Innovation and traffic safety: analysis of four Norwegian county authorities

Background

Innovation has proven to be highly important in the field of traffic safety (Belin et al., 2012).Fagerberg (2005) defines innovation as an idea for a new product or a process that is implemented in a specific context. The context is key to defining innovation. To count as an innovation, it is sufficient for the idea, practice, or object to be perceived as new by the implementer, even if others have already implemented it (Rogers, 2003). Innovation is also crucial in the traffic safety management of public entities. Previous studies show that Vision Zero, which is a radical innovation in traffic safety, has contributed to Norway and Sweden having the lowest number of traffic fatalities in the world (Belin et al., 2012; Craens et al., 2022). Innovation is a crucial component of Vision Zero and Safe System (Elvik et al., 2023).

International and Norwegian studies also demonstrate that the regional governance level is fundamental in traffic safety work in countries with high levels of traffic safety (Belin et al., 2012; Craens et al., 2022; ITF OECD, 2022; Krogstad, 2020). In Norway, county authorities play a key role in the national traffic safety work, because they serve as the organizational intermediary between the state and municipalities. County authorities are responsible for recommending and coordinating traffic safety measures within the county. This entails coordinating the traffic safety measures of municipalities and collaborating with regional actors such as the Norwegian council for traffic safety (Trygg Trafikk), the police, and various other organizations. County authorities also function as regional planning authorities, school owners, responsible for public transport and public health, and procurers of public transport services (Nævestad et al., 2023). They are also major employers with staff who commute for work. All these roles provide opportunities for comprehensive traffic safety management.

The role of county authorities in Norwegian traffic safety work has, however, undergone significant changes in recent years. Through the regional reform, which came into effect on 1 January 2020, planned to be completed by 2024, the number of counties in Norway was reduced from 19 to 11. This involved reorganization and merging processes of old structures, groups, and personnel in 13 of the original county authorities. At the same time, county authorities took over the administrative responsibility for all county roads, which accounts for approximately half of the public road network in Norway, previously managed by the national public roads administration (NPRA). This responsibility includes planning, construction, management, operation, maintenance, and traffic safety work. This responsibility is significant for traffic safety, as contributing accident factors related to road conditions or the road environment were identified in 43% of fatal accidents during the period 2017–2020 (Hesjevoll et al., 2022). Additionally, the county road network has the highest accident risk among the road networks (Statens vegvesen, 2022) and faces significant maintenance backlog.

In January 2020, seven new county authorities were thus faced with a range of challenges in developing new merged administrations, and in developing new systems to manage county roads. At the same time, the county authorities were required to maintain their key functions in the field of traffic safety. These changes involved new compositions of personnel, organizations, routines, etc. The new county authorities were formed by merging old county authorities, while also incorporating many individuals who had previously worked with county roads in the NPRA. These are change processes that likely require a significant amount of time and energy. At the same time, we can imagine that such changes can provide opportunities for learning and innovations that have the potential to enhance traffic safety, for example, in the development of new systems, methods, and routines for traffic safety management. Given the importance of these different mechanisms, it is important to study the status of the county authorities' traffic safety management three years after the regional reform.

Recent research shows that innovation is a key premise for traffic safety improvement in Vision Zero and the Safe System approach (Elvik et al., 2023). This research does, however, not examine the organisational and cultural conditions which allow traffic safety innovation to take place. This is the topic of the present paper. Defining the conditions facilitating and impeding traffic safety innovation is key to continued progress in traffic safety management within the Safe system approach.

Previous research shows that there is a relationship between innovation culture and self-reported innovation outcomes in organizations (Bower et al., 2003; Mathisen et al., 2008; Newman et al., 2020; West & Anderson, 1996). Therefore, we assume that there is a relationship between a positive innovation culture in the county authorities and innovations and learning in traffic safety work (Hypothesis 1) .

Previous research also shows that reforms and organizational changes can have consequences for innovation culture. Innovation culture refers to the extent to which members of organizations perceive that the organization encourages, supports, and enables innovation (Newman et al., 2020). Reforms and reorganizations often lead to new compositions of people, new communication networks, and new patterns of interaction (Eisenhardt & Tabrizi, 1995; Karim, 2009). This can create a new climate for learning and innovation by bringing together new perspectives and challenging them in ways that foster the emergence of new ideas (Gherardi et al., 1998; Karim, 2009; Meyer & Marais, 2009).Newman et al. (2020) refer to innovation climate as shared perceptions at the group or organizational level regarding the extent to which group or organizational processes encourage and enable innovation. The changes associated with the regional reform and the dissolution of the common road administration particularly involve new groups starting to work in the same organization, including individuals from merged county authorities and those who previously worked at in the NPRA. Additionally, the new county authorities were mandated to create new organizations, systems, and routines. Based on previous research, we assume that the new compositions of people and professional groups have created a new climate for learning and innovation (Hypothesis 2) in the county authorities, three years after the regional reform.

However, it should also be noted that if reorganizations are too extensive or energy-consuming, they do not lead to learning and innovation but rather to dissatisfaction, uncertainty, and organizations being ill-equipped to manage their key tasks (Karim, 2009; Meyer & Marais, 2009) (Hypothesis 3). Therefore, it is not guaranteed that the county authorities have developed a strong innovation culture.

Previous studies show that learning is one of the key elements of a strong safety culture (Pidgeon & O’leary, 2000; Reason, 1997). Therefore, we assume that the county authorities' innovation culture is strongly related to their safety culture (Hypothesis 4) .

Aims

In this section, we present the aims of the study. Based on what has been described above, this study aims to examine:

• Innovation outcomes in the county authorities' traffic safety work

• Innovation culture in the county authorities

• Factors influencing innovation outcomes in the county authorities

• Factors influencing innovation culture in the county authorities.

Innovation in traffic safety work

While invention refers to the emergence of a new idea, innovation is about putting that idea into practice. Innovation is not just about inventing something new; it can also involve adopting something in a new context or organization (Fagerberg, 2005). Introducing a new traffic safety measure in one country, inspired by something similar in another country, is also an innovation.Elvik et al. (2023) analyze the significance of innovation in the traffic safety management of public actors in a study of Norwegian traffic safety plans over the past 20 years. Some of the most important innovations related to the Vision Zero concept are conceptual innovations, which involve new ways of seeing the world, new perspectives, or purposes for traffic safety work in general. Vision Zero represented a new approach to responsibility, aiming to create a road system where road users are not killed or seriously injured if they comply with applicable laws and regulations (Belin et al., 2012; Elvik, 2022). This also assigns significant responsibility to road authorities and owners. It should also be noted that after 10 years of Vision Zero in Norway, in 2010, new interim goals for the number of fatalities and serious injuries were set on the path to Vision Zero. This occurred in both Norway and Sweden and is linked to further reductions in the number of fatalities and serious injuries in both countries (Elvik et al., 2023; Varhelyi, 2016). In the first 10 years of Vision Zero, it was considered immoral to have a goal for fatalities and serious injuries that was not zero. In the work on Vision Zero, innovations related to safety indicators were also developed, which were monitored annually. As the number of fatalities and serious injuries decreased, indicators related to road users (speed, alcohol impairment, seatbelt use), road infrastructure (kilometers with median barriers, traffic safety inspections), and vehicles (age of the vehicle fleet, percentage of vehicles with full Euro NCAP score) were developed (Elvik et al., 2023).Elvik et al. (2023) conclude that these gradual innovations in the continuous improvement of traffic safety work in Norway have contributed to fewer fatalities and serious injuries in traffic.

Innovation culture

If we accept that innovation is important for traffic safety work, it is crucial to understand what promotes (and hinders) innovation. Several studies have identified strong relationships between innovation culture and self-reported innovation outcomes in their own organizations (Bower et al., 2003; Mathisen et al., 2008; Newman et al., 2020; West & Anderson, 1996).

In this study, we define innovation culture as aspects of organizational culture that are relevant to innovation.Schein (1992) defines organizational culture as: ‘A set of shared basic assumptions that a group has learned while solving external/internal problems and that has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems’. ‘Basic assumptions’ refer to whatSchein (1992) calls the deepest cultural level of ‘taken-for-granted’ assumptions that govern what we pay attention to, what things mean, how we react emotionally, and how we act. Organizational culture and organizational climate refer to different analytical levels related to the same empirical phenomenon. Organizational climate represents snapshots, or manifestations, of organizational culture. The climate concept is often used to describe quantitative measurements of culture in organizations (Flin et al., 2000). However, the terms organizational climate and culture are often used interchangeably. In this study, we adopt the same approach.

Innovation culture is about the extent to which members of organizations perceive that the organization encourages and enables innovation (Newman et al., 2020). This may involve specific encouragement of innovation from managers and colleagues, autonomy and freedom, as well as access to necessary resources such as expertise, materials, information, etc.

Measurements of innovation culture and innovation climate typically focus on the following characteristics of the organization:

• Facilitation and emphasis on innovation among managers and employees, i.e. the extent to which employees perceive the organization as open to change, supportive of new ideas, and tolerant of diversity among members.

• Availability of resources, i.e. the degree to which resources are perceived as sufficient within the organization (Newman et al., 2020).

Regarding factors that influence innovation culture, previous research indicates that managers play a particularly important role. Leaders can foster a culture of innovation by encouraging and rewarding creativity, implementing new ideas, establishing expectations (Anderson & West, 1998). Additionally, managers can facilitate innovation by influencing organizational structure, group composition, problem-solving routines, resource allocation, competence, and granting autonomy, among other factors (Keller, 2001; Puccio et al., 2008; Somech, 2006; Tierney et al., 1999).

Innovation culture is influenced by organizational structure (Zennouche et al., 2014). High levels of centralization and strict hierarchy are seen as hindrances to learning and innovation, as they limit information flow and communication. Conversely, decentralization fosters greater participation and facilitates the exchange of multiple perspectives in creative ideation processes. However, excessive decentralization may impede coordination and the facilitation of systematic learning across different actors. The concept of ‘balanced empowerment’ reflects this balance between top-down and bottom-up dynamics in innovation processes (Sundbo, 1996). Additionally, the availability of resources is positively related to innovation since resources are needed to develop new ideas (Zennouche et al., 2014).

As mentioned, organizational changes and reforms that lead to new compositions of people and new patterns of interaction can impact innovation culture and outcomes (Eisenhardt & Tabrizi, 1995; Karim, 2009). Studies of various interventions for organizational learning show that learning and cultural change often occur when different individuals' opinions, interpretations, and experiences of practices are confronted (Richter & Koch, 2004). This happens through comparisons of perspectives and reflection on one's own practices (Gherardi et al., 1998; Richter & Koch, 2004). If the organizational changes are too extensive or energy-consuming, they do not lead to learning and innovation but rather dissatisfaction, uncertainty, and organizations becoming less equipped to manage their key tasks (Karim, 2009; Meyer & Marais, 2009).

The innovation culture of the county authorities will also be conditioned by and influence the interaction with the government on one side and the municipalities on the other. Such a broader approach to understanding innovation and learning culture beyond the framework of individual organizations reflects the insight that innovation is systemic. The research literature on innovation systems emphasizes that innovation cannot be seen as an isolated phenomenon but rather arises through the interaction between different types of actors (Edquist, 1997; Klein Woolthuis et al., 2005; Lundvall, 1992). Central to such a systemic understanding is the innovation potential that lies in the interaction and feedback between actors possessing different competencies and complementary perspectives. Such a systemic understanding has also been further developed and applied at a regional level, where the focus has been on how interaction and collaboration among various actors take place within a specific regional context, with its respective actors, networks, and institutional and infrastructural characteristics (Asheim & Isaksen, 2002; Cooke, 2004; Maskell & Malmberg, 1999).

Safety culture

In addition to the factors mentioned above, innovation culture in organizations is also influenced by safety culture. Safety culture can be understood as a focused aspect of organizational culture (Antonsen, 2009; Guldenmund, 2000; Haukelid, 2008; Nævestad, 2010). In line with this, safety culture can be defined as safety relevant characteristics of organizational culture (Antonsen, 2009; Nævestad, 2010). More specifically, we define it as shared and safety-relevant ways of thinking or acting that are (re)constructed through negotiations among people in social contexts (Nævestad, 2010).

There has been extensive research on the characteristics of good safety cultures. Leadership commitment to safety, employee engagement in safety, mutual trust, and reporting are often emphasized as core factors (Nævestad, 2010).Reason (1997) has identified five essential aspects that characterize organizations with a good safety culture:

• Informed culture: The organization gathers data on both accidents and incidents (near-misses) and conducts proactive measures such as safety audits and safety climate surveys.

• Reporting culture: All employees report incidents and near-misses and participate in safety climate surveys, among other reporting activities.

• Just culture: Employees are encouraged to report incidents, as they trust that the management will handle incident reports and implicated individuals in a fair manner.

• Flexible culture: The organization has the ability to adapt and change its practices

• Learning culture: The organization has the ability to learn from reported incidents, safety audits, etc., in order to improve safety.

Most studies on safety culture and safety management systems are based on this understanding of safety culture. Several studies from various sectors find correlations between different aspects of safety culture (such as management and employee commitment to safety, learning, reporting) and other safety outcomes (accidents, behavior, incidents, violations) in aviation (Hudson, 2003; Patankar, 2019), railway (Zuschlag et al., 2016), maritime sector (Lappalainen et al., 2012), and the road sector (Naveh & Katz-Navon, 2015). However, it is challenging to determine which specific safety culture traits have the greatest impact on safety. It is often emphasized that management commitment to safety is crucial, combined with a systematic and group learning process that involves identifying potential hazards, implementing measures, and following up on the process (Nævestad et al., 2018). This aligns with whatReason (1997) refers to as a learning culture.

Research traditions study different sub aspects of organizational culture, such as innovation culture (Newman et al., 2020) and (learning) safety culture (Antonsen, 2009). Since these sub aspects of organizational culture appear to be relatively overlapping, we can expect relationships between different sub aspects of organizational culture, such as innovation culture and safety culture. It is not surprising, for example, if organizations that have a learning safety culture (Reason, 1997) also have a strong innovation culture (Newman et al., 2020). However, as far as we know, there are no studies examining the relationship between innovation culture and safety culture. The research on innovation culture and safety culture does not appear to overlap.

Quantitative survey

Qualitative interviews

Characteristics of the respondents

In Table 2, we show a comprehensive overview of the respondents. Although there are some variations regarding age and gender, their differences are not statistically significant. When it comes to respondents' job position level, County authority 4 stands out with a relatively low number of employees among its respondents (58%). In the other county authorities, the proportions of employees range between 77% and 88%. The differences are statistically significant at the 1% level (p = 0.003). Approximately two-thirds of the respondents were hired in 2020. The exception is County authority 4, where over one-third of the respondents were hired in 2022. We observe that approximately 60% of the respondents hired in 2020 came from the NPRA before being employed in the studied county authorities in 2020. These are largely people working with county roads, who chose to switch employer to the county authority as the county authority took over the administrative responsibility for the county roads. Before the regional reform in 2020, the four county authorities were comprised of eight county authorities. Only 8% of the respondents worked in one of these before 2020. We have standardized the departments sub-divisions across county authorities, for these to be comparable across the counties.

Innovation and learning in traffic safety work

This section is related to the first aim of the study, which is to examine innovation outcomes in the county authorities' traffic safety work. First, we present results from the quantitative survey and then results from the qualitative interviews.

Results from the survey

To provide an overview, we present the results from the survey as figures displaying the share of respondents agreeing to the different statements. Figure 1 shows the percentage of respondents who agree and strongly agree with statements in each county authority regarding items listed under innovation and learning. These were introduced with: ‘Have you used the changes as an opportunity’ to e.g ‘introduce new ways to collaborate with municipalities in traffic safety work’, introduce new methods in traffic safety work, or general statements about learning, like: ‘We have a more holistic perspective on traffic safety than before’, ‘I have gained a better understanding of how the county authority's work can influence traffic safety’. See Appendix A for the full wording of the questions and distribution of responses for each question.

Based on Chi-square tests, we observe statistically significant differences between the county authorities regarding the various statements. County authority 1 consistently has the highest proportion of respondents who agreed with the statements, while County authority 3 has the lowest proportion of respondents who agreed. The differences between the county authorities are statistically significant at the 10% level for the first statement, at the 1% level for the statements about new methods and systems, and at the 5% level for the statement regarding new collaboration with municipalities. However, the differences in the proportions of the county authorities for the statement about collaboration with external parties are not statistically significant. Finally, the differences between the county authorities for the last two statements about learning are statistically significant at the 1% level. The highest proportion of respondents agree with the last two statements in County authority 1, and the lowest in County authority 3.

Comparing statements across county authorities, we notice that there are more respondents who agree that they have used the reform as an opportunity to reorganize the departments working with traffic safety, and the second highest number of respondents agree with the statement that they have gained a better understanding of how the county authority's work can impact traffic safety.

We created a sum score index for innovation in traffic safety work (min: 7, max: 35 points). The results on the index are presented in Table 3.

Table 3: The four county authorities' scores on the index for innovation and learning in traffic safety work (min: 7 points, max: 35 points)

County authority	Mean	N	Standard deviation
County authority 1	18.9	44	3.9
County authority 2	17.5	64	4.3
County authority 3	14.4	34	4.1
County authority 4	17.9	20	2.9
TOTAL	17.3	162	4.3

In line with the results from the Figure 1, we can see that County authority 1 has the highest score on the index, while County authority 3 has the lowest. The differences between the county authorities are statistically significant at the 1% level (p < .001).

Innovation culture in the county municipalities

This section is related to the second aim of the study, which is to examine innovation culture in the county authorities. First, we present results from the quantitative survey and then results from the qualitative interviews.

Factors influencing innovation culture and innovation

In this section, we focus on factors influencing innovation outcomes in the county authorities (third aim), and factors influencing innovation culture in the county authorities (fourth aim), based on qualitative data. We organize the below text based on themes reported in the interview.

The organizational changes have required significant time and energy. Interviewees in the merged county authorities experienced that the process of merging required a considerable amount of time and energy. They emphasized that the counties and the merged groups had different routines, systems, cultures, organizational structures, and collaboration methods with municipalities and other external entities. All of these aspects needed to be combined while creating new organizations, and decisions had to be made on how to proceed. Designing new common organizational structures and routines that catered to all needs as much as possible demanded a lot of time and energy. One interviewee mentioned that the changes and reorganization took away time and resources that could have been used for innovation:

‘Much of the time we used for innovation and development is now spent on reorganization. (...) There are some exciting technologies coming up, such as inexpensive sensors at critical locations that can be implemented in a management system. We started looking into it. Vehicle sensor technology to assess road conditions. (...) These projects were put on hold when we have to start reorganizing.’ (Interviewee, County Authority 3).

Split up and reorganization fatigue. Interviewees who previously worked in the NPRA expressed that they are tired of ongoing reorganizations that have taken place over many years. First, they spent their two last years in the NPRA preparing to change workplaces before the regional reform (2018–2020). Then, they had two years in the county authority (2020–2022) while working to establish a new organization. In 2022, they faced two more years in the county authority (2022–2024) while waiting for the split-up, and the new county authority where they will work. Reversals of the mergers (i.e. the split-up), which were decided in some county authorities, led to reduced job satisfaction. The reversals resulted in two more years of reorganization processes, which diverted attention away from regular work tasks. Most interviewees had negative views about the reversals. This was because specialized working environments, which were initially small, were now expected to become even smaller when newly merged county authorities were split up into more county authorities. They anticipated that the most significant consequences would be within operations and maintenance. Even among employees who did not come from the NPRA, there were fears that the specialized environments would become smaller, leading to a loss of the established expert communities that were created after the mergers and the regional reform.

COVID-19. Many mentioned the COVID-19 pandemic as a central barrier in establishing and coordinating the new county authority organizations. The fact that employees were working from home and couldn't meet physically was noted as a major challenge in the process of setting up new organizations, fostering good relationships, and promoting collaboration among new colleagues.

Resources: financial and expertise. An essential theme that emerged from the interviews was that, along with the responsibility for county roads county authorities also inherited a significant backlog in the maintenance of the county road network. Several interviewees mentioned that there is no room in the budget to carry out the necessary improvements, and there are also sudden demands for repairs. This can impact the construction of school roads, pedestrian pathways, and other transportation plans. Several interviewees noted that there are now smaller specialized environments and thinner staffing among those working on road maintenance and administration of the county roads, and they have lost several support functions that they had in the NPRA.

Relationship with regional and local politicians. For the interviewees working with county roads, it a new experience to have a more active engagement with county politicians. Several of them also experienced a different type of interaction with municipal politicians than before. Some mentioned examples of municipal politicians who contacted county politicians and/or the media to advocate for their issues and influence decisions, particularly related to specific road measures (speed limits, speed bumps, etc.). Many stated that this takes time and resources. On the other hand, some suggested that this contact could have positive consequences because regional politicians have resources and decision-making authority. These experienced changes can be attributed to the fact that the county authority is a political entity, whereas the NPRA is not.

The size of the new county authorities. Interviewees from County authority 2 explained that the size of the county authority and the number of municipalities posed a challenge that influenced the quality of traffic safety work. They described the difficulty of maintaining good communication with all municipalities.

Multivariate analyses

In this section, we focus on the third and the fourth aim, respectively, which are to examine factors influencing innovation outcomes and innovation culture in the county authorities. We examine this based on the quantitative data, in multivariate regression analyses.

Innovation results and innovation culture

The first and second objectives of the study were to examine innovation results and innovation culture in the traffic safety work of the county authorities. The county authorities scored moderate on the index for innovation culture and the index for traffic safety innovations. The average score on the index for innovation culture was 3.3 for all questions, indicating slightly better than ‘neither agree nor disagree’ on average. The average score on the index for traffic safety innovations corresponds to an average of 2.5 for all questions, which is between ‘disagree’ and ‘neither agree nor disagree’. These scores clearly suggest opportunities for improvement.

Despite relatively modest scores on the index for traffic safety innovations, we found several examples of traffic safety innovations in the interviews. County authority 1 had implemented a new method for risk assessment of curves on county roads. Some interviewees mentioned new and better adapted quality systems for the administration of county roads. Interviewees from County authority 2 talked about new ways of collaborating with municipalities and other external partners in traffic safety work. They also mentioned several conceptual innovations, i.e. new principles for traffic safety work. We found most traffic safety innovations in County authority 2, and the interviewees reported a strong innovation culture in this county authority after the implementation of the regional reform.

The results from the survey seem to be somewhat in contrast to this: County authority 2 did not score highest on either innovation culture or traffic safety innovations. One possible reason for the discrepancy between qualitative and quantitative data could be the time aspect. The interviewees in the interviews who talked about the innovation culture in County authority 2 primarily referred to the period immediately after the regional reform, while the survey questions measuring innovation culture were distributed three years after the reform and did not have the same retrospective perspective. Additionally, County authority 2 will be split up again into three new county authorities in 2024—a decision which has had a significant negative impact on the innovation culture.

The most consistent innovation across all county authorities was that the interviewees reported a more comprehensive and holistic perspective on traffic safety. They attributed this to the co-organization with individuals who previously worked in the merged county authorities and those who came from the NPRA.

These findings reflect an innovation system perspective, i.e. how the composition of actors with different expertise and perspectives can complement each other and thus be a central prerequisite for mutual learning and innovation in the field of traffic safety (Asheim & Isaksen, 2002; Cooke, 2004; Maskell & Malmberg, 1999). Previous studies also emphasize the importance of analyzing the regional level from an innovation system perspective (Edquist, 1997; Klein Woolthuis et al., 2005; Lundvall, 1992). The results of our study on learning effects from the co-organization of different professional groups illustrate the significance of facilitating continuous interaction and learning across actors and perspectives.

The third objective of the study was to examine factors influencing innovation results in the county authorities. We hypothesized that there is a relationship between the county authorities' innovation culture and their innovation results in traffic safety work (Hypothesis 1). We conducted multivariate analyses to test this hypothesis while controlling for several contextual factors. The results show that the most significant variable influencing innovation results is innovation culture. This is consistent with previous research. Several studies have found strong correlations between innovation culture and respondents' self-reported innovation outcomes in their organizations (Bower et al., 2003; Mathisen et al., 2008; Newman et al., 2020; West & Anderson, 1996). Another key variable influencing both innovation results and innovation culture was the top management's focus on traffic safety. Leadership focus is crucial in both innovation culture research (Anderson and West, 1998) and safety culture research (Flin et al., 2000). We did not find that respondents' perception of the status of available resources in traffic safety work contributed significantly, which contradicts previous research (Keller, 2001; Puccio et al., 2008; Somech, 2006; Tierney et al., 1999).

Factors influencing innovation culture

The fourth objective of the study was to examine factors influencing innovation culture in the county authorities. Previous research indicates that reforms and organizational changes that lead to new compositions of people, new communication networks, and new patterns of interaction can create a new climate for learning and innovation, allowing new perspectives to meet and confront in ways that give rise to new ideas (Eisenhardt & Tabrizi, 1995; Gherardi et al., 1998; Karim, 2009; Meyer & Marais, 2009). The changes related to the regional reform and the dissolution of the shared road administration particularly involve new groups starting to work in the same organization, i.e. individuals from merged county authorities and individuals who previously worked in the NPRA. Additionally, the new county authorities were mandated to create new organizations, systems, and routines.

Therefore, our hypothesis 2 suggested that the new compositions of people and professional groups have created a new climate for learning and innovation in the county authorities, three years after the regional reform. The quantitative data do not support Hypothesis 2. The scores on the index for innovation culture were moderate in the studied county authorities. However, the qualitative data support Hypothesis 2. This is particularly true for the interviewees from County authority 2, who described a strong innovation climate after the regional reform. They attributed this to the regional reform creating an opportunity to build something new, and this was seen as an opportunity to implement new solutions.

Another reason why the qualitative data support Hypothesis 2, is that the most consistent innovation in the study is precisely due to new compositions of people, new communication networks, and new patterns of interaction (Eisenhardt & Tabrizi, 1995; Karim, 2009). The most pervasive innovation across all county authorities is a more holistic perspective on traffic safety. This is an example of a conceptual innovation that emerged from the co-organization of individuals who worked in the merged county authorities with individuals who came from the NPRA. This new composition of people created new communication networks and new interaction patterns, leading to new perspectives on traffic safety (Gherardi et al., 1998; Karim, 2009; Meyer & Marais, 2009). It appears that this conceptual innovation occurred to a lesser extent in County authority 3, where there was less co-organization of individuals from the merged county authorities and individuals from the NPRA.

Conceptual innovations provide new perspectives, purposes, and views of the world that challenge and replace previous perspectives. This can also be linked to the concept of a paradigm shift. It involves seeing a new phenomenon not seen before (framing) or seeing an existing phenomenon in a new light and understanding it in a new way (reframing) (Berg, 1985). Framing and reframing create new frames of reference that underlie new interpretations and new actions. Framing and reframing a phenomenon are essential as they motivate and legitimize new strategies and actions.

Both the quantitative and qualitative data support Hypothesis 3, indicating that if reorganizations become too extensive or energy-consuming, they do not lead to learning and innovation but rather to dissatisfaction, ambiguity, and organizations becoming less equipped to manage their key tasks (Karim, 2009; Meyer & Marais, 2009). The qualitative data show that the following factors impeded the innovation climate in the county authorities: the reorganizations requiring significant time, energy, and resources, fragmentation, and organizational fatigue. Additionally, the impact of COVID-19 was mentioned. In the multivariate analyses, we observed that the index measuring whether the reorganizations demanded significant time, energy, and resources significantly and negatively influenced the innovation culture. Simultaneously, we found a positive relationship between top management's focus on traffic safety and innovation culture, which is in line with previous research (Anderson & West, 1998; Flin et al., 2000).

Innovation culture and safety culture

We hypothesized that the county authorities' innovation culture would be strongly related to their safety culture. Our results support this hypothesis. In the multivariate analysis of factors influencing innovation culture, safety culture was the variable with the most substantial impact. The strong correlation we find between safety culture and innovation culture is likely due to several factors.

First, learning is a key characteristic of a good safety culture.Reason (1997) defines a good safety culture, among other things, as a learning culture where the organization can learn from reported incidents, safety reviews, etc., to improve safety. In line with this, most studies of safety culture understand it as strong management engagement in safety combined with a systematic and group-based learning process, where potential hazards are identified, measures are implemented, and the process is followed up (Hudson, 2003; Lappalainen et al., 2012; Naveh & Katz-Navon, 2015; Patankar, 2019; Zuschlag et al., 2016). Organizational learning is closely linked to innovation, because the part of organizational learning that involves creating new knowledge is often a prerequisite for innovation. However, new ideas also need to be implemented for something to be called innovation.

Second, both the indexes for innovation culture and safety culture measure employee involvement. Both good innovation culture and good safety culture are characterized by encouraging and rewarding employees' initiatives. In a good or learning safety culture, it is crucial for employees to be vigilant and report relevant information within the organization (Reason, 1997). Similarly, a positive innovation culture is characterized by employees taking the initiative for new solutions (Newman et al., 2020).

Third, effective management is a crucial variable that can influence both good safety culture and good innovation culture. This involves leaders signaling to employees that they desire input and involvement, rewarding initiatives and suggestions, and facilitating open communication.

Despite finding a strong correlation between innovation culture and safety culture, which is also theoretically expected, there are to our knowledge no studies that examine the relationship between innovation culture and safety culture. The results indicate that innovation culture should be understood as an element of a learning safety culture. Since measurements and studies of safety culture do not focus on innovation in the traffic safety domain, this should be included in future studies.

Questions for future research

Practical implications: what can we learn from the results?

In this study, we have examined traffic safety innovations, innovation culture, and safety culture in four Norwegian county authorities, with the regional reform as the backdrop. The background is that in 2020, the county authorities had to establish new organizations, routines, and systems. The starting point of the study was that these changes could offer opportunities for new ways of thinking and innovation in traffic safety work, while also demanding significant time, energy, and resources. We studied the county authorities three years after the reform came into effect and found considerable differences between them.

Our results suggest that some county authorities have performed better than others in the changes we studied. County authority 1 appears to have performed the best, while County authority 3 consistently scores lower on most outcome measures. Our findings indicate that reorganizations that require substantial time, energy, and resources are negatively related to innovation culture and traffic safety innovations. Lengthy change processes (and reversals of reforms) are likely to be particularly challenging.

On the other hand, a strong management focus on traffic safety and a good safety culture are associated with innovation outcomes and innovation culture, even in county authorities that have undergone substantial changes and reorganizations. These factors appear to mitigate the negative effects of reorganizations that demand considerable time, energy, and resources.

A very relevant question, that is evoked by our results is whether it is necessary to re-organise in order to facilitate organisational learning, and new and broader perspectives on traffic safety. Although we have found that this has happened, especially in some of the studied county authorities, we should not use the results of the study to recommend re-organisation to facilitate learning and new perspectives. The purpose of the reorganization related to the regional reform in Norway, was primarily political (i.e. to make larger counties and give them more tasks) and not primarily to contribute to learning (although benefits related to giving county authorities a comprehensive responsibility for traffic safety also was mentioned, e.g. a more holistic approach to traffic safety management). Nevertheless, we find that learning and innovation has occurred, especially in some county authorities. Our study shows, however, that the learning and new perspectives have come at considerable costs, related to time, energy, and resources. Some of the county authorities have experienced higher costs in this respect, and fewer gains (e.g. county authority 3). The reorganization has also delayed the maintenance of key functions related to traffic safety for some time (Nævestad et al., 2023). The reversal of the reform in some counties (County authority 2 and 3) was also political, and has induced more costs on the personnel working with traffic safety in these county authorities. Our study has, however, pointed to some factors that may influence negative and positive results of reorganization. To conclude this discussion, one of the main positive mechanisms related to learning that has been highlighted by this study is that new compositions of people, new communication networks, and new patterns of interaction can create a new climate for learning and innovation by bringing together new perspectives and challenging them in ways that foster the emergence of new ideas (Gherardi et al., 1998; Karim, 2009; Meyer & Marais, 2009; Richter & Koch, 2004). It is not unreasonable to expect that this also can happen by bringing people from different organizations, scientific disciplines and with different work tasks etc. together in different types of cooperation and communication arrangements (i.e. not through permanent reorganization). This is indicated by previous research (Gherardi et al., 1998; Richter & Koch, 2004). On the other hand, results also indicate that the implementation of some innovations (e.g. new methods and systems) seem to require new organisations, presumably as reorganization provides the opportunity for the fulfilment of new ideas. This points to interesting questions for future research: which innovation require reorganisations (if so: how much?), and which do not?

Methodological limitations

We have not included all county authorities in Norway. In this study, we only focus on four county authorities, and we do not know how representative they are of all the Norwegian county authorities. However, it should be noted that these four county authorities are diverse. For instance, we have a large county authority, one that was not merged with any other county authority in 2020, and two county authorities that are planned to be split up into (reversing the 2020 reform). Nevertheless, to obtain a comprehensive picture of the situation in Norway, we would need to include more county authorities.

We cannot assume that all respondents work directly and indirectly with traffic safety, and this may influence their level of knowledge about all the questions regarding traffic safety. There are considerable proportions of respondents who have answered ‘do not know/not relevant’ to the questions about traffic safety innovations (up to around 30% for some questions). We have excluded these respondents from the analyses of the shares indicating agreement with the statements about traffic safety innovations, and in the multivariate analyses of the factors influencing this. The reason for calculating the shares agreeing after excluding those who answered ‘do not know’ is to focus on the proportions agreeing with the statements among those who actually have knowledge of and/or an opinion about this. Our interest is not in measuring the degree of knowledge about innovations, but the extent of innovations. We cannot assume that all respondents have knowledge about all these statements. By removing the ‘do not know’ responses, we can also reduce the effects of potential biases among the respondents in the county authorities. The consequence of filtering out those who answered ‘do not know’ is that the number of participants included in these analyses may become relatively small, as they only consist of those who did not answer ‘do not know’ for any of the questions involved in the index. Moreover, it is also important to discuss what it means to work with traffic safety. Many of the county authorities' employees work on matters that have indirect implications for traffic safety (e.g. purchasing public transportation, public health, zoning plans, etc.), and it is important to include them in the study as well. One of the main results of our study is that the people working indirectly with traffic safety also learned that their work has implications for traffic safety.

The samples include many individuals coming from the NPRA. Approximately 60% of all participants in this study are from the NPRA, and most of them were employed by the county authorities in 2020. In total, 65% were employed in 2020, while 17% were employed before that. This suggests that the participants may be somewhat skewed, but this does not necessarily have to be the case. The number of participants working with recommending and coordinating traffic safety measures in the county authorities is not very large compared to the number working with county roads and coming from the NPRA. Based on information from some of our contact persons in the county authorities, the proportion of former employees from the NPRA should have been even greater in the survey, because the number of employees who work with county roads (and who come from NPRA) is several times higher than the number of employees who work on recommending and coordinating traffic safety measures (for example 200 against 15 employees).