A Survey Study on Public Risk Perception of Digital Governance in Chengdu

Research background

In today’s era of digitalization, data resources have become a common element permeating every aspect of our lives. During the “13th Five-Year Plan” period, China achieved significant results such as leading global information infrastructure, the transformation and development of new digital business models, and steady progress in digital government construction. However, with the deepening of a new round of technological revolution and industrial transformation, how to adapt to the new trends of digital change, enhance the capabilities and levels of digital governance, and seize the initiative in data resources, is a contemporary challenge we need to address. A key component of modernizing the national governance system is the modernization of urban governance, and smart city construction perfectly aligns with this requirement. It makes cities smarter and is an inevitable path to advancing the modernization of national governance. The development of smart cities in China began in 2008 when IBM introduced the concept of a Smart Planet, which was understood by Chinese city operators; with the deepening of urbanization and the advancement of information technology, various regions started exploring independently around 2012 ; by 2 015 and 2016, the construction of new-type smart cities entered a new phase, and big data technology became a national strategy; today, there is a surge in the craze for smart cities across the country. As a megacity in southwestern China and located within the Chengdu-Chongqing Economic Circle, Chengdu stands out in this regard. With the construction of “Smart Chengdu” as a starting point, we will accelerate the construction of new smart cities and promote higher quality development in all fields of Chengdus economy and society.

Objectives of the study

As the concept of smart cities continues to evolve, current smart city construction has shifted from an initial focus on technology to a greater emphasis on people-oriented approaches. Therefore, this paper adopts a public perspective and primarily employs both quantitative analysis and qualitative research methods. By investigating citizens views on specific application systems and scenarios in Smart Chengdu, it explores the public’s perception of risks associated with digital governance and proposes corresponding preventive measures for potential risks. We hope that these measures can help the digital construction management platform mitigate certain risks, promote the healthy development of digital construction management, while upholding the people’s position and exploring humanistic governance through digital means.

Research significance

By reviewing relevant literature, we find that previous studies on smart cities have mostly explored the intrinsic potential and advantages of technology and data empowerment from a macro perspective. Since President Xi Jinping pointed out in his report at the 19th National Congress of the Communist Party of China: “We must rely on scientific and technological power to provide strong support for building a digital China and smart cities, and we must put people at the centre, adhering to improving people’s livelihoods in development.” More scholars have thus focused on the “human” factor, analysing the current status of smart city construction from aspects such as citizen satisfaction, smart city performance, and public service efficiency. However, studies that delve into specific cities and use public

experience to reflect on the risks of smart city construction remain scarce. Therefore, this paper selects some citizens in Chengdu and investigates their perception of the risks associated with Smart Chengdu, aiming to analyze the current status and existing issues of smart city construction more objectively and humanely, thereby providing a reference for the healthy development of digital governance.

Research methods Questionnaire survey method

This study conducted online and offline questionnaire surveys, distributing questionnaires in Sichuan Province, particularly within Chengdu City. The online sources for samples included sample services, WeChat, and mobile submissions. After collecting the questionnaires, they were screened based on the location of the sample, the duration of completion, and the logical consistency of the responses, retaining valid samples. Subsequently, tools such as IBM SPSS Statistics and Excel were used to investigate the correlation and regression of variables, generating a series of charts and function expressions. Data was organized to establish analysis tables and related models, followed by statistical data collection. Finally, various data and materials were compiled to verify the rationality of the hypotheses and propose corresponding recommendations based on citizens perceptions of risks associated with digital governance in Chengdu.

In-depth interview method

We conducted a survey on the four smart Chengdu construction units, namely Chengdu Smart Chengdu Research Institute, Chengdu Online Governance Office, Chengdu Transportation Commission and Chengdu Urban Management Commission, and conducted indepth interviews with relevant informed persons to obtain firsthand interview materials.

Case study method

With the popularization of digitalization, smart city management has been implemented in many cities. This paper focuses on Sichuan Province, with particular attention to Chengdu City, comparing it with samples from other regions in Sichuan to observe the prevalence of Smart Chengdu. The sample includes six types of professions, and we separately analyze the questionnaire results from government departments, comparing them with the overall sample to study the differences in outcomes brought by different professions.

Theoretical assumptions

a) Hypothesis 1: Public support for digital governance as a whole
b) Hypothesis 2: The public’s perception of different risks of smart Chengdu is different.
c) Hypothesis 3: The risk perception of different groups of people in smart Chengdu is different.
d) Hypothesis 4: The risk perception of the public on smart Chengdu is influenced by certain application scenarios.

Data acquisition

First of all, a questionnaire was released through Questionnaire Star, and 411 valid samples in Sichuan province were randomly collected as the research object of public risk perception of digital governance in Chengdu.

Secondly, a random interview was conducted with several residents in Sichuan to investigate their attitudes towards the practicality and convenience of risk perception factors, as well as their views on the possible risks of smart Chengdu digital governance (including age adaptability, security, degree of data sharing, data consistency and algorithmic decision).

Measurement of indicators Risk perception degree

After collecting 411 samples from Sichuan Province, the average values of age-friendliness, safety, data sharing degree, data consistency, and algorithmic decision-making survey results for each sample were taken to obtain the risk perception level of each sample. Then, the average risk perception level of all samples was calculated to derive the overall average risk perception level of all samples.

Independent variable

The selection results of the 15thi and 16th questions for all samples were assigned by the IFS function in excel software, which was assigned 0, 1, 2, 3 and 4 according to the degree from small to large. The value assigned by the selection results of the 15th question was practicality, and the value assigned by the selection results of the 16th question was convenience.

Control variables

The familiarity is taken as a control variable, and the specific operation is the same as the independent variable. In order to analyze and compare the publics familiarity with the 14 software of Smart Chengdu, familiarity values are adopted.

When investigating the impact of familiarity on average risk perception, considering reliability, samples with a familiarity level of 0 and usage frequency of 0 will be excluded. Samples with an experience rating of “very satisfied” will be selected for analysis, with their corresponding familiarity level data being “average familiarity.” In the partial correlation analysis using SPSS software, the influence of each factor on risk perception will be analyzed, with average familiarity serving as a control variable to exclude its effect.

Related models

In this data analysis, the primary linear regression model was used. First, samples with a knowledge level of 0 and usage frequency of 0 were excluded. Samples that reported “very satisfied” usage experience were selected and their corresponding data on practicality, convenience, and average risk perception were analyzed for correlation. If a linear relationship exists, a x-y linear regression equation can be plotted to determine its positive or negative correlation and the degree of influence (judged by coefficients).

Demographic variables statistics

Statistics show that the ratio of men to women is about 3:2, and the majority are between 20 and 40 years old, about 52%, while those over 60 years old only account for 4%. This shows that fewer elderly people use digital software.

Public perception of different risks

The survey data shows that the public’s perception of different risks is different. As shown in Figure 1, the public’s perception of the risk of data sharing is the highest, and the perception of the risk of data consistency is the lowest. Therefore, Smart Chengdu can improve and pay attention to some higher perception risks.

Influence of influencing factors on the degree of risk perception Practicality

Figure 2 shows the correlation analysis between practicality and average risk perception after controlling variables and averaging familiarity. As shown in Figure 2, the significance (double tail) =0.021<0.05, so practicality has an influence on average risk perception.

In terms of practicality, some respondents made the following comments:
a) The elderly do not have such needs, and the operation is also troublesome.
b) Easy
c) For me, this kind of APP should be easy to use.
d) Okay
e) It’s easy to operate. For example, Tianfu Health Pass, electronic social security card, etc.

As can be seen from the above comments, respondents do not have much difficulty in the practicality of Smart Chengdu, but there is still room for improvement in the practicality of smart Chengdu.

Figure 4 shows the correlation analysis between convenience and average risk perception after controlling variables and averaging familiarity. As shown in Figure 4, the significance (double tail) =0.001<0.05, so practicality has an impact on average risk perception.

In terms of convenience, some respondents made the following comments:
a) Convenience
b) It makes my life more convenient
c) It is rarely used
d) Generally
e) It brings great convenience. For example, electronic social security card, etc., I don’t have to take cards out of my house. It is very OK for young people.

As can be seen from the above comments, respondents generally have a positive attitude towards the convenience of smart Chengdu. This is the publics support for smart Chengdu, which shows that the implementation of smart Chengdu has effectively facilitated people’s lives in terms of livelihood.

Related summary Practicality and convenience

Practicality and convenience both influence the average perceived risk level, showing a positive correlation. Comparing the coefficients of their impact, it is evident that practicality has a greater effect on the average perceived risk level than convenience. Therefore, the digital governance practice platform Smart Chengdu can balance practicality and convenience with the average perceived risk level to optimize digital services for the public.

Risk perception of different groups

Different groups have different degrees of perception of different risks, with the highest degree of data sharing and the lowest degree of data consistency.

The average risk perception of enterprise workers, government department personnel, and students is highest when data sharing is the most common, and lowest when data consistency is the most important. This result may be related to the nature of the aforementioned work and learning units, where personnel may have more information to collect or be collected, and during the statistical period, they hope for more open and shared data to improve work efficiency.

The average risk perception of public institutions and public service institutions, as well as freelancers or flexible workers, is the highest for safety and the lowest for aging adaptability. The above units have strong flexibility in work, and they obviously pay more attention to privacy protection in the context of data openness.

The average risk perception of retirees is highest when it comes to data sharing and lowest when it comes to age-friendliness. This survey was limited to retired seniors who use mobile questionnaires on WeSurvey. Since these users already have basic skills in operating smartphones, and such samples are relatively small, accounting for only 4% of the total sample, the results may be biased from reality. However, this limitation also highlights an issue: retirees who can use mobile apps find digital life not too difficult. In contrast, retirees who fail to complete the questionnaire via their phones likely lack the ability to use smart city software. Therefore, age-friendliness should be given due attention.

Public support for digital governance

Through a survey on the publics familiarity with various applications of Smart Chengdu, we found that although the Smart Chengdu system is still under construction, it has already gained considerable influence. Smart Chengdu covers areas such as traffic management, smart communities, education, healthcare, emergency management, and more, totalling “6+8+N.” While there are differences in public familiarity across different application fields, every system mentioned in the questionnaire has been heard of or understood by the public and received generally positive or favourable evaluations. This indicates that digital governance methods are widely recognized and supported by the public, with promising prospects. Therefore, we need to analyse potential risks from different perspectives to promote its healthy development.

Public perception of different risks varies

Among the five listed risks, the perception of data sharing, algorithmic decision-making, and security is the highest. Regarding the frequent re-entry of basic information, we believe that effective barriers should be broken down between system back ends and departments to facilitate data circulation. Grasping citizens basic information or directly retrieving it from relevant departments with their consent can significantly improve platform efficiency, reduce the frequency of public re-entry of information, and enhance its accuracy. Due to the frequent entry of personal basic information across different platforms, there is a notable concern about the risk of privacy leakage. Therefore, system platforms, enterprises, and governments need to improve data management and confidentiality measures to reduce citizens anxiety about providing basic information. In terms of algorithmic decisionmaking, the data provided by platforms is a key factor in guiding government decisions. To improve decision accuracy, the precision of data must be enhanced; precise strikes require precise policies. To prevent algorithms from replacing human decision-making, government departments need to combine reality when making decisions. Algorithms serve as a basis but also require practical consideration to minimize the risk of algorithmic data hindering departmental decision-making, ensuring that algorithms do not become the masters of decisions. For Data Consistency is the least perceived risk by the public, which means that the data presented by different system platforms for the same information is not very different. Therefore, it also proves that most of the data provided by the smart Chengdu system are accurate and consistent, which can also bring great convenience to citizens daily life.

Different groups have different risk perception of smart Chengdu

Regarding the ranking of risk perception among different groups, we found that government departments have the lowest risk perception, followed by enterprises. This is because smart digital management systems are primarily led by government or enterprise entities, with significant involvement from government personnel and enterprise managers in research, design, application, and statistics. Therefore, they have more direct perception but pay less attention to the software’s usage effects, system details, and specific application scenarios. They mainly play the role of data statisticians and analysts rather than data entry operators or primary users of the platform, making it harder for them to detect risks in these systems. In contrast, retired individuals have the highest risk perception regarding smart Chengdu. This group mostly consists of people aged 60 and above, who have more free time in their lives and may be more meticulous in studying various apps, thus making it easier for them to identify risks in the system. This also reminds us to pay attention to the age-friendliness of smart systems, focusing on the special needs of elderly groups and improving the system platform, such as adding large font mode, information prompt navigation, and guidance modes, to provide guidance for digital operations for the elderly.

Public risk perception of smart Chengdu is affected by application context

After data measurement, we found that familiarity, practicality, and convenience are all positively correlated with risk perception, with the positive correlation between practicality and risk perception being the most significant. This indicates that citizens are more likely to perceive risks and identify issues in systems that are closer to their daily lives and frequently used. Therefore, while enhancing technical capabilities, platforms should pay greater attention to public feedback and promptly respond to improve the system, ensuring it truly serves the people.

In addition, we found that both government staff and other citizens are much more familiar with the health codes and venue codes of Tianfu Health Pass compared to other Smart Chengdu Apps. This demonstrates that the governments appeal is highly effective in emergencies, and special circumstances facilitate the promotion and operation of smart city systems. Therefore, the application of smart city systems should focus on reality, seize the tide of the times, and integrate digital technology with management in special and critical areas, implementing large-scale and powerful policies and management (Annex 1-6).

The survey focus on the usage of digital software for the “Tianfu Citizen Cloud” or “Citizen Cloud Services,” “Tianfu Easy-handling Platform” or “Tianfu Easy-enjoying Platform,” “Tianfu Health Code,” and “Venue Code,” as well as the “Smart City Spatiotemporal Big Data Platform,” “Electronic Social Security Card” or “Rong e HR”，”Smart Community” or “Smart Security Neighborhood,” “Smart Education” cloud platforms, “Cultural Tourism e Management” or “Cultural Tianfu,” “Smart Transportation” (TOCC), “Traffic Management 12123,” “12345” citizen hotline, “Smart Fire Rescue” or “Smart Emergency Response,” “Smart Urban Management,” and the “Park City” smart platform. A total of 413 data points were collected.

There are 411 valid samples in the sample. The survey area is Sichuan Province, including 280 samples in Chengdu and 131 samples in non-Chengdu.

questionnaire 1. Have you heard of “Tianfu Citizen Cloud” or “Citizen Cloud Service”? How often do you use it? How do you experience it?
2. Have you heard of “Tianfu Easy-handling Platform” or “Tianfu Easy-enjoying Platform”? How often do you use them? How is the experience?
3. Have you heard of the “Health Code” and “Place Code” of “Tianfu Health Code”? How often do you use them? How is the experience?
4. Have you heard of “Smart City Spatiotemporal Big Data Platform”? How often do you use it? How is the experience?
5. Have you heard of “electronic Social Security Card” or “Rong e HR”? How often do you use it? How is the experience?
6. Have you heard of “smart community” or “smart security community”? How often do you use it? How do you experience it?
7. Have you heard of the “Smart Education” cloud platform? How often do you use it? How is the experience?
8. Have you heard of “cultural tourism e-manager” or “Cultural Tianfu”? How often do you use it? How is the experience?
9. Have you heard of the “Transportation Intelligent Platform (TOCC)”? How often do you use it? How is the experience?
10. Have you heard of “Traffic Management 12123”? How often do you use it? How is the experience?
11. Have you heard of the “12345” citizen hotline? How often do you use it? How is your experience?
12. Have you heard of “smart fire rescue” or “smart emergency”? How often do you use it? How do you experience it?
13. Have you heard of “smart city management”? How often do you use it? How is the experience?
14. Have you heard of the “Park City” smart platform? How often do you use it? How is the experience?

Table 1:

Note: The above questions are cascaded questions. The drop-down options are as follows:

15. For the above platform/system/software, please answer the following questions according to your personal feelings: For you, is the function and interface clear and easy to use? 0= very difficult, 4= very easy
16. Does it bring convenience to your work and life? 0= very inconvenient, 4= very convenient
17. How friendly is it for special groups such as the elderly? 0= very friendly, 4= not very friendly
18. Do you have any concerns about your personal privacy? 0= No worry, 4= Very worried
19. Do you enter similar basic information repeatedly when you log in to different platforms/systems/software? 0=Never encountered, 4= Very frequent
20. Will there be significant differences between different platforms/systems/software? 0= no difference, 4= large difference
21. Do you think these digital platforms have the potential to influence or even manipulate the decision-making behaviour of the public sector? 0= impossible, 4= likely Note: Questions 15-21 are evaluation questions, and each option is assigned a corresponding score
22. What is your age range?
A. Under 20 years of age B. 20-40 years of age C. 41-60 years of age D. Over 60 years of age
23. What is your gender?
A. Male B. Female
24. What is the nature of your work/study unit?
A. Government departments B. Public institutions and public service agencies C. Enterprises D. Self-employed and flexible employment E. Students F. Retired.

Practicality

Through the screening of samples with “very satisfied” usage experience in questions 1-14, we explored their choices on question 15: “Is the function and interface clear and easy to operate for you?”

Using the IF function, it is not easy to assign values to different choices, such as =0, not easy to assign values to 1, generally =2, easy to assign values to 3, and very easy to assign values to 4.

Using the AVERAGE function, the average value of the corresponding value of each softwares 15 options was taken for all samples to obtain the practicality of each software.

Convenience

Method is the same as 1

Familiarity: The publics familiarity with the 14 applications of Smart Chengdu can be obtained through the questions “Have you heard of XX?” and “How often do you use it?”

Usage experience: The user’s familiarity with the usage experience can be obtained through the question “How is the usage experience?” and the usage experience can be used as a screening condition in the later data processing.
1. The average familiarity of the public with different software

The public is most familiar with the “Health Code” and “Venue Code” of Tianfu Health Code, while they are least familiar with Smart City Management. This reflects that during the special period of the pandemic, the public has fully utilized the software under Smart Chengdu. It can be seen that by using certain methods to increase demand for a particular software, the public’s familiarity with it can be enhanced.
2. Average familiarity of government departments with different software

Government workers were most familiar with the “health code” and “venue code” of Tianfu Health Code, and least familiar with “smart fire rescue” or “smart emergency”. Government workers were most familiar with some management software.

Let the degree of understanding k be the corresponding option (as follows) in question 1-14 of questionnaire 1-14, “Have you ever heard of XX software?” The corresponding value is 3,2,1,0 respectively.

Table 2:

Let u be the usage frequency of question 1-14 in the questionnaire “How often do you use it?” and the corresponding options (as follows) are 3,2,1,0 respectively.

Table 3:

The total sample size is n.
Average familiarity = sum (degree of understanding i * frequency of use i) / total sample size n
That is, average familiarity =Σk*u/n

There are 411 valid samples.
Gender: 59% male and 41% female (see chart below).

Age: 12% of the sample were under 20 years old, 52% were between 20 and 40 years old, 32% were between 41 and 60 years old, and 4% were over 60 years old.

Nature of work/unit: Government departments account for 5%, public institutions and public service institutions account for 20%, enterprises account for 37%, freelancers and flexible employment account for 16%, students account for 18%, retirees account for 4%.

a) Model description

A linear regression model was established to investigate the correlation between average familiarity, practicality and convenience and average risk perception.

b= (Σ Familiarity i * risk perception degree i-n * average familiarity * average risk perception degree)/ (Σ Familiarity i²-n * average familiarity ²)
A = average risk perception degree-b * average familiarity
y=b * familiarity + a

b) Model results

1. Average familiarity —— average risk perception

Table 4:

2. Practicality —— average risk perception degree

Table 5:

3. Convenience —— average risk perception

Table 6:

1. Yu Na (2022) Research on the Evaluation of Smart City Construction Level in Beijing. Beijing University of Civil Engineering and Architecture.
2. Lyu Xiaofan (2022) The Impact of Smart City Construction on Public Service Efficiency. Southwestern University of Finance and Economic.
3. Hu Xingchen (2022) Research on the Impact of Smart City Construction on High-Quality Economic Development of Cities. Sichuan University.
4. Deng Qi (2022) Research on the Problems and Countermeasures in the Construction of Smart City in Chengdu. Sichuan University.
5. Sun Liying (2022) Research on Performance Evaluation of New Smart City Construction from the Perspective of Public Perception[D]. Qingdao University of Technology.
6. Jiao Manman (2022) Influencing Factors and Optimization Strategies of Public Participation in the Construction of Smart City in Xuzhou. China University of Mining and Technology.
7. Dang Xifeng (2023) The Impact of Smart City Construction on the Economic Development of Megacities. Cooperative Economy & Science (05): 8-11.
8. (2022) General Office of the Chengdu Municipal Committee of the Communist Party of China. Notice on Issuing the Implementation Plan for the Construction of the Smart Chengdu Operation and Management Platform. Chengdu, Chengwei Ting pp. 14.
9. (2022) Office of the Smart Chengdu Construction Leading Group, Chengdu. The 14th Five-Year Plan for New Smart City Construction in Chengdu (Draft for Review). Chengdu.
10. (2021) People's Government of Sichuan Province, Notice on Issuing the "14th Five-Year Plan for Digital Economic Development in Sichuan Province". Chengdu: Chuanfu Fa 32.
11. Wenzelburger G, König PD, Felfeli J, Achtziger A (2022) Algorithms in the public sector. Why context matters. Public Administration 102(1-21).