Influence of organization environment on contractor’s inter-regional market entry practice in China: an industry-level approach based on Bayesian logistic regression model

Abstract

Inter-regional market entry poses a complex challenge for contractors, particularly in the construction industry with its unique characteristics. Existing research predominantly focuses on an international scale and target markets, often overlooking the influence of the organizational environment, especially the home market. This study aims to fill this gap by analyzing how the construction organizational environment affects contractors’ inter-regional market practices. Grounded in theories of organizational management, we developed a theoretical model with seven hypotheses examining the impact of organizational environment. We tested hypotheses with Bayesian logistic regression model based on data from 426 construction organizations in China. Our findings reveal that a contractor’s size significantly impacts the practice compared to other attributes like ownership and experience. Additionally, a thriving local economy strongly supports contractors in exploring other geographical markets. Interestingly, while local competition intensity and protectionism create barriers for external firms, they do not significantly affect the diversification activities of local contractors. These insights contribute to understanding the dynamics of inter-regional market entry in the construction industry and can guide practitioners in strategizing for sustainable growth. Future research could explore more detailed strategies such as market selection and entry mode, enhancing the depth of understanding in this area.

Keywords:

• Organizational environment
• Inter-regional strategy
• Bayesian logistic regression
• Construction market
• Market Diversification

Introduction

The inter-regional market entry strategy in the construction industry is critical for contractors seeking to expand their geographic market and is important for the growth and sustainability of their businesses. However, products in the construction industry are featured as long-lasting outputs and irreproducible units, which can cause a huge economic loss when inappropriate geodiversity strategies are made. Therefore, the strategy on whether to enter the market in a new place or not must be a careful trade-off between benefits and costs shaped by organizational environments.

Theories in organization management define the organizational environment includes two aspects: the external environment produces the images of opportunity and threat that will be faced by an organization and the internal environment creates the images of strength and weakness owned by an organization (Rizal et al. 2017). Compared to other industries, the construction industry is not only labour and capital-intensive (Shiferaw 2017) but has a temporary nature of project-oriented construction teams (Addyman and Smyth 2022). While the temporary nature impairs the ability of construction firm organizations to create cumulative organization knowledge which is common practice in many other sectors of the economy (Eltigani et al. 2020), the intensive industry environment also put challenges for contractors’ competitiveness, such as organization capability and available resources, to enter new geographical markets (Wang et al. 2020). These studies indicate that when making strategies, organization managers need to take comprehensive consideration of both the internal and external organizational environment.

Researchers in the construction industry have extensively explored the dynamics of contractors’ inter-regional operations at an international level. The discourse encompasses a range of decision-making aspects inherent in the process of inter-regional market entry. These aspects typically follow a logical development sequence, including (1) the initial go/no-go decision (Utama et al. 2019); (2) market selection (Chen et al. 2016, Preece et al. 2016, Wang et al. 2020); and (3) choice of entry mode (Jia et al. 2016, Sullivan et al. 2019, Kumar Viswanathan and Neeraj Jha 2020). However, such a focus predominantly at the international level tends to treat each country as a homogenous environment, overlooking the diverse and often complex sub-regional economic landscapes. This might lead to the oversimplification of market dynamics, potentially leading to strategic missteps in countries with diverse economic, cultural and regulatory environments such as China. Besides, most research pays more attention to target regions like the host country, while less attention has been paid to the local environment. However, the environment where an organization is located, generally identified as the ‘home-market effect’ (Corsetti et al. 2007), also influences the organization’s strategies.

Therefore, this study aims to provide a more granular perspective on inter-regional market entry strategies by examining organizational environmental factors at the sub-national level and extending the focus to include the “home-market effect.” To achieve this goal, we analysed the inter-regional market entry practices of contractors in China. This is because China’s sub-national regions present diverse economic and regulatory environments, and its growing construction market offers numerous opportunities for contractors to geographically expand their businesses into these varied sub-national regions. By integrating insights from organizational management research and specific characteristics of the Chinese construction market, the study formulates hypotheses about how organizational environmental factors influence contractors’ choices between local and inter-regional markets. To validate these hypotheses, the study employs Bayesian logistic regression analysis, utilizing industry data to provide empirical evidence. The outcome is expected to offer a better understanding of how construction organization environments influence contractors’ inter-regional market entry practice.

This article is organized as follows: following this introduction, we delve into a comprehensive literature review on the existing body of research on inter-regional construction market entry practice. The subsequent section outlines the theoretical background and formulates our hypotheses, setting the stage for our investigative approach. The research method section describes the Bayesian logistic regression model we use and elucidates the data collection and sampling process. Then the result and discussion section presents the model result and discussion in light of previous research and the contextual setting of the study. Finally, the conclusion section summarizes the main contributions of the study, discusses its implications for current inter-regional construction market entry practice and suggests potential direction for future research.

Literature review

Regions for different geographic market divisions are very complex social entities, encompassing multiple types of social and business activity and existing within several types of geographic locations, such as cities, counties, states and nations (Romanelli and Khessina 2005). Market entry theories such as Porter’s Diamond Model, the Eclectic Paradigm and the Transaction Cost Analysis offer a foundational understanding of how firms assess and engage with new geographical markets (Holtbrügge and Friedmann 2016, Rahman et al. 2018). These frameworks emphasize factors in both firm and market environments such as firm resources, ownership, market demand and competition advantages. In considering the market-level environment, the target market plays a pivotal role, providing the context within which firms must adapt and thrive. However, the interaction between a firm’s home market environment and its strategies for entering and navigating new regional markets is equally crucial. This interplay shapes not only how a firm approaches market entry but also influences its operational tactics and long-term sustainability in diverse regional markets (Cuervo-Cazurra et al. 2018). Recent studies have also revealed the impact of home market characteristics on a firm’s innovations and foreign expansion (Hoskisson et al. 2013) and how the “distance” between the home and target market affects the geographical expansion of firms (Luo and Shenkar 2011, Ang et al. 2015)

For the construction industry, there has been a growing body of research focusing on the practices of international market entry. The studies identified environmental factors that influence the market entry strategy from both the internal and external organization environment. The internal organization environment influences how organizations evaluate opportunities and risks and whether the organization can handle the business in the target market (Odediran and Windapo 2017). The detailed factors generally involve the organization’s features such as previous experience, company size and ownership structure. For example, through a survey of 45 construction firms, Preece et al. (2016) claim that the firm’s ability to assess market signals and opportunities, the firm’s resources, and the firm’s experience in similar works will influence their strategy development on entry location, entry timing and entry mode.

For the external market environment, research on inter-regional construction market entry often focuses on the target market, commonly referred to as the host country in this context. The environment of the host country influences the strategy of the whole market entry process such as the initial go/no-go decision (Utama et al. 2019); market selection (Kumar Viswanathan and Neeraj Jha 2020) and enter mode (Jia et al. 2016). For example, Kumar Viswanathan and Neeraj Jha (2020) studied Indian construction firms and revealed that host market potential and risk level greatly influence their preference when making international market selection decisions. In addition to the strategy employed, the target market also influences the subsequent performance of construction firms, particularly after they have successfully navigated the initial entry phase. For instance, Jang et al. (2019) conducted a study examining the impact of diversification – across regional, product and industry dimensions – on the performance of construction firms within international markets. This research highlights the significance of strategic market choices in determining the success and sustainability of construction firms post-market entry.

Despite the rich body of research on market entry strategies in the construction industry, two significant gaps are evident. Firstly, existing studies predominantly focus on international market entry, treating each country as a monolithic market. This approach overlooks the substantial regional variations within countries, which can be critical in shaping market entry strategies and outcomes. The nuanced complexities of inter-regional markets, especially in countries with diverse economic, cultural and regulatory landscapes, remain underexplored. Secondly, the literature heavily concentrates on the dynamics of the target or host markets, with limited attention to the influence of home market conditions. However, unlike products that can be manufactured in one location and shipped globally, construction services require a physical presence at specific sites, making each project highly reliant on and unique to its location. This characteristic of the construction industry binds firms more intimately to their local environment. Consequently, it demands a nuanced approach to geographic market entry strategies, especially in the initial phase of deciding whether to undertake a project outside the local region or not.

Therefore, this study aims to fill the gaps in construction industry research by analysing how organizational environment factors, including internal firm characteristics and the home market environment, influence contractors’ market entry practices within China’s subnational regions. This research will contribute to a better understanding of market entry strategies in the construction industry, particularly in the context of countries with significant internal regional diversity.

Theoretical model development

For the definition of construction organizations’ inter-regional activity, researchers may either choose the headquarter location or registered place as the base (Ren et al. 2015). As contractors may register different subsidiaries under local regulations when starting projects in different provinces, the headquarters location is generally regarded as the original region. When the project location is different from the headquarter location, we define it that the firm took an inter-regional market practice. In this research, factors on go/not go practice can be divided into internal and external organization’s environment in the local market. For the internal organization environment, identified factors are firm size, ownership structure, organization history, operation experience, number of employees, etc. For the external organization environment, identified factors are local economic, industry scale, local protection, market risk, competitive intensity and market potential. While the measurement of some factors cannot be quantified or accessed through industry open accessible data, as Figure 1 shows, the remaining factors for theoretical model development in this study are (1) organization size, (2) ownership structure, (3) organization experience, (4) market potential, 5) industry size, (6) local protection and (7) Local competition. The following subsections will discuss how these factors influence inter-regional practice and propose related hypotheses.

Figure 1. Theoretical model of contractor’s inter-regional market entry practice.

Internal organization environment

Organization size

As resources-based theory states, the ability and resources available to the organizations are important when they are engaged in inter-regional expansion and competing in new markets (Wernerfelt 1984). Compared with the manufacturing or service industry, the construction market is often more capital-intensive. During the long-lasted construction period, payments are usually given to the contractor with the actual completion state. Contractors often need to pay the bill with their own money in advance, especially at the beginning of a project when there is no physical progress but with a lot of preparation work. While the capital return often takes a long time, there is an asset threshold for contractors to undertake some projects. Though some scholars still hold the view that due to the flexibility and specialization, small organizations may adapt better to the complexity of the environment in a new inter-regional market (Johanson et al. 2015), there is empirical evidence proves that local small and medium contractors are usually not well prepared in the new market expanding despite strong encouragement and support from the government (Teo et al. 2007). Also, studies focusing on market selection (Chen et al. 2016) and entry mode (Odediran and Windapo 2017) indicate that the organizational size of contractors will influence their market entry strategy.

It is therefore hypothesized that:

Hypothesis 1 (H1): With other variables held constant, large construction organizations are more likely to perform inter-regional expansion.

Ownership structure

In general terms, ownership structure refers to the composition of a firm’s ownership, which can include individual owners, groups of people, corporations, charitable foundations, or governmental entities. Compared with organizations owned by a private entity, in the construction industry, firms with government backup will have more networks and capacity to deal with risks involved in inter-regional market entry. The impact of ownership structure on the organization’s inter-regional market entry strategy has been acknowledged in organization research (Li et al. 2009). In the construction industry, the empirical study of Wang et al. (2020) shows that among contractors entering the inter-regional market, state-owned ones account for a relatively high proportion. To maintain sustained growth and profitability, it is key to continue to obtain new projects. While the supply of land resources in different regions is generally not balanced, construction organizations will inevitably face the choice of inter-regional management strategies after developing to a certain scale. For state-owned organizations with abundant local resources, the willingness to expand to new regional markets may not be high. However, unlike the continuous and stable general commodity consumption market, the limited land resources and spatial limitations of the population carrying capacity make the construction market in most regions easily reach saturation. For sustainable development, state-owned construction organizations also need to explore new markets to sustainably survive and increase long-term profits (Choi and Russell 2004). State-owned organizations are relatively easier to establish relationships with target market institutions, which can help them obtain more information and speed up the adaptation process when entering a new market.

It is therefore hypothesized that:

Hypothesis 2 (H2): With other variables held constant, state-owned construction organizations are more likely to perform inter-regional expansion.

Operation experience

When entering a new regional market, previous experiences can accelerate the process of gaining a competitive advantage. Because experienced companies can more easily predict the potential opportunities and possible challenges in the target market, they can invest more efficiently in enterprise resources. Study on market entry strategy influences shows that experience plays an important role in their entry time, mode and market selection (Preece et al. 2016, Odediran and Windapo 2017). Also, for the basic entry practice, studies on market entry strategy show that experienced contractors are more likely to participate in the international market (Viswanathan and Jha 2019). However, pure experience is hard to quantify. We assume that the longer an organization exists, the more experience it tends to gain. Some scholars’ research has also proven that the length of the establishment can affect an organization’s developing strategy (Parnell et al. 2012).

It is therefore hypothesized that:

Hypothesis 3 (H3): With other variables held constant, long-established construction organizations with experience are more likely to perform inter-regional expansion.

Effect of the local market

Economic development

In the process of China’s economic development, the construction industry is a pillar industry that promotes the development of the country’s economy. The government has repeatedly used increasing investment in infrastructure construction as a means to expand domestic demand, making the construction industry an important driving force for the growth of per capita GDP and one of the important sources of increasing social wealth (Jiang 2007). In turn, contractors in the developed region are more backed in terms of resources to explore new regional markets. Besides, regional economic development often correlates positively with urbanization levels (Solarin 2017). When urbanization reaches a certain scale in developed regions, construction is more likely to saturate. So the government tends to pull more developments toward areas with undeveloped land.

It is therefore hypothesized that:

Hypothesis 4 (H4): With other variables held constant, construction organizations located in the more developed region are more likely to perform inter-regional expansion.

Industry scale

The local market’s influences on an organization’s entry strategy have been studied in the international market (Root 1994). If the domestic market attractiveness is enough, which means a large market with stable demand, many organizations will focus on the domestic market and become less interested in geographic expansion (Chen and Messner 2011). In contrast, for contractors who have suffered from dwindling local markets, it is important to explore the new regional market to maintain. While industry scale can be an important index reflecting local market attractiveness.

It is therefore hypothesized that:

Hypothesis 5 (H5): With other variables held constant, construction organizations located in regions with bigger industry scales are less likely to perform inter-regional expansion.

Local protection

Local protection is a variety of behaviours in which local governments exercise their power to protect the interests of economic entities within their jurisdiction (Zhang and Wu 2006). For example, governments may stipulate that local units can only receive services from local organizations and limit foreign goods or services from entering the local market through some rigorous regulations. It constitutes a kind of entry barrier for contractors outside the region, which led to the decentralization of domestic, making it difficult to give play to the comparative advantages of local markets in economic operation. The form of barrier is generally to prevent the entry of new rivals and increase the profitability of incumbent firms (Islami et al. 2019) and often influences new arrivals’ entering strategy (Kumar Viswanathan and Neeraj Jha 2020). Wang’s research on construction industry barriers also shows that local protection is one of the main factors that block contractors’ regional market entry (Wang et al. 2020). In general, local protectionism restricts the free flow of resources and goods between regions. Organizations located in high-protection regions often face less competition from organizations outside the region and enjoy greater market size compared with those located in more free-trade regions, which may lower their passion to make business expansion.

It is therefore hypothesized that:

Hypothesis 6 (H6): With other variables held constant, construction organizations located in regions with higher local protection are less likely to perform inter-regional expansion.

Local competition

Limited land provision and high competition in the very region force firms to take inter-regional expansion to increase the firm’s growth in size and profitability to survive in local construction markets. Generally, doing business in highly competitive regions is less profitable, when the competition reaches a degree and seeking inter-regional opportunities costs less for survivors of contractors. Musso and Francioni (2012) found that the competitive intensity of candidate markets is considered when using shortlisting selections. However, in Chen et al.’s (2016) research on factors for market selection, results show that contractors tend to enter highly competitive countries, which seems to be contradicted by the traditional concept that the higher the competition tensity of a market, the lower the probability that contractors. In this study, we would like to verify whether the local competitive intensity will influence the practice of contractors to expand their business in other markets.

It is therefore hypothesized that:

Hypothesis 7 (H7): With other variables held constant, construction organizations located in regions with higher local competition are more likely to perform inter-regional expansion.

Research method

To test the proposed hypotheses regarding the factors that influence inter-regional construction market entry practice, this study employs a Bayesian logistic regression model. This approach is chosen for its adeptness in managing uncertainties inherent in hypothesis testing, an aspect where it holds an advantage over the likelihood-based frequentist methods typically used in standard logistic regression (Hilbe 2011). This is especially important given that factors such as the traits of decision-makers, which are not directly measured in our study, can also influence construction market entry practice. In addition, the Bayesian logistic regression model offers a distinct advantage in analysing the identified influencing factors. Rather than providing a singular, determined parameter estimate, it yields a range and distribution for each factor, providing all possible outcomes. This feature is particularly valuable in the complex context of deciding whether to take the inter-regional market entry practice for construction organizations. To ensure our findings align with the broader body of empirical research, this study also incorporated a standard logistic regression model to complement the Bayesian analysis with a frequentist perspective. This established method facilitates a meaningful comparison and interpretation of our conclusions within the broader research community of prior empirical studies.

In the Logistic regression model used in this paper, whether a construction enterprise has inter-regional market entry behaviour is the dependent variable. When this behaviour occurs, the dependent variable Y is valued as 1 with probability P; otherwise, there is (1 − P) probability for Y valued as 0. To solve the S-shaped function evident in the original model, we define the probability of Y = 1 is defined as p/(1 − p) (p < 1), which also known as odds. The natural logarithm of the probability log p/(1 − p) is called the logarithm of Y, denoted as logit (Y). Therefore, the relationship between the dependent variable Y and the independent variable X₁, X₂, X₃…in this logistic regression can be expressed as: (1) $$\text{Logit}\left(\mathrm{Y}\right){\text{\hspace{0.17em}=\hspace{0.17em}}\mathrm{\beta }}_0{\text{\hspace{0.17em}+\hspace{0.17em}}\mathrm{\beta }}_1{\mathrm{X}}_1{\text{\hspace{0.17em}+\hspace{0.17em}}\mathrm{\beta }}_2{\mathrm{X}}_2{\text{\hspace{0.17em}+\hspace{0.17em}L \hspace{0.17em}+\hspace{0.17em}}\mathrm{\beta }}_{\mathrm{k}}{\mathrm{X}}_{\mathrm{k}}$$(1)

Where, β_0, β_1, β_2。。。β_k are parameters in the model, therefore, we have (2) $$\mathrm{P}=\frac{e^{{\beta }_0\hspace{0.17em}+\hspace{0.17em}{\beta }_1{\mathrm{X}}_1\hspace{0.17em}+\hspace{0.17em}{\beta }_2{\mathrm{X}}_2\hspace{0.17em}+\hspace{0.17em}\dots \hspace{0.17em}+\hspace{0.17em}{\beta }_k{\mathrm{X}}_k}}{1+e^{{\beta }_0\hspace{0.17em}+\hspace{0.17em}{\beta }_1{\mathrm{X}}_1\hspace{0.17em}+\hspace{0.17em}{\beta }_2{\mathrm{X}}_2\hspace{0.17em}+\hspace{0.17em}\dots \hspace{0.17em}+\hspace{0.17em}{\beta }_k{\mathrm{X}}_k}}$$(2)

Thus, if the chosen function is correct, a plot of log p/(1 − p) against x will produce a straight line, which is easier to estimate. The hypothesis test of relevant parameters in logistic regression analysis is completed using the Wald test, in which the parameter value estimated by the original equation is substituted into the constraint condition to check whether the constraint condition is true.

To build the Bayesian logistic regression, the likelihood function is the same as pure logistic regression. We’ll also assume independence among the priors and express our prior understanding of the model baseline through the centred intercept: (3) $$\begin{array}{c}\begin{array}{c}\mathrm{\text{Data}}:Y_i\left|{\beta }_{0,}{\beta }_{i\mathrm{ }}\text{with Logit}\right(\frac{p}{1-p})={\mathrm{\beta }}_0+{\beta }_i\mathrm{ }{x}_i(0<p<1)\\ \mathrm{\text{Priors}}:{\beta }_0\sim N\left(0,1\right)\end{array}\\ {\beta }_i\sim N(0,1)\end{array}$$(3)

The thing we add is prior for the intercept and coefficient of variables. Since it can be anything in real data, consider our prior tunings, for a given contractor without considering its organizational environment, there is a roughly 50% chance to perform geodiversity strategies, so Thus, we set the prior mean for β₀ on the log(odds) scale to 0: (4) $$\mathrm{\text{log}}\hspace{0.17em}\left(\frac{p}{1-p}\right)=\mathrm{\text{log}}\hspace{0.17em}\left(\frac{0.5}{1-0.5}\right)=0$$(4) so we assign priors a standard normal distribution with a function as (0,1). As for the error, make it an exponential distribution. To validate the model, we apply the general regression model validation methods by using random 80% data to do the training, and the remaining untrained data to do the validation work. For the training chains set, we let the chains as 4, iteration as 5000*2 and seed as 84,735.

Variables for statistical analysis

Variables in the hypothesis can be generally measured by multiple methods. The principle we choose for the measurement way lies in two aspects. One is that the variable measurement should be widely used in current peer-reviewed journals. The other consideration is whether the involved measurement can be supported by open data since one of the contributions of this study is to use openly accessible data to verify the hypothesis. For variables related to the organization’s internal organization environment, the organization scale could be measured by the number of employees, paid-in capital, or total assets, etc. This paper uses the number of employees to quantify organization scale, for which the standard has been regulated as those organizations with more than 3000 employees are large, medium-sized organizations with 600–3000 employees and small ones with less than 600 employees (Shigang and David 2011). For operation experience, this paper takes the assumption that the longer the contractors exist, the more experience they will have in undertaking construction projects. Therefore, the current year is used as the benchmark to subtract the time when the organization was established. Generally, the earlier the organization is established, the more its experience is.

The ownership structure is the reflection of an organization’s property rights, which is, therefore, generally divided into state-owned and private based on whether the organization’s property rights belong to the state. According to the definition of the National Bureau of Statistics, state-owned organizations in a broad sense include purely state-owned, state-controlled and state-owned joint-stock. In a narrow sense, the state-owned only refers to organizations whose national equity ratio exceeds 50%. Existing studies mostly use broad definitions, which treat pure state-owned, state-holding and state-owned joint-stock all as state-owned organizations (Ren et al. 2015). Referring to the information in the National Enterprise Credit Information Publicity System in China, this paper also categorizes an organization’s ownership structure in this way.

In terms of variables related to the local market, for economic development measurement, per capita GDP has been used most frequently and has also been regarded as an indicator of the market potential in some research (Chen et al. 2016). It is an effective tool for governments and people to understand and manage the macroeconomic operation of regions. This paper thus selects per capita GDP over 2015–2019 as an indicator to measure regional economic development. As shown in Table 1 per capita GDP is obtained by dividing the gross domestic product realized by a country during the accounting period (usually one year) by the country’s permanent population (currently the registered population). In the existing literature on the construction industry, the measurement indicators of industry scale mainly include the gross output value, the added value and the completed area of buildings. Among them, the gross output value of the construction industry is one of the most common indexes in industry-related analysis and is especially important for the overall arrangement of the national economy (Papadakis and Barwise 1997). Therefore, the paper also chooses it as the index of the construction industry scale. Generally, the competition intensity could be measured by the competitor number (Ye et al. 2014). Industry organization economics also assumes that competitive intensity increases when the number of competitors increases in the industry, which puts downward pressure on market share and marginal profit rates. Also, considering that the same number of competitors will bring different competitive pressures under different market sizes, we use the ratio of contractor numbers to market size to represent the competitive intensity of the local market.

Table 1. Definition and measurement of variables.

Variables	Definition	Measurement	Data source
Organization scale	The degree to which factors of production and products are concentrated in an organization	The number of employees	Organization’s official website
Operation experience	The length of time an organization has existed	2020–established year
Ownership structure	Reflection of organization’s property rights	Whether the organization’s property rights belong to the state	National Enterprise Credit Information Publicity System in China
Economic development	The degree of economic prosperity in a specific region	Per capita GDP	National Bureau of Statistics of China (NBS) 2019
Industry scale	The output scale or operational scale of the industry.	The gross output value
Local protection	The tendency of local authorities to negatively exercise their power to safeguard local interests	The Local Protection Index (LP).
Local competition	The competition tensity in the local market	The number of contractors

Table 2. Demographic information of examined organizations.

Variable	Category	Number	Percentage
Inter-regional organizations	Yes	114	26.76
	No	312	73.24
Organization size	Below 600	88	20.66
	600–3000	201	47.18
	Above 3000	137	32.16
Organization history	Below 20	36	8.45
	20–50	232	54.46
	Above 50	158	37.09
Ownership structure	State-owned	232	54.46
	Not state-owned	194	45.54

As for the measurement of the degree of local protection, while there are no direct indicators in existing research, there are indirect indicators to quantify the local protection level through the Local Protection Index (LP) (Sun et al. 2014), namely: LP _at =1/(X _at*Y _at), in which, (5) $$X_{\mathit{\text{at}}}=\left|\frac{{\mathit{\text{GDP}}}_{2t}}{{\mathit{\text{GDP}}}_t}-\frac{{\mathit{\text{GDP}}}_{2\mathit{\text{at}}}}{{\mathit{\text{GDP}}}_{\mathit{\text{at}}}}\right|Y_{\mathit{\text{at}}}=\left|\frac{{\mathit{\text{GDP}}}_{2\mathit{\text{at}}}}{{\mathit{\text{GDP}}}_{\mathit{\text{at}}}}/\frac{{\mathit{\text{GDP}}}_{2t}}{{\mathit{\text{GDP}}}_t}\right|$$(5)

In this function, subscript 2 represents the secondary industry, and t and a represent the specific time and region, respectively. Among them, X_at is the Krugman Industrial Structure Convergence Index, which is used to measure the degree of industrial structure differences between regions. The larger this value is, the greater the degree of differentiation between local industries and other regions will be, thus the threshold for foreign companies to enter the local market could be higher, leading to lower motivation for local governments to implement local protection. Y_at is the Hoover Localization Coefficient, which is commonly used to describe the industrial concentration of a region. The higher this value, the more concentrated the industry is in a geographic location, which means the higher professionalization level of the region’s industry. Thus, the necessity and enthusiasm of local governments to implement protection could be lower. Overall, the greater the LP in this function, the greater the degree of local protection.

Data collection and sampling

To analyse how the organizational environment affects contractors’ inter-regional market entry practice, we collected the organizational samples through the longitudinal dataset of the National Quality Award Projects (NQAP) in the Chinese construction industry during the period of 2000–2019. China Construction Industry Association (CCIA) established the National Quality Award in 1987 under the guidance of the Ministry of Housing and Urban-Rural Development (MOHURD). Through 30 years of development, it becomes the most influential and prestigious award in the Chinese construction industry, and the award-winning organizations are widely regarded as the most competitive construction firms in the industry (Wang et al. 2020). The award has been issued every two years since 2010 and covers all types of construction projects throughout the construction industry. Organization data of all the award-winning projects are publicly available on the official website of the CCIA (http://www.zgjzy.org/). Due to the availability of its project data and the strict official appraisal process, the National Quality Award provides a relatively attractive research setting to comprehensively analyse Chinese contractors’ inter-regional market entry practice. The paper focuses on construction organizations undertaking inter-regional undertaking inter-regional projects as independent contractors. Since contractors’ competitiveness may be affected by their project-based collaborative relationships (Cao et al. 2017), construction projects involving multiple general contractors or informal contracting organizations, such as command and military units, are excluded. There are 426 contractors left in the sample. In the meantime, we collect regional market samples according to political boundaries in mainland China, which leads to 31 such divisions, classified as 22 provinces, 4 municipalities and 5 autonomous regions. The following section will further clarify how we get more detailed information about the organization and regional market research samples.

Organization-specific data

Among 426 contractor organization samples, 114 contractors had inter-regional market entry practices, the other 312 contractors did not perform this strategy. We obtained detailed organization information such as organization size, location and establishment year mainly through the organization’s official website and got the information on the organization’s ownership structure from the National Enterprise Credit Information Publicity System in China. Information publicized on this system comes from relevant government departments that are responsible for the authenticity of their public information. Due to the lack of detailed information on the process of screening and collecting the information above, another 10 organizations have been filtered out again. The demographic characteristics of contractor samples are shown in Table 2.

Table 3. Logistic regression model results.

Independent variables	Model 1			Model 2
	Coefficient	Conf. low	Conf. high	Coefficient	Z value	Sig
Ownership structure	0.499	−0.041	1.058	0.422	1.210	0.226
Operation experience	−0.05	−1.985	1.892	−79.36	−2.261	0.023
Organization size	1.514	0.978	2.099	1.631	4.851	0.010
Local economic development	1.643	0.326	2.991	2.910	2.577	0.000
Local construction market	0.555	−0.335	1.476	−0.226	−0.392	0.695
Local protection	−0.095	−0.697	0.498	0.078	0.218	0.827
Local competition	−0.380	−2.259	1.526	−1.228	−0.265	0.791

Region-specific data

Data related to regions are taken from the panel data published on the website of the National Bureau of Statistics of China (NBS), supplemented by the statistic yearbooks of the provinces (municipalities and autonomous regions). To reduce the effects of time-related accidental influence, the study takes the natural algorithm transformed average statistic data from 2017 to 2019 in NBS. Figure 2 shows the distribution of regional economic development, construction market scale and local protection level. In general, the economic development and construction market scale are both higher and larger on the eastern seaboard and Bohai Rim is relatively higher than other regions in China, which follows the fact that China’s economic development and construction spending have historically been concentrated along the eastern seaboard. As for the local protection level, Figure 2 shows a different trend than that of economic development. In general, the west and north parts of China have higher local protection than the east and south parts. This regional data to some degree indicates China’s geographical imbalances in various regions. It seems an open and diversified economic development environment may provide organizations with stronger backups to support their expansion activities. The following part will analyse how this environment affects the contractor’s geodiversity strategy through Bayesian logistic regression analysis.

Figure 2. Distribution of regional economic development, construction market scale and local protection in China.

Analysis results

Before analysing the results, we run the Bayesian logistic regression model on the training and test dataset with diagnostic functions, which include trace, dens-overlay, mcmc_acf and pp_check (Gabry et al. 2019). Trace plots provide an important tool for assessing the mixing of a chain. In this case, as Figure 3 shows, the trace plots show no obvious flat bits or too many consecutive steps in one direction, which indicates the MCMC sampler mixes well. Dens-overlay shows the distributions of many replicated datasets drawn from the posterior predictive distribution compared to the empirical distribution of the observed outcome. Overlapped chains in the density figure indicate that chains stated from different starting points can converge to the same density plots, which indicates the model is represented well in the dataset. Mcmc_acf () is a grid of autocorrelation plots by chain and parameter. The lags argument gives the maximum number of lags at which to calculate the autocorrelation function. Autocorrelation function (ACF) plots measure how the sample autocorrelation between the terms of the chain decreases as a function of their lag. Plots, in this case, show that autocorrelation is large at short lags, but then goes to zero pretty quickly, indicating that the model is quite good with no obvious autocorrelation problem. The final pp_check is the plot of posterior predictive checks, which simulates replicated data under the fitted model and then compares these to the observed data. The plot shows how prediction results are distributed around the test data. They are basically in the same trace in this case. Accordingly, the accuracy returned by the model on test data is 85.88%.

Figure 3. Diagnose result of Bayesian regression model-(a) the trace, (b) dens-overlay, (c) mcmc_acf (d) and pp_check.

To test the hypothesis, we also ran data on a normal binary regression model (model 2). Table 3 shows the results of both models. For the Bayesian logistic regression model, a 95% confidence interval indicates whether the influence of a specific parameter is significant. If the 95% confidence interval includes the null value, then there is no statistically meaningful or statistically significant effect. In this case, only organization size ([0.978, 2.099]) and local economic development ([0.326, 2.991]) show significant influence. Though the coefficient shows that local protection and local competition will have a negative influence on contractors’ seeking opportunities outside, the influence is not statistically significant. For the frequentist logistic regression model, we used Wald tests to evaluate the significance of coefficients Z value. Generally, A higher absolute value of the Z-value and a significant p value (less than .05 or .1) show that the independent variable plays a certain explanatory role in the model. Model 2 results in Table 3 also show organization size (β = 1.630, z = 4.851, p < .01) and local economic development (β = 2.910, z = 2.577, p < .01) are significant, which is in accordance with the result of model 1. However, Model 2 shows the influence of Operation Experience is significant (β= −79.36, z= −2.261, p < .05), while the Bayesian model does not indicate a significant influence, with the interval ranging from −1.985 to 1.892 and encompassing zero.

Therefore, for the internal organizational independent variables, we accept the hypotheses: HI large organizations are more likely to perform inter-regional expansion and reject the hypotheses: H2 state-owned organizations are more likely to perform inter-regional expansion. As for hypothesis H3, there is a discrepancy between the results of the Bayesian and frequentist regression models regarding the impact of operational experience. While the frequentist model indicates significant influence, the Bayesian model suggests otherwise, with its credible interval encompassing zero. Given these conflicting outcomes, we refrain from making a definitive conclusion about accepting or rejecting H3 at this stage. Instead, we acknowledge the complexity and uncertainty inherent in the analysis and further discuss the implications of these differing outcomes in the discussion section.

For the external environment independent variables, we accept the hypothesis: H4 organizations located in the more developed region are more likely to perform inter-regional expansion. And rejected the other three hypotheses: H5 organizations located in regions with bigger industry scales are less likely to perform inter-regional expansion. H6 organizations located in regions with higher local protection are less likely to perform inter-regional expansion. H7 organizations located in regions with higher local competition are more likely to perform inter-regional expansion.

Discussion

The accepted hypothesis on contractors’ organizational size in the modelling results follows the existing organizational research. In the theory of organization management, when an organization has more resources than it needs for normal production and operation, it obtains abundant resources. This increases the organization’s tolerance for potential failures, so they are more inclined to try new business strategies. Existing research shows that organization scale can represent the capacity and resource abundance inside the organization (Denicolai et al. 2021). The larger the organization’s scale is, the higher the organization’s financial capacity will be, and decision-makers are more likely to make expansion decisions. On the contrary, small-scale contractors are relatively weaker in risk defence even when they successfully enter the inter-regional construction market, so they have relatively weaker motivation and ability to perform the expansion strategy. The findings also agree with earlier studies reporting that firm size is one of the critical moderators of risk perception for new market entry strategies (Odediran and Windapo 2017). However, as the most basic unit of economic development, small and medium-sized organizations are an indispensable part of the market system, especially in the construction industry, which is characterized by fragmentation. To reduce the restriction of the external environment on its long-time development and broaden the strategic decision choice, these contractors must increase the organization’s abundant resources and strive to make it stronger. Now many studies are focusing on the performance and sustainability of small and medium enterprises (Wentzel et al. 2022), geodiversity strategy of the organization in construction has still been paid less attention to.

Our findings also shed light on the role of ownership structure in the construction firm’s inter-regional market strategy. While the literature shows that private firms face more challenges and risks when entering a new market (Zhang et al. 2022), the result shows that whether the firm is state-owned or private does not influence whether it will perform entry practice to new geographical markets. Through digging more into studies on firm ownership structure, the literature also points out that not only state-owned firms but also privately owned firms that receive policy support from governments are in strong positions to expand (Cravo and Piza 2019). This underlines the importance of governmental policy support in facilitating firms’ growth and geographical expansion, which adds a new perspective to current understanding.

For Operational Experiences, the divergence results from the foundational differences between the two models. The Bayesian approach incorporates a broader scope of uncertainty, which leads to a more conservative estimate, while the frequentist approach is based on observed data frequencies. Additionally, the difference might reflect practical significance, where the Bayesian model’s output offers a probability distribution that accounts for real-world complexities. While there is limited study specifically on sub-national construction markets, existing research in international market contexts indicates that experience is a crucial factor in contractors’ strategies for entering new markets (Isa et al. 2017, Viswanathan and Jha 2019). This observation invites further discussion on the differences between national and international market diversification strategies. A potential explanation for the observed discrepancy is that more experienced construction firms have already established a strong presence in larger markets, leading them to explore international rather than domestic opportunities. This hypothesis suggests a strategic shift in focus for experienced firms, prioritizing international expansion over national market exploration. Therefore, future research could benefit from a deeper investigation into these discrepancies. It would be valuable to examine how contractors’ experiences influence their market entry strategies, particularly in contrasting domestic and international contexts. Such research could provide insights into the strategic decision-making processes of construction firms and contribute to a better understanding of market entry dynamics.

The results of the external environment show that a prosperous economic region provides a stronger backup for the organization’s head to support its sub-department performing inter-regional market activities. Also, the motivation comes from peer pressure for the limited local construction market in the well-developed region, which generally has more competitive contractors. This result is also supported by the international market entry practice in the construction industry. Compared with other industries, such as manufacturing that rely on consumption, less developed countries often indicate more project opportunities, while a high economic development level also indicates a high urbanization level. Thus, firms from the highly developed region will seek project opportunities in less developed as they still have more room to explore.

The interesting point of the regression results is that the degree of local protection and local competition intensity have no obvious influence on an organization’s outside market-seeking activity. In research on targeted regional market barriers (Wang et al. 2020), local market protection has a significant negative impact on inter-regional market entry practice. While this local protection can block outsiders, it seems to have little influence on local contractors performing expansion strategies. The market segmentation caused by local protection hinders contractors from both regions from developing, which is not conducive to the development of the local economy either. As for local competition intensity, while some studies point out that competition in the market and competitive advantage are part of the decision factors (Sullivan et al. 2019), Chen et al. (2016) show that contractors tend to enter highly competitive host countries. This result to some degree verified our study as a highly competitive host region may indicate more opportunities that attract outside contractors, and the indigenous contractors will also concentrate on the market.

By combining the influence of both internal organizational factors and external market factors, our research findings also underscore the capital-intensive nature of the construction industry, a characteristic that significantly influences a firm’s ability to penetrate new geographical markets. In construction projects, there is a need for a significant amount of both equity and debt capital to finance projects, especially in the initial stage of some project delivery, where effective use of capital can often be a critical determinant of a firm’s success (Williams 2016). Our findings reinforce the role of firm size, linked with the local market, shapes a firm’s expansion strategy in market expansion strategies. Larger firms, typically with stronger backups from the local market, are better positioned to accumulate the working capital necessary for expansion into new markets. Generally, a robust domestic market, supported by local protection measures, creates a conducive environment for larger firms to thrive and accumulate resources necessary for expansion (Sullivan et al. 2019). This insight complements existing literature on firm size and market penetration.

The specific context of this study makes the influence of regional economic disparities and regulatory frameworks more obvious in shaping contractors’ inter-regional market entry strategies. Compared to other countries, China’s sub-national regions feature a variation from highly urbanized and developed coastal areas to rapidly growing but less developed inland regions. This diversity not only offers market opportunities but also presents varied challenges for contractors. In more developed regions, firms may encounter saturated markets with intense competition, pushing them to seek opportunities in less developed areas where the market potential is untapped. Conversely, in these emerging markets, contractors might face challenges such as adapting to local business practices, navigating different regulations, dealing with local market protection and building relationships with local stakeholders.

However, it is important to acknowledge the unique context of China, which may limit the generalizability of our findings to other countries’ sub-national regions. The influence of government policies and incentives in China’s different subnational regions, for instance, plays a distinct role in shaping the construction market. In this study, we categorized these influences as “local protection,” a factor that might be more pronounced in the Chinese market compared to other countries. Though this variable does not show significant influence as part of the home market effect, the unique characteristic still adds layers of complexity to the construction market dynamics in China. Such context-specific factors need to be carefully considered when applying our findings to other settings.

Conclusions

This study analyses the influence of internal and external organizational factors on inter-regional market practices within the construction industry. By doing so, it bridges a significant gap in existing research, which predominantly focuses on international construction and target market environments. Based on organizational theory and the unique attributes of the construction market, we formulated seven hypotheses to investigate these dynamics. Utilizing Bayesian logistic regression for analysis, our findings highlight the critical role of firm size, the impact of a robust domestic market and the importance of policy support in shaping market entry strategies in the construction industry. Moreover, our results indicate that while local competition intensity and protectionism may put barriers to external entrants, these factors do not significantly impede the geographical diversification activities of local contractors.

The practical implications of these findings are significant for industry stakeholders, offering valuable guidance for construction firms looking to expand their market reach. For larger firms, the study underscores the advantage of leveraging their resource abundance in strategizing for inter-regional expansion. On the other hand, smaller firms should focus on consolidating their position in the local market to build a robust foundation for future expansion. These insights also suggest that collaborations and strategic alliances could be particularly beneficial for smaller firms for entering a new market, which can be further explored in terms of market entry mode strategy. For policymakers, these findings highlight the importance of creating a supportive environment that balances the need for local protection with the opportunities for market diversification. Policies that encourage fair competition and provide support to construction firms can foster a more dynamic and sustainable construction industry. Overall, these insights contribute to a more comprehensive understanding of the construction industry’s market dynamics, offering implications for both construction firms and policymakers in strategizing for sustainable growth.

While our study offers valuable insights, it is important to acknowledge its limitations. While the hypotheses regarding contractors’ internal organization features and external environment were grounded in organizational management theory, other potential factors such as market risks could also influence inter-regional market entry behaviours of construction companies. Bayesian logistic regression method employed in our analysis helps to mitigate some of the uncertainties associated with factors not directly included in the study, this still leaves room for more factors to be addressed in future research. Additionally, our research focused primarily on whether construction companies expand their business into new geographical markets. Other aspects of market strategy, including market selection, entry timing and entry modes that involve detailed characteristics of target markets, also present opportunities for further exploration.

Moreover, it is important to note that this study focuses specifically on the Chinese construction market, incorporating several context-specific variables that may not be entirely applicable to other countries’ contexts. This specificity might limit the direct applicability of our findings to different international settings. Therefore, future research can aim to explore and compare market entry strategies both within various subnational regions of China and in different countries. Such comparative studies could provide more comprehensive insights into how diverse organizational environments, each with its unique economic, cultural and regulatory context, shape the market entry decisions of construction firms. By examining these variations, we can deepen our understanding of inter-regional market entry dynamics in the construction industry on a global scale.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Some or all data, models, or codes that support the findings of this study are available from the corresponding author upon reasonable request.