How Does Public Opinion Influence Production Safety within Small and Medium Enterprises in the Sustainability Context?

Abstract

Safety accidents are a major hazard to the community and impact sustainable development. Production safety, as the basic corporate social responsibility (CSR), is mostly overlooked by small and medium-sized enterprises (SMEs) in China. The traditional safety management model of SMEs has some loopholes. Therefore, it is critical to explore other non-traditional and effective approaches. With the rapid development of the internet and the wide use of social media, the influence of public opinion on the production safety of SMEs has now become more and more powerful. However, the influence mechanism hasn’t been fully studied and understood. Based on grounded theory, this paper conducted an in-depth study on how SME behavior in production safety is influenced by public opinion. The study found that public opinion influences SME production safety through three channels: public awareness, media response, and government guidance. Public opinion influences the production safety of SMEs through intermediaries such as government supervision, the willingness of SMEs to produce safely, the behavior of employees to participate in safety management, and the self-disciplinary behavior of the industry. The impact of the willingness of SMEs to produce safely is affected by the resource guarantee ability of SMEs, therefore, the limited public opinion resources should be combined with the reality of SMEs through information exchange to promote the optimum game relationship between bodies, in order to make up for the lack of production safety resources capacity of SMEs. The conclusion of this study is crucial in understanding how to improve the production safety of SMEs by the means of public opinion to promote SMEs to engage in sustainability practices and reduce the occurrence of safety problems.

1. Introduction

In China, the numerous small and medium-sized enterprises (SMEs), accounting for more than 90% of the total number of enterprises, are fundamental to the economy developing healthily and stably. As the basic CSR, production safety is seriously ignored by SMEs in many cases in China [1]. Many studies showed that due to the large number and the geographical dispersion of SMEs and limited resources in developing appropriate safety management systems, identifying safety hazards, controlling safety risks, and maintaining a high safety standard in the workplace [2], SMEs have higher accident rate and more hazardous work environment than large-scale enterprises [2,3,4,5,6,7,8]. Even as individuals, SMEs may have relatively little influence, but taken as a whole their influence becomes much larger and crucial within the sustainability field. For solving the production safety problem of SMEs, it is natural and inevitable to improve government supervision [9,10]. At the high costs of supervision, although compulsory supervision achieves greater safety results [11], major safety accidents still emerge in SMEs, which reveals the helplessness and loopholes of compulsory supervision [12]. As the traditional safety management mode of SMEs has reached a bottleneck, other perspectives are needed for regulation.

The existing studies, mostly focusing on the motivation and boundary conditions of the safety management of SMEs, present various external factors, such as government regulation [13], laws and regulations [14], and various internal factors such as enterprise production environment [15,16,17], enterprise safety input [18,19], safety awareness [20], psychological factors of employees [21,22], and supply chain [23]. Only a few studies explore the influence of public opinion on the production safety of SMEs. With the rapid development of modern technological communications and the acceleration of social transformation, new methods of network utilization have emerged, rupturing the traditional mass media’s control over information sources [24]. Cyberspace offers a means for public expression and alters the mode of communication and ecology of public opinion profoundly [25,26]. Moreover, sudden safety accidents in SMEs can easily become a public issue. However, unlike listed companies that can steer media reports consciously, public opinion tends to focus almost exclusively on superficial phenomena [27]. Therefore, with the frequent occurrence of SME safety incidents, it is necessary to explore how to solve the production safety problems of SMEs from public opinion perspective instead of the original framework of business thinking, to emphasize the safety cost spillover. This is most significant for reducing the occurrence of production safety problems in practice.

Based on interviews and network data, this paper breaks away from the established idea of business management and uses grounded theory to explore the influencing mechanism of public opinion on the production safety of SMEs, showing that public opinion influences the production safety of SMEs through the following intermediaries, government supervision, the willingness of SMEs to produce safely, the behavior of employees to participate in safety management, and the self-disciplinary behavior of the industry.

The purpose of this paper is to clarify the mechanism of the impact of public opinion on the production safety of SMEs on the basis of clarifying the structure of public opinion. The structure of this paper is as follows. In Section 1, the research background, research purpose, and possible innovations of the research topic are introduced. In Section 2, the related literature is reviewed to highlight the purpose of this paper. In Section 3, the research methods and data sources are described. In Section 4, the research data based on the grounded theory are analyzed and the impact model of public opinion on the production safety behavior of SMEs is built. In Section 5, the model built in Section 4 is explained and analyzed. In Section 6, the research results and revelations are summarized and research development is proposed.

2. Literature Review

The research on the governance mechanism of production safety mainly concerns government regulation in China, while research on the improvement in enterprises’ own ability and social regulation is relatively rare [28]. Safety regulations and administrative intervention actions executed by the Chinese government have played an essential role in solving SME production safety problems [29]. However, it is difficult to rely only on the supervision power, which has limited resources to monitor such a large number of SME entities [29,30,31]. Therefore, safety governance in China shows emergent hybrid characteristics. The production safety of SMEs involves different parties such as the government, enterprises, employees, the public, and the media. There are complex game relationships between them. With respect to the game relationship of all parties, the government, driven by economic interests, would relax the safety supervision for promoting the economic objectives of SMEs. NGOs and industry associations still remain in the early development stages of safety management and are of little importance in current regulatory bodies [32]. Meanwhile, for short-term interests, most SMEs will choose not to invest in production safety, while employees, as a vulnerable group, lose their voice regarding production safety [33] and that will inevitably incur greater safety costs. Consequently, product safety probably becomes opportunistic and most impracticable for enterprises. Therefore, it is necessary for civil society to perform its power-restriction function of the government, through free expression, function sharing, public opinion, and participation in decision-making [34]. In essence, civil society governs itself by the majority opinions, which offers a mechanism of controlling the power through public opinion.

Studies that have examined the influence of public opinion on enterprise behavior are scattered throughout the literature concerning production safety and employee rights protection. Public perceptions have a direct or indirect impact on improving the safety situation of enterprises [35,36,37]. Basu and Palazzo found that enterprises should always pay attention to external norms including public ideas, internal coordination, and other legitimacy requirements [38]. However, public participation is still not enough in regard to safety supervision in enterprises [29], and the public has little concern for safety work in enterprises [32]. Mitchell pointed out that encouraging the public to report production safety problems with enterprises can increase the cost of illegal production and improve the government’s regulatory capacity [39].

Media supervision was an unofficial but effective external mechanism for corporate governance [40,41], an important tool for long-term impact on producer behavior [29], and could be considered an important driver of CSR. Westerwick highlights that media supervision reflects a general idea seeking to reduce negative externality in production safety, and represents the basis for optimizing public opinion environment [42]. With the development of the economy and society, the public’s requirements for safe production are constantly improving, and the media’s attention and reports on production safety accidents are also increasing, which plays an irreplaceable role in coal mine production safety supervision [43]. Chen et al. pointed out that although safety governance is mainly driven by the government, media reports also play a complementary role [44]. But because the media is influenced by the government [45,46,47], there may be a limitation on disclosure and bias could occur.

The public’s perception of safety risks depends largely on media reports. Media significantly mitigated the serious information asymmetry problem between the public and producers [48]. The public thinks that some media lack objective and accurate reports on the safety problems of SMEs, which cannot meet the information needs of the public [49]. The publicity power and influence of the media greatly limit the willingness and ways of the public to participate in safety decision-making within SMEs [43]. When the media repeatedly or constantly reported the behavior of enterprises to the public, the accumulative effects on the public would significantly affect the enterprises’ production and sales. From this point of view, the media could put pressure on managers to correct their management behavior [41,50].

Public opinion is an effective social mechanism that compels corporate managers to adopt safety behaviors [51] and regulate the tense relationship between other stakeholders and enterprises [52]. The consistent, long-term pressure generated by public opinion can be utilized to promote all aspects of production safety, and the re-occurrence of safety issues in SMEs is largely attributed to the failure of maintaining such pressure [43]. For SMEs, one of the major reasons for neglecting production safety is the lack of good public opinion on environmental management and supervision [43]. The above-mentioned literature studies the impact of the public or media or public opinion on production safety as a whole, instead of the driving mechanism of them for production safety. Based on network data, Ye et al. [1] introduce public opinion into management, using grounded theory, dividing the public opinion environment into three dimensions: public awareness, media response, and communication channels, and indicate that the influence of public opinion on the production safety of SMEs is mediated by the government supervision and the degree of advice.

Although the existing research views public opinion as an intrinsic variable that influences decision-making and studies the interactive relationship between public opinion and production safety of SMEs, the original data used in the current research are mainly obtained from news topics and user posts after the occurrence of major safety accidents in China. In other words, the information mainly comes from stakeholders such as the public and the media, which generates two obvious shortcomings in the research results. (1) In 1993, Japanese scholar Ito Yang-yi states that the mass media, the government, and the public are the three most important poles affecting public opinion, which are independent of each other, jointly promoting the formation and development of public opinion. The three-pole model reasonably and concisely grasps the basic process of the formation of public opinion [53,54]. However, the current research subdivides the public opinion environment into two main bodies: the public and the media without considering the role of the government in the formation of public opinion. (2) The influence of stakeholders, such as the government, shareholders, and employees, is not considered in the production safety of SMEs, resulting in the proposed model not highlighting the responsibility subject of enterprise production safety.

Generally speaking, the relationship between public opinion and the production safety of SMEs has achieved some success. However, the model still needs to be expanded and deepened to evaluate the public opinion environment about the mechanism of SME production safety. This implies that there are some limitations to the theories. Therefore, it is necessary to explore how to promote SME production safety by the means of public opinion. Consequently, based on the grounded theory, to ensure production safety much more internally, this paper deeply explored the conceptionual category of public opinion by combining interview questionnaires and networking data and excavated various public opinion rules or system defects that may affect the specific management process of the production safety in SMEs.

3. Research Methods and Data Sources

3.1. Research Methods

To achieve the research objectives, the study adopted grounded theory. This is often used to form a theory grounded in the raw data systematically gathered and analyzed and then to describe and explain behaviors under research [55,56,57,58,59]. The emergence process of production safety in SMEs is an example of an “incompletely understood phenomenon”, which needs theoretical construction. The text information collected by grounded theory through interviews is rich, informative, and deep, allowing people to explore the hidden deep factors and connections through the surface phenomena of things, which is suitable for explaining the questions of “how” and “why”. Since Glazer and Strauss applied this method to social science in 1967, grounded theory has been widely applied in various fields, including anthropology, psychology, management, etc. [55].

3.2. Sample Selection and Data Collection

Considering that the general public does not pay much attention to production safety, and the focus of this study has a certain degree of professionalism, this paper mainly uses two channels to collect data.

(1)

Interview records

Through the introduction of social relations of the research team and on-the-job postgraduates, this paper selects the emergency management government staff, SMEs owners, enterprise safety management staff, workshop managers, as well as experts and scholars in related fields who have a certain understanding of enterprise production safety and public opinion to conduct semi-structured interviews. These interviewees not only were experienced in safety management and willing to participate but also had enough time and could clearly express their views [55]. According to the “theory saturation principle” of the grounded method, 27 subjects were interviewed. The samples contained different occupation backgrounds and different educational levels and were distributed in the following seven cities: Beijing, Shanghai, Nanjing, Xuzhou, Lianyungang, Zhengzhou, and Hefei. To ensure they had sufficient knowledge of the field, the interviewees’ selection was limited to those with at least five years of work experience.

Table 1. Basic information of interviewees.

Gender	Number (Percentage)	Age	Number (Percentage)	Education Level	Number (Percentage)
Male	20 (74.1%)	24~30	2 (7.4%)	Associate degree or below	5 (18.5%)
Female	7 (25.9%)	31~40	12 (44.4%)	undergraduate course	16 (59.3%)
		41~50	7 (25.8%)	Master’s degree or above	6 (22.2%)
		51~60	6 (22.2%)
Working experience	Number (Percentage)	Occupation	Number (Percentage)
5–10 years	5 (18.5%)	government staff	3 (11.1%)
More than 10 years	22 (81.5%)	SMEs owners	3 (11.1%)
		safety management staff	9 (33.3%)
		workshop managers	8 (29.7%)
		experts and scholars	4 (14.8%)

shows the demographic breakdown of the participants.

Based on the relevant literature, a group of experts worked out and adjusted the interview outline according to the interview effect. (The interview outline is shown in

Table 2. Interview outline.

Interview Themes	Specific Question
Basic information	Age, sex, education level, occupation
Public Concern and Action on Safety Accidents in SMEs	Do you often pay attention to news or media reports about safety accidents in SMEs? Why do you do this? What actions have you taken in response to safety incidents that cause you concern?
Media Reports on Production safety	What media channel do you focus on for SME production safety accidents? What do you think since the media has done poorly according to their reporting of safety incidents? Why do these problems occur?
Government’s Voice on Production safety	In your opinion, why is there a problem in relation to the government’s voice concerning safety-related accidents in SMEs? Why do these problems occur?
The influence of public opinion generated by the interaction of the public, media, and the government with respect to the production safety of SMEs	When a production safety accident occurs, which stakeholders could mostly influence a change in public opinion that would lead to a corresponding behavior or reaction, and thus affect the production safety of SMEs? What impact would these effects have? Is it positive or negative?

). The expert group is composed of two advisors, four Ph.D. students, and two SME owners with rich experience in production safety. In each interview, we interviewed participants with an open attitude, allowing them to express their views, and clarified any ambiguous answers. With the consent of the participants, we recorded the interview and converted this verbatim into text within 24 h for further analysis [55].

Considering the geographical constraint, formal interviews were conducted face to face, via QQ, WeChat, or telephone. In order to establish a bond, the interview is first conducted by chatting with the interviewee/interviewee. The discussion and interview took about 40–90 min. At the end of each interview, the interview materials were sorted and analyzed, the ideas and core conceptions were gradually clarified, and the interview scheme was further improved according to the analysis results. If the information obtained in the interview starts to repeat, and the new interviewee cannot provide new information—that is, the theory reaches saturation—the interview was stopped. About 50,000 Chinese characters of transcript were obtained.

(2)

Network data

In addition, after screening and comparison, this paper used second-hand information as verification materials, all of which are reports of major production safety accidents and high-quality user comments on websites, like Tencent, Netease, the China Production Safety Network, and others. The incidents include the accidental explosion at Ruihai Company Tianjin Port on 12 August 2015, the accidental explosion at Rongrong Metal Products Co., Ltd. in Kunshan Development Zone on 2 August 2014, and the accidental explosion at Jiangsu Tianjiayi Chemical Co., Ltd. on 21 March 2019. Interview data and network information complement and verify each other to improve the reliability and validity of the data.

4. Research Data Analysis and Model Construction

As proposed by Strauss and Corbin, the three-stage procedures of open coding, axial coding, and selective coding were used to analyze the data [55]. During the data analysis process, all analyses were strictly based on raw data. To ensure the consistency and reliability of the data analysis, the research team randomly selected 18 of the 27 valid interviews for data analysis performed by member X. The remaining 9 interviews were conducted to member Y to independently determine categories [56]. The categories identified by the two team members were cross-compared to complement the new categories. The details of the data analysis process are as follows.

4.1. Open Coding

Open coding is a word-for-word coding, labeling, and logging of the raw data for identifying initial conception through continuous comparative analysis [58]. In the label selection, this paper attempts to use the original words of the interviewees and to name the conceptions directly or extract related conceptions from them in order to eliminate the bias of the coder as much as possible. In the process of coding, objective and rational conceptions and categories were advised [56]. To comprehensively explore the influence of public opinion on the production safety behavior of SMEs, this paper excludes any relatively simple and excessively vague answers from the interview records. After the extensive processing and analysis of the labeled sentences, more than 500 original sentences and corresponding initial conceptions were obtained. Inconsistency in the initial conceptions was eliminated, and initial conceptions were selected if they occur more than three times. Through repeated research and analysis of the relationship between these conceptions, an abstract field level was formed, that is, a category. Therefore, the data were analyzed to reveal the connections between different conceptions through open coding, and hidden contexts and causal relationships were excavated [56]. As a result, 53 initial conceptions and 22 categories were obtained. For brevity, this paper only lists the representative original interview sentences and initial conceptions for each category, as shown in

Table 3. The process of conceptionualization and categorization.

Subject	Representative Sentences from the Interviews	Conception	Category
Public	I only pay attention to major safety incidents that are covered by many media reports.	Focus on hot topics	Safety requirement
	It is relevant to our immediate interests. By paying attention to safety incidents, we can see if there are similar safety risks around us.	Risk-Informed
	Keep an eye on the latest official news and determine what causes the accident.	Explore the truth
	Discuss with colleagues in the office and check other people’s comments on the relevant news on the Internet.	Express opinions	Concern public opinion
		Public interaction
	It’s so uncomfortable to have such a big safety incident when will we realize that production safety is the most important thing?	Safety sensitive	Public participation
		Safety baseline
Media	In order to cater to special interests, the media are unbalanced when disclosing the opinions of the government, enterprises, and the public.	Cater to interest	Truthfully report
	Usually, there was no report, but the accident was exposed.	information disclosure
	It is difficult to obtain information from production safety reports, and there are many obstacles to interviews, so the real cause of the accident has not been uncovered.	Cost of collecting and editing
		Explore the truth
		Information monopoly
	The social concern about production safety is not high. People pay more attention to entertainment news.	Social environment
	The interaction of public opinion is insufficient, which cannot fundamentally solve the problem of production safety.	Public opinion interaction	Reaction to public opinion
	Media-related reports are not true or comprehensive.	Public opinion quality
	The amount of relevant news items and the coverage of accident-related topics are limited, and the effect of public opinion is limited.	Public opinion intensity
	Accident reports are not comprehensive or investigative.	Public opinion tenacity
	Pay attention to safety accidents through newspapers, television, etc.	Traditional media	Transmission route
	Pay attention to safety incidents through online media sources such as Tencent, Sohu, and Netease.	Internet media
	Interpersonal communication of relevant interest groups and communities in which the company that experienced the accident is located.	Interpersonal communication
Government	The government has repeatedly stressed the need to uncover the truth and severely punish the person responsible for the accident.	Explore the truth	Information monopoly
	Government disclosure of safety information is inadequate. Feedback of safety supervision effect prefers to report only what is good rather than concealing what is unpleasant.	Regulatory effect feedback	Government voice
		Information disclosure
	To a certain extent, it also shows that in peacetime, safety inspections by the government are not prioritized, and front-line workers are the ultimate victims.	Selective law enforcement	Regulation implementation
		Collusion in rent-seeking
		Regulatory aliasing
	The government only pays attention to GDP and replaces supervision with fines, which leads to frequent safety accidents.	Interest-oriented
		Replaces supervision with fines
		Frequent accidents	Regulation effect
	Facing the pressure of public opinion, the government strengthened the supervision of production safety.	Public opinion reaction
Enterprise	Lessons paid for with blood are in front of us. We realize that we should not only pay attention to profits, but also carry out safety training, equip employees with safety equipment and facilities, check whether there are potential safety hazards, and make rectifications in the enterprise.	Willingness to offer safety training	Willingness to invest in safety
		Willingness to update equipment and facilities
		Willingness to inspect hidden dangers	Willingness to self-check and self-rectify
		Willingness to rectify hidden dangers
	We also want to be safe in production, but our company’s profit is so small that we can only pay attention to profits. Moreover, employees are generally low-skilled workers, and there is a lack of professional safety supervisors. It is very difficult to carry out safe production.	Meager profit	Fund guarantee
		Interest-oriented
		Staff quality	Human capital guarantee
		Lack of professionals
	There will be a negative impact, such as whether the enterprise needs them or not, and for the sake of insurance, the government may force the enterprise to carry out a safety investment or safety training.	Safety investment	Safety prevention
		Safety training
	Government regulation enforcement has increased the frequency of safety preparedness exercises for our enterprises.	Safety plan drills
	Our company began to pay attention to the construction of safety standardization.	safety standardization construction	Daily safety management
	In general, there are more positive effects. The pressure of the government and public opinion urges enterprises to conduct safety inspections and eliminate hidden dangers.	Safety check
		Investigation of hidden dangers	Safety rectification
	After the Xiangshui accident broke out, our company has also strengthened its commitment to rectifying hidden dangers.	Rectification of hidden danger
	Our company requires investigating and dealing with the safety accident seriously. Lessons are learned to prevent repeat accidents.	Dealing with safety incidents
Employee	The company or government does not provide sufficient safety training, and employees cannot understand the production safety risks in the enterprise.	Safety training	Risk-Informed
		Understanding risks
	At present, there are many channels for reporting. Employees can report irregular production or potential safety hazards to relevant departments.	Hidden dangers tip-off	Hidden dangers tip-off
	Express safety-related opinions and suggestions to business managers.	Expression of opinion	Safety suggestions
	Seeing other people who operate illegally; I want to persuade others, but I don’t know how to say it.	Communication method
Industry Association	Frequent safety incidents have reflected the substantial shortcomings of industry associations in obliging enterprises to implement systemized health and safety management, ensure the intrinsic safety of equipment and facilities, and develop the ethics of employees.	Construction of Safety and Health Management System	Safety standards construction
		Construction of Essential Safety of Equipment and Facilities
		Construction of Moral Standards	Moral standards construction

. These listed items show the process of classifying the factors that influence production safety among SMEs. Five conceptions, including the common features among two or more subjects, are in Table 3: explore the truth (public, media, and government), information monopoly (media and government), interest orientation (government and enterprise), safety standardization construction (industry associations and enterprises), and safety training (enterprises and employees). This simply reflects some degree of interaction between the parties.

4.2. Axial Coding

Axis coding is the clustering analysis of the segmented data in open coding in order to establish the connections between different categories. When forming connections, clues are examined whether categories are latently connected conceptionally [56]. Considering the intentions and contexts of the subjects’ responses, the associations among different conceptions were established based on the principle of compliance [56]. This stage of data analysis included the investigation of the relationships between conceptions and categories as well as the relationships between categories and main categories. The paper analyzes the 53 captions and 22 categories in depth. By deducing the logical relationship between them, the author summarizes 9 main categories, as shown in

Table 4. The axial coding process.

Subject	Category	Main Category
Public	Safety requirement	Public awareness
	Concern public opinion
	Public participation
Media	Truthfully report	Media response
	Reaction to public opinion
	Transmission route
Government	Government voice	Government guidance
	Information monopoly
	Regulation implementation	Government regulation
	Regulation effect
Enterprise	Willingness to invest in safety	The willingness of production safety in SMEs
	Willingness to self-check and self-rectify
	Fund guarantee	Enterprise resource guarantee ability
	Human capital guarantee
	Safety prevention	Production safety of SMEs
	Daily safety management
	Safety rectification
Employee	Risk-Informed	Employee safety participation behavior
	Hidden dangers tip-off
	Safety suggestions
Industry Association	Safety standards construction	Self-discipline behavior of Industry association
	Moral standards construction

.

4.3. Selective Coding

In selective coding, the researchers discover the core category with more clustering significance by further inducing the main category, systematically connecting one core category with other categories by describing a “storyline”, and further collecting information to confirm this connection.

The government, media, and public do not affect the production safety of SMEs independently [60]. There are differences in terms of their value judgments and interests, which prompt them to play games for their own interests. The government is dominant in guiding public opinion. In the omni-media era, the government interacts with the public directly or indirectly. The government will selectively disclose production safety information for its own interests. When the media report on production safety incidents, they often need the instruction of the administrative leaders, to improve the ratings and click-through rates, the media often ignore or fail to report the issue of production safety objectively. Although the public will express their opinions or views on the issue of production safety through the mass media, their opinions are often filtered by the media. The three subjects form public opinion and reflect public opinion in the continuous interaction, which affects the production safety of SMEs.

Therefore, the main categories of the public, the media, and government guidance in its main categories are further generalized as ‘public opinion’, which includes ‘public participation’, ‘media response’, and ‘government guidance’, as shown in Figure 1.

Based on clarifying the subdivision structure of public opinion, this paper is to explore the public opinion impact on the production safety of SMEs. Firstly, it highlights the influence path of public opinion on the production safety of SMEs as the core category, and then extracts the complete storyline based on the core category: public opinion affects the production safety of SMEs through the following four major factors, government supervision behavior, the willingness of SMEs to produce safely, the behavior of employees to participate in safety management, and the self-disciplinary behavior of industry associations. In addition, the impact of SMEs’ willingness to produce safely is affected by the ability of their resource guarantee. Finally, the mechanism model of public opinion on the production safety of SMEs is formed. Figure 2 shows the control structure for this core category over other categories.

4.4. Saturation Test

In order to fully identify the impact of public opinion on the production safety of SMEs, we should test whether the information contained in the new sample can provide new conceptions or categories. This paper tests from two aspects: On the one hand, the remaining nine theoretical samples were selected for coding analysis by member Y of the research group. On the other hand, we conducted coding analysis on the collected network data. The three levels of qualitative analysis based on the grounded theory were conducted independently [56]. The results obtained were consistent with the factors found in the earlier analysis, and no new relationships or categories were discovered, which demonstrates that the theory is saturated. Therefore, the model passes the saturation test, and data collection and analysis were stopped.

5. Model Interpret and Analysis

5.1. Interpretation of Model Mechanism

The theoretical model in this paper focuses on the centralized response of many institutions, such as the government, the media, the public, industry associations, and employees, to the impact of the operations of enterprises in the Chinese context. Its mechanism is explained as follows:

The occurrence and development of safety incidents are closely related to the public interest, and the public should have the right to know. From the perspective of its self-interest, the government’s dissemination of information to the public is often limited to a certain extent, by communicating its own view of production safety issues through the media to inform the public and lead public opinion. While gaining the attention of the public, the media is motivated to pursue influential news items. However, the social environment obviously places an insufficient emphasis on production safety in SMEs. Thus, it is difficult for the media to conduct interviews about safety incidents. This situation leads to the inadequate disclosure of safety-related information, and it is impossible to penetrate the information monopoly. Moreover, the number of reports is relatively small and it is difficult to form a hot spot of public opinion that demonstrates quality, intensity, and toughness. By utilizing various means, the public has an obligation to comment on, verify, and disseminate the existing information and combine it with their own personal experience, and the intensification of public opinion ultimately increases the pressure on the government or responsible persons, thereby promoting the development of the situation that is in the public interest [61], which is vital for resolving issues about the production safety of SMEs.

The behavioral response of the government, the media, and the interaction of the public form a targeted public opinion. This process can be considered in three dimensions, namely, the ‘public awareness’, the ‘media response’, and ‘government guidance’, and referred to the 23 conceptions and 8 categories that relate to the media, public, and government.

The government will adjust its supervision behavior with respect to the production safety of SMEs since it perceives the change in the corresponding structure as well as the pressure arising from public opinion. For SMEs, government supervision is mainly to supervise their production process by the means of relevant policy guidance and on-the-spot supervision, which makes sure that production behavior complies with its rules. However, the government generally focuses on the economic objective and deals with the regulatory objectives selectively without feedback on the regulatory effect. This eventually leads to frequent safety risks and accidents in SMEs under information control. A change in public opinion has led to more intense supervision behavior by government departments, whose behavior is not only adapted to the requirements of the accident itself but also affected by the pressure and scope of public opinion. This type of government behavior can be highlighted by considering two categories, namely, ‘regulation implementation’ and ‘regulation effect’ as well as the corresponding seven conceptions. This path is in line with Delmas et al. [62] and Jia et al. [30]. They point out that companies perceive greater pressure from market mechanisms and non-market mechanisms, and are more likely to assume social responsibility. This finding is also in line with Holtbrügge and Dögl [63], who point out that, if the CSR of managers mainly comes from external regulatory pressure and punishment, as opposed to a rewards-based system, this can represent a more effective means of promoting CSR.

The owners of SMEs will adjust their willingness to produce safely by perceiving the changes in the corresponding structure and the pressure from public opinion. Facing the pressure of public opinion, the owners of SMEs will have the willingness for safety investment and self-examination, and self-rectification, which constitutes the cornerstone of the production safety of SMEs. However, the fact that it could be fulfilled or not is dependent upon the willingness and the ability of enterprises to invest their resources, particularly their fund guarantee and human capital guarantee. The above corporate attitudes can be considered in terms of the ‘willingness to invest in safety’, ‘willingness to self-monitor and self-rectify’, ‘fund guarantee’, and ‘human capital guarantee’, as well as the corresponding eight conceptions. This path conforms with Koster’s Knowledge-Attitude-Practice (KAP) theory, which was developed in 1960 and proposes that the knowledge and the attitude of decision-makers will affect their behavior. At the same time, the resource-based theory holds that CSR is generally consistent with the company’s own resource capacity [14]. Namely, the influence of SMEs’ willingness to produce safely on production safety should be regulated in accordance with the strength of the enterprise.

With respect to participating in safety management within enterprises, employees will adjust their behavior by perceiving the changes in the corresponding structure and the pressure from public opinion. The excessive focus on profitability results in an inadequate safety management system. Managers are also insufficiently motivated to prioritize safety management and tend to avoid or repress suggestions of potential hazards. The pressure from public opinion encourages employees to more actively propose suggestions about safety to protect their own interests. This type of employee behavior can be emphasized by considering three categories, namely, ‘safety risk informed’, ‘hidden danger tip-off’, and ‘safety suggestions’, as well as the corresponding five conceptions. This conclusion is not verified in Ye’s [28] study. However, we should believe that some improvement in the cultural quality of employees and a better awareness of safety risks would encourage employees to be more willing to participate in safety management in enterprises.

In terms of the production safety of SMEs, industry associations will adjust their self-disciplinary behavior in response to their perception of changes in the appropriate structure and pressure from public opinion. Confronted with public opinion pressure, industry associations should reinforce the development of standardized safety measures and employee ethical standards. This type of industry association behavior can be addressed by considering two categories, namely, ‘safety standardization construction’ and ‘moral standards construction’, as well as the corresponding three conceptions. This is consistent with Sharma et al.’s [64] study, which highlights that public opinion can promote industry associations to supervise member enterprises better, and enable industry associations to use their own opinion and leadership role in the supervision process to promote the production safety of enterprises.

SMEs not only perceive changes in the corresponding structure and pressure of public opinion, but also adjust their behavior by sensing changes in the areas including government regulation behavior, the willingness of owners to produce safely, the employees’ safety participation behavior, and the industry associations’ self-disciplinary behavior. In the absence of relevant pressures, economic interests continue to represent the most dominant behavior of SMEs, and their safety prevention, daily safety management, and safety rectification measures are largely emergency reactions. The changes in the pressure arising from public opinion, government, owners of SMEs, employees, and industry associations have stimulated changes in the original production safety behavior of enterprises. The above enterprise behavior can be considered with respect to ‘safety prevention behavior’, ‘daily safety management behavior’, and ‘safety rectification behavior’, as well as the corresponding eight conceptions.

5.2. Analysis

The model can essentially reflect the impact of public opinion on the production safety of SMEs. SMEs lack resources and capacity, lack conditions for systematic governance, and their safety objectives are clear, but often are ignored, unable to enter the decision-making agenda in time, and safety management is mostly emergency response. Large enterprises have a strong professional safety management system, which is more than enough to deal with superficial and obvious safety problems. The focus of safety management is to find the deep hidden danger. The role of safety governance in SMEs is to strengthen the supervision of known problems and possible resource coordination. Therefore, SMEs need the external environment to provide a strong standard to participate in the correction and filling. Timely and accurate interaction helps to urge SMEs to make correct decisions, but the external environment obviously lacks such interaction, which is where public opinion can play a role.

The positive influence of public opinion is to urge SMEs to implement the neglected safety objectives. SMEs perceive the intensity and structure of public opinion pressure and adjust their decisions to promote the implementation of safety management. However, social attention to the production safety of SMEs is obviously insufficient. Accordingly, the media lack the dynamic to report, and the reporting structure focuses on hot events, which cannot sustain attention for a long time. The insufficient total amount and shallow structure of public opinion further promote the pressure dependence of public opinion—that is, safety accidents are ignored without exposure, and decision-making behavior is not out of the internal needs of enterprise safety management, but catering to governance. In addition, the channels of communication are limited, and the relevant stakeholders deliberately reduce or even hinder information dissemination for economic goals, which leads to the weak social mobilization ability in the field of production safety of SMEs, and it is difficult to attract more resources to cooperate. This is the crux of the safety management dilemma of SMEs.

In addition, in China, an in-depth study of the impact of public opinion on the production safety of SMEs needs to clarify an embedded structure: the relationship between the government and enterprises. The government and SMEs pursue economic benefits together, which increases the risk that the safety objectives of enterprises are ignored. Different from the game relationship between large enterprises and the government, due to weak resource capacity, and facing relatively high regulatory requirements, SMEs can only commit themselves to power for protection. Therefore, the relationship between the government and SMEs is largely dominant–dependent. SMEs are not independent decision-makers. Production safety behavior is largely controlled by the degree of government supervision, and other external influences are largely realized through government supervision. A correct understanding of the relationship between the government and SMEs is the key to a deep analysis of the dilemma of production safety in SMEs. The negative influence of public opinion is also closely related to the relationship between the government and enterprises, especially after the occurrence of major safety accidents. In the face of public opinion pressure, the government may engage in all-around activities for the sake of insurance. For example, whether the enterprise needs a certain safety investment or safety training, enterprises are forced to carry this out. On the surface, these regulation objectives meet social expectations, but they are not in line with the reality of SMEs. On the contrary, they will interfere with the safety management of SMEs. This is also an important reason for the formation of the government’s control over SMEs. Therefore, public opinion should truthfully reflect the actual situation of SMEs, mobilize more social resources to interact, and make the government’s regulation objectives adapt to the actual situation of SMEs to reduce the behaviors of SMEs such as transferring risks to employees and seeking power protection. In addition, the relationship between SMEs and employees is largely economically dominant–dependent, which also affects the normal play of the positive role of public opinion. The degree of employees’ safety participation is very low, and generally, they can only leave their job to avoid the possible safety spillover cost.

6. Conclusions and Revelations

6.1. Conclusions

Based on grounded theory, this paper analyzes the influence mechanism of public opinion on the production safety of SMEs and obtains the following valuable findings.

(1)

Public opinion can influence social resources and the multi-stakeholder game, and promote SMEs to assume the responsibility for production safety. However, the insufficiency of public opinion resources and the unbalances of the structure, prevent society from paying attention to them; consequently, a low-level cycle of “low attention → low reporting → shallow structure → low attention” forms. Therefore, to solve this problem, we should start with public opinion and adjust the total amount and structure of public opinion resources in the field of production safety of SMEs.

(2)

Public opinion influences the production safety of SMEs in three respects: public awareness, media response, and government guidance. Public opinion influences the production safety of SMEs through the following intermediaries, such as government supervision behavior, the willingness of SMEs to produce safely, the behavior of employees to participate in safety management, and the self-disciplinary behavior of industry associations. The impact of SMEs’ production safety is also affected by the resource capacity of enterprises.

(3)

Public opinion has both positive and negative effects on the production safety of SMEs. Therefore, the limited public opinion resources should be based on the reality of SMEs, and promote the game relationships among the main bodies through full information exchange to make up for the lack of production safety resource capacity of SMEs.

6.2. Revelations

6.2.1. Theoretical Aspects

This study begins from the area of public opinion to study the impact of public opinion on the production safety of SMEs. SMEs are unable to mobilize social resources to influence public opinion. The driving force behind the production safety of SMEs can be largely attributed to external environmental pressure, which is verified by the grounded analysis presented in this study. For this reason, an obvious means of resolving the issue of production safety in SMEs is to intensify the pressure generated by public opinion. Specifically, it adds other stakeholders into the unbalanced relationship from the government → the enterprise→ the employee and helps to build a closed-loop restraint: that is from public opinion (public-media-government interaction) → the government → Industry Association → the enterprise → the employee → public opinion.

6.2.2. Practical Aspects

SMEs are large in number, small in scale, distributed widely, and lack the necessary conditions for systematic safety management. We should break away from the original business management, and explore some ways of solving the production safety problems of SMEs in the views of public opinion.

(1) The higher the pressure of public opinion perceived by enterprises, the more likely it is to implement production safety. In the field of the production safety of SMEs, we should realize the importance of public opinion, and make use of the joint efforts of stakeholders to perform production safety, while the dissemination or absorption of enterprise production safety information through public opinion can attract more social resources so that more stakeholders can share and burden the enterprise decision. (2) We will carry out public safety literacy education, encourage the public to participate in the expression of opinions on production safety, meet the public’s requirements better for information disclosure, and encourage more NGOs to participate in the field of production safety. (3) The government should build a “constructive partnership” with the major media, intensify propaganda and education on the production safety of SMEs, consciously enhance the voluntary responsibility for the production safety among SMEs, and facilitate financing, financial support, tax relief, industry access, and other aspects for assisting the development of SMEs effectively. (4) China can learn from the United States in the promulgation of relevant laws to encourage employees through public opinion publicity, expose the production safety problems, and improve the independence and influence of labor unions so that they can better protect the rights of employees in production safety. (5) Industry associations can be encouraged to participate in the field of production safety to enforce the additional operational standards, and ensure the more effective supervision of the production safety in SMEs.

In recent years, European and American countries have achieved great safety performance through compulsory supervision, but the cost of supervision is very high. Many Western corporate social responsibilities are imposed under the pressure from NGOs, consumers, or investors, as well as other alternative mechanisms. Therefore, the model in this paper conforms to international practice. The most fundamental solution to the problem of production safety in SMEs is to ensure the constant optimization of the rules of public opinion and to enforce the government, media, and public opinion environment to participate in a fully interactive game, thereby encouraging the public opinion environment to engage with the production safety of SMEs, while further allowing more stakeholders to participate in the enterprise governance. We should also pay attention to the fact that there are many factors affecting the production safety of SMEs. Successful governance cannot be accomplished in a single stroke, yet relies on the accumulation of the enterprise’s development efforts and the continuous evolution of the institution.

6.3. Development

Unfettered by the scope of operation management, this paper makes use of the grounded theory and builds an internal relationship between public opinion and enterprise decision-making behavior. However, this study lacks empirical support. For enhancing the persuasiveness and promotion value of the results, in the future, we will design the relevant questionnaires according to the main categories of a model, and take an empirical test for the model concerning objectivity and scientificity.