologies/gensim) in Python 3.6.5. Evaluation method (F1 score) and cross-validation with the 302 manual decoded results are then conducted to validate whether the LSA can be scaled up to the 369 303 CSR reports from all the 68 ICCs. 304 305 Finally, the LSA-assisted text mining process is applied to all the collected reports to extract the data 306 of environmental and social practices for this study. The text mining process is quite useful and robust. 307 However, it is too technical and lengthy to be presented in this paper. Therefore, it is added as 308 supplementary material to keep the main storyline of this paper. 309 310 3.1.3 Extracted data of environmental and social practices 311 After the extraction process, the documents can be allocated to a specific topic with the doc-to-topic 312 similarity score and the score can be further converted into a “0-1” binary score based on the threshold 313 of the topic. “1” indicates ICCs have conducted the practice while “0” indicates not. Therefore, we can 314 get “0-1” data for the environmental and social practices of each ICC in specific host countries. The 315 data were further cleaned by removing items with the same home and host countries. As a result, 811 316 items describing environmental and social practices in 55 host countries for 66 ICCs are derived for this 317 study. 318 319 3.2 Variables and measures 320 3.2.1 Dependent variables 321 The dependent variables would be environmental and social practices of ICCs in host countries, which 322 are shown in the secion “Extracted data of environmental and social practices”. There are only two 323 possible values for the dependent variable, represented by a dichotomous variable “1” and “0”. One 324 assumption here is involved that the country names were mentioned in the reports only when ICCs 325 conduct specific environmental or social practices. 12 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 326 327 3.2.2 Independent and moderating variables 328 Based on the hypothesis, independent and moderate variables are different for environmental and social 329 practices. Basically, institutional distance and host country contexts (including environmental 330 performance index and human development index) are the two variables, either as an independent or a 331 moderating variable. 332 333 Institutional distance is described as the degree of dissimilarity in institutions between two countries 334 (Gaur and Lu 2007). To measure institutional distance, the Worldwide Governance Indicators (WGI) 335 are selected, encompassing the broadest range of institutional issues (Ando and Paik 2013). WGI is a 336 research dataset summarizing the views on the quality of national governance, with six dimensions 337 “voice and accountability, political stability and absence of violence, government effectiveness, 338 regulatory quality, the rule of law, and control of corruption”, describing a wide range of evaluation on 339 governance including “the capacity of a government to effectively formulate and implement sound 340 policies, the respect of citizens, and the state for the institutions that govern social and environmental 341 interactions among them” (WorldBank 2021). To operationalize institutional distance, this research 342 used the Euclidean distance measure following Gaur and Lu (2007) and Konara and Mohr (2019) (see 343 Equation 1). 344 𝐷𝐼𝑆_𝐼𝑁𝑆 = √∑6𝑘=1(𝐼𝑘𝑖 − 𝐼𝑘𝑗 )2 345 Where, 𝐷𝐼𝑆_𝐼𝑁𝑆 is the institutional distance between country i and country j, 𝐼𝑘𝑖 is country i’s score 346 of the kth dimension in WGI, there are 6 dimensions involved in WGI. Equation 1 347 348 Environmental Performance Index (EPI) is a method of quantifying and numerically marking the 349 environmental performance of a country’s policies on the environment, which includes 32 indicators 350 organized into 11 issue categories in two policy objectives of environmental health (40%) and 351 ecosystem vitality (60%), such as air quality, sanitation, and drinking water, heavy metals, waste 352 management, biodiversity and habitat, ecosystem services, climate change, and water resources. This 13 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 353 index provides a data-driven summary of the state’s sustainability around the world, which can reflect 354 the operating context for environmental practices. 355 356 Human Development Index (HDI) is a concept developed by the United Nations, describing a 357 comprehensive approach comprising health, education, and economic dimensions (UNDP 2016). Each 358 country has scores ranging from 0 to 1 based on the three dimensions. The HDI considers national 359 policy choices to measure economic growth and human development. It is thus regarded as a proxy to 360 describe the development levels of countries, reflecting the health of the economy, the development of 361 industries, the rise of non-profit organizations, and communication and education, thus describing the 362 operating context for social practices. 363 364 3.2.3 Control variables 365 Two corporate-level control variables are chosen for the analysis: firm size and degree of 366 internationalization. Firm size. With greater influence on society, ICCs with larger firm sizes are desired 367 to conduct more environmental and social practices. They are also assumed to have the ability and 368 financial support for the practices. Firm size is measured by using a natural logarithm of the average 369 total revenue from 2011 to 2016, which reflects the average size during the period of collected CSR 370 reports. Degree of internationalization. When their degree of internationalization is higher, ICCs may 371 have more opportunities to implement environmental and social practices in overseas markets to 372 establish their legitimacy. The degree of internationalization is measured by the average proportion of 373 overseas revenue in total revenue during the period from 2011 to 2016. 374 375 Two variables are used to control the characteristics of the host markets. Regions indicate whether the 376 host markets are in the same region as the country-of-origin of ICCs. It is assumed that host markets 377 which are geographically close to the country of origin should have more transportation and 378 communication links. The degree of openness of the host markets to foreign companies is controlled by 379 using the prevalence of foreign ownership in the Global Competitiveness Index. 380 14 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 381 3.3 Estimation methods 382 Three steps are conducted for a logistic regression model. Step 1 is to form a logistic regression model 383 and to test the relationships between the dependent, independent, and moderating variables (described 384 above). In this step, p-values for variables are calculated to indicate the goodness fit of the coefficients. 385 Step 2 is to carry out the odds ratio test to indicate the effects of independent variables. An odds ratio 386 is a measure of association between the independent variable and the dependent variable. An odds ratio 387 greater than 1 indicates that the independent variable is connected with higher odds of the dependent 388 variable; while an odds ratio of less than 1 indicates the opposite situation (Menard 1995). Step 3 is to 389 carry out tests to indicate the goodness fit of a logistic regression model, such Wald test, and the 390 Hosmer-Lemeshow test. The Wald test is a statistical test used for comparing the goodness of fit of two 391 statistical models. If the p-value is less than 0.05, it indicates the model fits well. Hosmer-Lemeshow 392 test is used to further indicate the goodness fit of the logistic regression model. A Hosmer-Lemeshow 393 (H-L) statistic with a p-value greater than 0.05 is considered a good fit (Hilbe 2009). 394 395 4. Analyses and results 396 4.1 Correlation matrix and descriptive statistics 397 Table 3 shows the correlation matrix and descriptive statistics. The mean values indicate that, on 398 average, 35% of the whole items (N=811) show engagement in environmental practices, and 53% in 399 social practices. The correlation matrix in Table 3 shows that the institutional distance has correlations 400 with the environmental and social practices (r=-0.18, p<.01 and r=0.087, p<0.05 respectively), the host 401 country environmental performance index (Host_Envi) correlates with environmental practices 402 (r=0.217, p<0.01) and the host country human development index (Host_HDI) correlates with social 403 practices (r=-0.146, p<0.01). The host country environmental performance index and human 404 development index are highly correlated (r=0.896, p<0.01). Whether the region of the host country is 405 the same as the country of origin (the dummy variable) has negative correlations with the institutional 406 distance, indicating that geographical distance can affect the institutional distance. 407 <> 15 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 408 409 4.2 Results of hypotheses testing 410 The hypotheses testing was conducted by using R programming. Three parts of results following three 411 steps of logistic regression models are presented in Table 4, including the regression results, the 412 goodness of fit of coefficients, the results of odds ratios tests, and the Hosmer-Lemeshow tests to 413 indicate the goodness fit of the model. Models 1 and 2 reflect the effects of institutional distance and 414 environmental contexts of host countries on environmental practices; while Model 3 and Model 4 report 415 the effects of institutional distance and social contexts of host countries on social practices. 416 <> 417 418 Model 1 presents a significant positive effect of Host_EPI on environmental practices of ICCs with the 419 coefficient of 0.033 (p<.001). The odds ratio for Host_EPI is 1.0337, indicating that the odds of a one- 420 unit increase in Host_EPI would make the odds of environmental practices increase by 3.37% 421 (calculated by 1.0337 minus 1). The p-value for the Wald test is below 0.001, indicating the addition of 422 independent variables to the model is significantly better than the constant-only model. The Hosmer- 423 Lemeshow statistic for Model 1 is 13.481, resulting in a p-value of 0.096, which provides statistical 424 evidence of a well-fitted model. H1 is thus supported that the host country environmental context has 425 positive impacts on environmental practices. By adding the indicator of institutional distance, Model 2 426 presents the effects of Host_EPI and the interaction with institutional distance on environmental 427 practices, but the results show a non-significant effect of institutional distance and the Host_EPI’s effect 428 is significant at the 10% level. Therefore, H2 can be supported. 429 430 Model 3 concerns the social practices of ICCs. The coefficient of Host_HDI is -3.337, which is negative 431 and significant (p<.001). The odds ratio for Host_HDI is 0.0356, indicating that the odds of a one-unit 432 increase in Host_EPI would make the odds of social practices decrease by 96.44% (calculated by 0.0356 433 minus 1). The p-value for the Wald test is below 0.001, indicating the addition of independent variables 434 to the model is significantly better than the constant-only model. The Hosmer-Lemeshow statistic for 16 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 435 Model 3 is 4.143 with a p-value of 0.844 presenting statistical evidence of the well-fitted model as well. 436 H3 is thus supported. 437 438 Model 4 presents an interesting result that Host_HDI’s effects on social practices become not significant 439 when considering the institutional distance. It shows that the institutional distance has a significant 440 positive impact on social practices at the 10% level, and the positive impact would be negatively 441 affected by the Host_HDI since the coefficient of the interaction term of Host_HDI and institutional 442 distance is negative and significant at the 10% level. The odds ratio of the institutional distance is 0.7937, 443 indicating that the odds of a one-unit increase in the institutional distance would make the odds of social 444 practices decrease by 20.63% (calculated by 0.7937 minus 1). The p-value for the Walt test in Model 4 445 is below 0.001 and the Hosmer-Lemeshow statistic is 4.146 with a p-value of 0.844, indicating the good 446 addition of the independent variables and a well-fitted model. The results can refine H4 that social 447 practices are positively correlated with the institutional distance and the host country social context 448 would decrease the effects of the institutional distance. 449 450 The firm size is found to have no effects on both environmental and social practices, which means these 451 practices are conducted regardless of their size influence on society. It is found that the degree of 452 internationalization matters to the environmental and social practices at the 10% level, indicating that 453 ICCs with a higher degree of internationalization have more possibilities to conduct environmental and 454 social practices. Whether the host markets are in the same region as the country-of-origin of ICCs does 455 not matter to the environmental and social practices. But the degree of openness of the host markets has 456 positive impacts on social practices at the 1% level, indicating that when the host market is more open, 457 ICCs are more likely to conduct social practices. 458 459 5. Discussions 460 The macro-institutional pressures from the companies’ embedded environment have influences on their 461 environmental and social practices (e.g. Campbell 2007; Jackson and Apostolakou 2010). The 17 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 462 influences would be high for construction companies in particular, due to the industry’s fixed final 463 product, long production cycle, and disruptive nature of construction projects to the environment and 464 society. Ye et al. (2020) emphasize the impacts of the local grounding on environmental and social 465 initiatives for ICCs. Our results confirm that the host country contexts where ICCs are embedded have 466 impacts on environmental and social practices (in line with H1 and H3). 467 468 The results further show that these two kinds of practices on the environment and society present distinct 469 concerns. The higher environmental requirements of host countries lead to more environmental 470 practices (consistent with H1) since environmental practices are under formal power acting as the 471 standards-based practices to establish regulative and normative legitimacy (DiMaggio and Powell 1983). 472 For example, tougher regulations on green buildings and waste management for projects lead to better 473 practices on green buildings and waste management when comparing the green building projects of the 474 US and China (Chi et al. 2020). Social practices, on the contrary, are conducted in the absence of 475 regulation to meet societal needs and gain cognitive legitimacy. The societal needs are opposite to the 476 development levels of the host countries so that higher development levels lead to fewer social practices 477 (consistent with H3). This may be because social practices in less developed countries are easily 478 regarded as best practices to be reported for establishing legitimacy. 479 480 For construction companies, the influence of institutional distance cannot be investigated without regard 481 to the host country context due to the characteristics of the construction business as stated previously. 482 The results show that the institutional distance does not matter to ICCs’ environmental practices 483 (consistent with H2) by considering the impacts of both institutional distance and host country contexts. 484 On the one hand, when the compliance pressure in host countries is strong, i.e. with strict regulations 485 or policies on the environment, ICCs have no choice but to follow the regulations to establish regulative 486 legitimacy (Scott 2005). Under this circumstance, ICCs from large institutional distant countries have 487 no need to practice more to establish legitimacy because the baseline is so high that they need abundant 488 investments to meet the requirement. On the other hand, when the compliance pressure is low in host 489 countries, ICCs may promote an integrated approach among the company in case of environmental 18 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 490 issues as these issues are globally concerned (Bustamante 2011; Ye et al. 2020). In this sense, ICCs 491 may not consider too much on the institutional distance which reflects the in-between tensions of the 492 countries. 493 494 Social practices, however, are with diverse considerations. There are usually no policies or regulations 495 for practices with regard to social issues. Social practices are conducted with positive externality to 496 establish legitimacy and are expected as one way to deal with the concerns caused by the institutional 497 distance (Campbell et al. 2012). ICCs prefer to conduct straightforward social practices, such as 498 donation, community volunteer work, or disaster relief, to address the needs in the host markets as well 499 as to avoid cultural conflicts when the institutional distance is large (consistent with H4). The results 500 also find that the development levels of host countries negatively affect the relationship between 501 institutional distance and social practices. It is argued that ICCs are more likely to “do the right thing” 502 to recede the impacts caused by the institutional distance, and this can be easily addressed when the 503 societal needs are obvious in host markets where the development levels are low. 504 505 6. Concluding remarks 506 This research reports on whether the institutional distance between home and host countries matters to 507 the environmental and social practices a construction company conducts in a particular country. Using 508 empirical data, it tested the effects of host country contexts and institutional distance, as well as their 509 interactive effects in the context of international construction. The results support the effects of host 510 country contexts measured by the environmental performance index on environmental practices to be 511 significantly positive, but the institutional distance does not matter to the environmental practices in the 512 host countries. However, strong evidence is presented that social practices of ICCs are positively 513 correlated with the institutional distance between home and host countries while the positive 514 relationship would be less pronounced when the development level of the host country is higher. 515 19 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 516 This research makes three contributions to the knowledge of connecting institutional theory to 517 environmental and social practices. First, by extracting environmental and social practices from ICCs’ 518 CSR/sustainability reports, this research emphasizes the specific practices instead of evaluating 519 corporate social performance scores. Second, it is argued in this research that CSR is not a single 520 practice, but involving different kinds of concerns and practices. Environmental and social practices are 521 analyzed with distinct considerations. The effects of the institutional distance and host country contexts 522 are reflected to be different in this research and we also call for independent discussions on different 523 dimensions of CSR. Third, this research enriches the knowledge of the interactive effects of host country 524 contexts and the institutional distance on environmental and social practices in the context of 525 international construction responding to the characteristics of the industry, where host country contexts 526 cannot be neglected. 527 528 One practical implication is that ICCs need to examine the host country contexts as well as the tensions 529 of the home and host countries to engage in environmental and social practices and develop CSR 530 strategies in international construction markets. Other recommendations to the companies are host 531 country contexts are vital to the implementation of the construction projects and even the success of the 532 projects. ICCs need to evaluate the contexts to determine whether and how to exploit environmental 533 and social practices to establish legitimacy. Another implication is that ICCs are suggested to 534 distinguish environmental and social practices- one is under regulations with a purpose to decrease 535 negative externalities and the other is in the absence of regulation to increase positive externalities. 536 While environmental practices are conducted to reach the baseline, social practices have the role to 537 eliminate the negative effects caused by institutional distance and establish legitimacy. 538 539 This research is not immune from a shortcoming, therefore, requires future investigation. First, our 540 sample and data on environmental and social practices are largely based on CSR/sustainability reports. 541 Companies from developed countries with greater needs for CSR disclosure are more likely to be chosen 542 as the sample for this research. Future studies are recommended to exploit multiple methods including 543 interviews for collecting data. Second, to improve the accuracy of the logistic regression model applied 20 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 544 in the quantitative approach for hypotheses testing, future studies can use a more specific proxy for 545 measuring institutional distance. Third, future studies can extend the research to further explore other 546 CSR aspects such as practices regarding labor and human rights with considerations on the internal and 547 external legitimacy establishment. 548 549 7. Acknowledgement 550 Some part of the work described in this paper was extracted from the Ph.D. thesis supported by the 551 University of Hong Kong. 552 553 8. Data Availability Statement 554 Some or all data, models, or code that support the findings of this study are available from the 555 corresponding author upon reasonable request. 556 557 9. Supplemental Materials 558 Appendixes S1 and S2, Figs. S1 and S2, and Tables S1 and S2 are available online in the ASCE Library 559 (www.ascelibrary.org). 560 561 10. Reference 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 Aguilera-Caracuel, J., Hurtado-Torres, N. E., Aragón-Correa, J. A., and Rugman, A. M. (2013). "Differentiated effects of formal and informal institutional distance between countries on the environmental performance of multinational enterprises." J. Bus. Res., 66(12), 2657-2665. http://doi.org/10.1016/j.jbusres.2013.04.002 Ando, N., and Paik, Y. (2013). "Institutional distance, host country and international business experience, and the use of parent country nationals." Hum. Resour. Manag. J., 23(1), 52-71. http://doi.org/10.1111/j.1748-8583.2012.00201.x Babiak, K., and Trendafilova, S. (2011). "CSR and environmental responsibility: motives and pressures to adopt green management practices." Corp. Soc. Resp. Env. Ma., 18(1), 11-24. http://doi.org/10.1002/csr.229 Bondy, K., Moon, J., and Matten, D. (2012). "An institution of corporate social responsibility (CSR) in multi-national corporations (MNCs): Form and implications." J. Bus. Ethics, 111(2), 281299. http://doi.org/10.1007/s10551-012-1208-7 Bustamante, S. (2011). "Localization vs. Standardization: Global approaches to CSR management in multinational companies." Working Papers of the Institute of Management Berlin at the Berlin School of Economics and Law (HWR Berlin). 21 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 Campbell, J. L. (2007). "Why would corporations behave in socially responsible ways? An institutional theory of corporate social responsibility." Acad. Manage. Rev., 32(3), 946-967. http://doi.org/10.5465/AMR.2007.25275684 Campbell, J. T., Eden, L., and Miller, S. R. (2012). "Multinationals and corporate social responsibility in host countries: Does distance matter?" J. Int. Bus. Stud., 43(1), 84-106. http://doi.org/10.1057/jibs.2011.45 Castelló, I., and Lozano, J. M. (2011). "Searching for new forms of legitimacy through corporate responsibility rhetoric." J. Bus. Ethics, 100(1), 11-29. http://doi.org/10.1007/s10551-0110770-8 Chi, B., Lu, W., Ye, M., Bao, Z., and Zhang, X. (2020). "Construction waste minimization in green building: A comparative analysis of LEED-NC 2009 certified projects in the US and China." J. Clean. Prod., 256, 120749. http://doi.org/10.1016/j.jclepro.2020.120749 Chowdhury, S., Zhu, J., Rasoulkhani, K., Mostafavi, A., Jaselskis, E., Stoa, R., Li, Q., Banerjee, S., Alsharef, A., and Brannen, L. (2020). "Guidelines for Robust Adaptation to Environmental Regulations in Infrastructure Projects." J. Constr. Eng. M., 146(10), 04020121. http://doi.org/10.1061/(ASCE)CO.1943-7862.0001908 Collinge, W. (2020). "Stakeholder engagement in construction: Exploring corporate social responsibility, ethical behaviors, and practices." J. Constr. Eng. M., 146(3), 04020003. http://doi.org/10.1061/(ASCE)CO.1943-7862.0001769 DiMaggio, P. J., and Powell, W. W. (1983). "The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields." Am. Sociol. Rev., 48(2), 147-160. http://doi.org/10.1016/S0742-3322(00)17011-1 Duran, J. J., and Bajo, N. (2014). "Institutions as determinant factors of corporate responsibility strategies of multinational firms." Corp. Soc. Resp. Env. Ma., 21(6), 301-317. http://doi.org/10.1002/csr.1308 Eden, L., and Miller, S. R. (2004). "Distance matters: Liability of foreignness, institutional distance and ownership strategy." Adv. Int. Manag., 16(187-221). http://doi.org/10.1016/S07477929(04)16010-1 Gaur, A. S., and Lu, J. W. (2007). "Ownership strategies and survival of foreign subsidiaries: Impacts of institutional distance and experience." J. Manage., 33(1), 84-110. http://doi.org/10.1177/0149206306295203 Gokkon, B. (2018). "Environmentalists are raising concerns over China's Belt and Road Initiative." . (August 5, 2020). GRI (Global Reporting Initiative) (2014). "GRI-G4-Construction and Real Estate Sector Disclosures ", . Hernández, V., and Nieto, M. J. (2015). "The effect of the magnitude and direction of institutional distance on the choice of international entry modes." J. World Bus., 50(1), 122-132. http://doi.org/10.1016/j.jwb.2014.02.002 Hilbe, J. M. (2009). Logistic regression models, CRC press, Taylor & Francis Group, New York. HumanRightWatch (2019). "China: ‘Belt and Road’ Projects Should Respect Rights." . (2021, July 10th). Jackson, G., and Apostolakou, A. (2010). "Corporate social responsibility in Western Europe: an institutional mirror or substitute?" J. Bus. Ethics, 94(3), 371-394. http://doi.org/10.1007/s10551-009-0269-8 Jackson, G., and Rathert, N. (2016). "Private governance as regulatory substitute or complement? A comparative institutional approach to CSR adoption by multinational corporations☆." Multinational Corporations and Organization Theory: Post Millennium Perspectives, Emerald Publishing Limited, 445-478. Jacqueminet, A. (2017). "CSR implementation in MNEs: The role of distance and prioritization of demands." Distance in International Business: Concept, Cost and Value, Emerald Publishing Limited, 363-378. 22 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 Jiang, W., and Wong, J. K. (2016). "Key activity areas of corporate social responsibility (CSR) in the construction industry: a study of China." J. Clean. Prod., 113, 850-860. http://doi.org/10.1016/j.jclepro.2015.10.093 Jones, P., Comfort, D., and Hillier, D. (2006). "Corporate social responsibility and the UK construction industry." J. Corp. R. Estate, 8(3), 134-150. http://doi.org/10.1108/14630010610711757 Kolk, A. (2005). "Environmental reporting by multinationals from the Triad: convergence or divergence?" Manage. Int. Rev., 45(1): 145-166. http://doi.org/10.1007/978-3-322-91005-9_9 Konara, P., and Mohr, A. (2019). "Why we should stop using the Kogut and Singh index." Manage. Int. Rev. http://doi.org/10.1007/s11575-019-00378-7 Kostova, T. (1997). "Country institutional profiles: Concept and measurement." In Academy of Management Proceedings, 180-184. New York: Academy of Management. https://doi.org/10.5465/ambpp.1997.4981338. Kostova, T., and Zaheer, S. (1999). "Organizational legitimacy under conditions of complexity: The case of the multinational enterprise." Acad. Manage. Rev., 24(1), 64-81. http://doi.org/10.5465/AMR.1999.1580441 Landauer, T. K., Foltz, P. W., and Laham, D. (1998). "An introduction to latent semantic analysis." Discourse processes, 25(2-3), 259-284. http://doi.org/10.1080/01638539809545028 Lee, K.-W., Han, S. H., Park, H., and David Jeong, H. (2016). "Empirical analysis of host-country effects in the international construction market: An industry-level approach." J. Constr. Eng. M., 142(3), 04015092. http://doi.org/10.1061/(ASCE)CO.1943-7862.0001083 Lu, W., Ye, M., Chau, K. W., and Flanagan, R. (2018). "The paradoxical nexus between corporate social responsibility and sustainable financial performance: Evidence from the international construction business." Corp. Soc. Resp. Env. Ma., 25(5), 844-852. http://doi.org/https://doi.org/10.1002/csr.1501 Lu, W., Ye, M., Flanagan, R., and Ye, K. (2016). "Corporate social responsibility disclosures in international construction business: trends and prospects." J. Constr. Eng. M., 04015053. http://doi.org/10.1061/(ASCE)CO.1943-7862.0001034 Lu, Y., Cui, Q., and Le, Y. (2013). "Turning green to gold in the construction industry: Fable or fact?" J. Constr. Eng. M., 139(8), 1026-1036. http://doi.org/10.1061/(ASCE)CO.19437862.0000676 Mahfouz, T., and Kandil, A. (2010). "Unstructured construction document classification model through latent semantic analysis (LSA)." Proc., Proceeding of the 27th International Conference on Applications of IT in the AEC Industry (CIB-W78 2010). Ottawa: International Council for Research and Innovation in Building and Construction. Menard, S. W. (1995). Applied logistic regression analysis, Sage Publications, Thousand Oaks, Calif. Meyer, J. W., and Scott, W. R. (1983). "Centralization and the legitimacy problems of local government." Organizational environments: Ritual and rationality, J. W. Meyer, and W. R. Scott, eds., London: Sage, 199-215. Mio, C., Panfilo, S., and Blundo, B. (2020). "Sustainable development goals and the strategic role of business: A systematic literature review." Bus. Strateg. Environ,. http://doi.org/10.1002/bse.2568 Prahalad, C. K., and Doz, Y. L. (1987). The multinational mission: Balancing local demands and global vision, New York: Free Press. Qiu, L., Hu, D., and Wang, Y. (2020). "How do firms achieve sustainability through green innovation under external pressures of environmental regulation and market turbulence?" Bus. Strateg. Environ,. http://doi.org/10.1002/bse.2530 Quer, D., Rienda, L., Andreu, R., and Miao, S. (2019). "Host country experience, institutional distance and location choice of Chinese MNEs: The moderating effect of government official visits." Cross Cult. Strateg. M., 26(1), 24-45. http://doi.org/10.1108/CCSM-02-2018-0028 Rasoulkhani, K., Brannen, L., Zhu, J., Mostafavi, A., Jaselskis, E., Stoa, R., Li, Q., Alsharef, A., Banerjee, S., and Chowdhury, S. (2020). "Establishing a Future-Proofing Framework for Infrastructure Projects to Proactively Adapt to Complex Regulatory Landscapes." J. Manage. Eng., 36(4), 04020032. http://doi.org/10.1061/(ASCE)ME.1943-5479.0000794 23 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 Rathert, N. (2016). "Strategies of legitimation: MNEs and the adoption of CSR in response to hostcountry institutions." J. Int. Bus. Stud., 47(7), 858-879. Salomon, R., and Wu, Z. (2012). "Institutional distance and local isomorphism strategy." J. Int. Bus. Stud., 43(4), 343-367. http://doi.org/10.1057/jibs.2012.3 Scott, W. R. (2005). Institutional theory: Contributing to a theoretical research program. In K. G. Smith, & M. A. Hitt, (Eds.), Great minds in management: The process of theory development (pp. 460-484). Oxford: Oxford University Press. Scott, W. R. (2008). Instituions and Organizations: Ideas and Interests, Sage Publications, Los Angeles, CA. Suddaby, R., Bitektine, A., and Haack, P. (2016). "Legitimacy." Acad. Manag. Ann., 28. http://doi.org/10.5465/annals.2015.0101 UN (2015). "Sustainable Development Goals." . (April 10th, 2020). UNDP (2016). "Human Development Report 2016: Human Development for Everyone." United Nations, New York, USA. Wang, S., Wang, H., and Wang, J. (2019). "Exploring the effects of institutional pressures on the implementation of environmental management accounting: Do top management support and perceived benefit work?" Bus. Strateg. Environ., 28(1), 233-243. http://doi.org/10.1002/bse.2252 WorldBank (2021). "What is Governance?", . (July 12, 2021). Xu, D., and Shenkar, O. (2002). "Institutional distance and the multinational enterprise." Acad. Manage. Rev., 27(4), 608-618. http://doi.org/10.2307/4134406 Yalcinkaya, M., and Singh, V. (2015). "Patterns and trends in Building Information Modeling (BIM) research: A Latent Semantic Analysis." Automat. Constr., 59, 68-80. http://doi.org/10.1016/j.autcon.2015.07.012 Ye, M. (2018). The impacts of institutional distance on corporate social responsibility "glocalisation" : evidence from international construction, University of Hong Kong Libraries, Hong Kong. Ye, M., Lu, W., and Flanagan, R. (2018). "Corporate social responsibility localization in international construction business." Proc., Proceedings of the 21st International Symposium on Advancement of Construction Management and Real Estate, Springer, 553-564. Berlin: Springer. Ye, M., Lu, W., Flanagan, R., and Chau, K. W. (2020). "Corporate social responsibility “glocalisation”: Evidence from the international construction business." Corp. Soc. Resp. Env. Ma. http://doi.org/10.1002/csr.1831 Zhao, Z.-Y., Zhao, X.-J., Davidson, K., and Zuo, J. (2012). "A corporate social responsibility indicator system for construction enterprises." J. Clean. Prod., 29–30(0), 277-289. http://doi.org/10.1016/j.jclepro.2011.12.036 Zheng, Q., Luo, Y., and Maksimov, V. (2015). "Achieving legitimacy through corporate social responsibility: The case of emerging economy firms." J. World Bus., 50(3), 389-403. http://doi.org/10.1016/j.jwb.2014.05.001 24 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 730 Table 1. The numbers of CSR/sustainability reports collected Year CSR/sustainability reports 2011 2012 2013 2014 2015 2016 2017 Total 45 57 61 60 59 63 24 369 731 25 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 732 Table 2. Topics of environmental and social practices Code Category EN Environmental practice SO Social practice Topic EN1-Energy and carbon emission EN2-Biodiversity conservation EN3-Waste management SO1- Local community communication SO2-Donation SO3-Disaster relief SO4-Poverty caring SO5-Medical caring SO6-Youth and education 733 26 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 Table 3. The correlation matrix and descriptive statistics (N=811) Mean SD Max Min 1 2 3 4 5 6 7 8 9 1 Prac_Envi 0.35 0.016 1 0 1 2 Prac_Social 0.53 0.017 1 0 0.036 1 3 DIS_Ins 2.80 0.054 7.13 0.22 -0.180** 0.087* 1 4 Host_EPI 72.08 0.457 88.91 41.77 0.217** -0.127** -0.597** 1 5 Host_HDI 0.77 0.005 0.939 0.418 0.222** -0.146** -0.617** 0.896** 1 6 Comp_size 9.10 0.041 11.62 3.39 -0.052 0.004 -0.034 -0.026 -0.019 1 7 Comp_inte 0.48 0.010 1 0 0.079* 0.060` 0.013 0.076* 0.070* -0.220** 1 8 Dummy_reg 0.35 0.016 1 0 0.027 -0.009 -0.132** 0.088* 0.100** 0.043 -0.246** 1 9 Host_open 4.77 0.029 6.13 2.92 0.161** 0.002 -0.538** 0.603** 0.558** -0.075* 0.082* 0.053 1 Note: Prac_Envi= Environmental practice; Prac_Social= Social practice; DIS_INS= institutional distance; Host_Envi= host country environmental performance index; Host_HDI= host country human development index `p < .1, *p < .05, **p<.01 27 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 Table 4. Testing results of logistic regression models Variables Comp_size Comp_inte Dummy_reg Host_open Constant DIS_Ins Host_EPI DIS_Ins * Host_EPI Host_HDI DV= Prac_Envi (Odds Ratio) Model 1 Model 2 -0.063 -0.070 (0.9393) (0.9322) 0.495` 0.508` (1.6407) (1.6618) 0.094 0.066 (1.0989) (1.0686) 0.107 0.049 (1.1131) (1.0504) -3.290*** -2.345 (0.0372) (0.0959) -0.070 (0.9323) 0.033*** 0.029` (1.0337) (1.0294) -0.001 (0.9995) - DV= Prac_Social (Odds Ratio) Model 3 Model 4 0.042 0.045 (1.0430) (1.0462) 0.581* 0.490` (1.7884) (1.6327) 0.105 0.084 (1.1111) (1.0875) 0.305** 0.319** (1.3568) (1.3762) -0.527 -2.131* (1.6938) (0.1187) 0.761** (0.7937) - -0.231 (0.7937) DIS_Ins * Host_HDI -0.940** (0.3905) N 811 811 811 811 Log-likelihood -499.75 -498.32 -545.53 -541.78 (df=6) (df=8) (df=6) (df=8) Model fit Wald χ2 = 45.8 Wald χ2 =48.7 Wald χ2 = 30.5 Wald χ2 =38.0 (p < 0.001) (p <0.001) (p <0.001) (p <0.001) Hosmer-Lemeshow 13.481 8.501 4.143 4.146 C statistic# p-value p-value p-value p-value =0.096 =0.3861 =0.844 =0.844 Note: `, *, **, and *** indicates significance at the 0.1, 0.05, 0.01 and 0.001 levels, respectively. # 1. df refers to the degree of freedom; 2. A Hosmer-Lemeshow (H-L) statistic with a p-value greater than 0.05 is considered a good fit 28 -3.337*** (0.0356) - - Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 Figure captions: Fig. 1 The conceptual framework Fig. 2 A latent semantic analysis (LSA)-assisted content analysis for extracting environmental and social practices data 29 Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 Institutional distance Host country context (social) Host country context (environmental) H1 Environmental practices H2 H4 Corporate social responsibility (CSR) Other CSR practices Fig.1. The conceptual framework 30 H3 Social practices Journal of Construction Engineering and Management (ASCE) Published: Nov. 10, 2021; DOI: 10.1061/(ASCE)CO.1943-7862.0002226 Identifying topics for extracting environmental and social practices Manual decoding 8 ICCs with 50 CSR reports Nvivo Pro 11 Text mining (Piloting) 8 ICCs with 50 CSR reports Latent semantic analysis (LSA) gensim in Python 3.6.5 Results validation and feature selection Evaluation measures (F1 score); Cross-validation gensim in Python 3.6.5 Text mining (Scaling-up) 68 ICCs with 369 CSR reports Latent semantic analysis (LSA) gensim in Python 3.6.5 Extracted data of environmental and social practices Fig.2. A latent semantic analysis (LSA)–assisted content analysis for extracting environmental and social practices data 31