In the small scale-survey we commented on several unexpected results when we compared the Norwegian drivers, the drivers from WE, the CEE drivers and the drivers in the Norwegian II sample. These differences were for instance related to the safety commitment of managers and colleagues, self- reported accidents and self- reports of whether the drivers have ever fallen asleep behind the wheel, whether they ever drive while fatigued and speed and seat belt use of drivers in their companies. We found that CEE- and WE-drivers report a very high level of safety, and receive very high scores for some safety culture items in their firms. In some cases, they exceed the scores of Norwegian firms (Norwegian II sample) with a documented history of targeted safety work and very low accident levels, which would be expected to outperform any random group of HGV drivers.
The results are also not supported by our estimations of HGV accident risk in this study, showing that the accident risk for HGVs from CEE-countries and WE- countries is significantly higher than that of Norwegian HGVs. We therefore hypothesize that the results are not straightforwardly comparable between national samples, and should be used with extreme caution. Below, we discuss several potential explanations for this. It is important to note that many of these are hypotheses that should be examined further in future research:
Small samples
The samples are small (in the case of WE-drivers, extremely small), and respondents may not be representative.
The first sample of Norwegian drivers consist of 61 drivers recruited through a web link to the survey on the website of the Institute of Transport Economics. We therefore do not know the response rate of this sample. The web link was introduced on the Institute of Transport Economics website. A link to this site was also
presented on the Facebook website to members of the “Norwegian cabotage study”, which is a group for fans of a study attempting to map cabotage driving in Norway. We used this Facebook site, as we assumed that most of the members would be Norwegian HGV drivers.
It may be argued that some of these Norwegian are likely to be critical to foreign drivers, and that they therefore could answer strategically in the survey. The introduction text to the survey did, however, not tell the respondents that the data was collected on behalf of the Safe Foreign Transport project (SAFT), or that their results would be compared to those of foreign drivers. Moreover, we have also seen that these Norwegian drivers have less “favourable” scores than the foreign drivers on several issues, which do not indicate that they have answered strategically.
We compare the results of the survey involving the three above mentioned national groups with those of a previous study of safety culture among drivers (N=224) in three Norwegian haulage companies (Nævestad & Bjørnskau 2014). We refer to this sample as “Norwegian II”. The Norwegian II sample is included to supplement our interpretations of the differences between the three groups. Based on our previous qualitative research in the companies of the Norwegian II sample, we have seen that these companies work extensively with safety culture and safety management of work related factors with safety implications. Moreover, we have also seen that these companies have high scores when it comes to safety culture and work related factors with implications for transport safety.
Given the presumably random recruitment of respondents to the Norwegian I sample, and the fact that the Norwegian II sample companies were recruited based on their positive work on safety culture and safety management, we expected the Norwegian II sample to score somewhat higher than the sample of 61 Norwegian drivers on questions related to safety culture and work related factors with safety implications. The present study supported this hypothesis, indicating that the Norwegian sample does not seem to be substantially biased in a positive or negative direction.
The group of drivers from “Western European countries” (WE) is unfortunately too small to be useful for drawing any conclusions about this group. This group actually consists of 8 drivers from Nordic countries and 9 drivers from other European countries (mostly from Holland). The small size of the WE group reflects the limited kilometres driven by these foreign drivers on Norwegian roads. The kilometres driven by Other EU15 countries made up 1 % of the kilometres driven with HGVs in Norway in the period 2007-2012 and 18 % of the kilometres driven with foreign HGVs in Norway. Moreover, only 2 % of the inspected HGVs in the NPRA’s winter inspections in 2015 were from Western Europe, while 20 % were from Central and Eastern Europe. Thus, it may seem that the low share of Western European drivers in our sample to some extent reflects the fact that this group also has a relatively low share of drivers on Norwegian roads.
The group of drivers from CEE countries was established because of a small sample of foreign drivers. Other publications (e.g. Nævestad et al 2014) discern between “older” (Poland, Baltics) and “newer” (Romania, Bulgaria) EU members from Central Eastern Europe. In this sample, however, we group all these countries together. Nævestad et al (2014) did not find the differences between the accident risks of HGVs from “older” and “newer” EU members from Central or Eastern Europe in Norway to be statistically significant. The group of 52 drivers from CEE countries are distributed among the following nationalities: 29 Polish, 16 Baltic (mostly Lithuania), and the rest from Romania, Bulgaria, Hungary and Slovakia. The foreign drivers were recruited by Gunhild Meyer Levlin on rest stops, terminals and parking lots in the South Eastern region of Norway in May 2014. Levlin counted the number of drivers who did not want to answer her survey as she approached them, and she states that the drivers who were unwilling to answer either could not answer because of their language, or because they did not have time or because they did not want to. According to her estimates, a total of 33 % of the drivers that the approached were unwilling to answer the survey, giving her a response rate of 67 %. She states however, that most of the drivers who were unwilling to answer were unwilling because she did not have the survey available in their language (19 %).
Respondents in different countries have different points of reference
The drivers may refer to different baselines or have different anchoring: if safety standards vary substantially between different nationalities or cultures, evaluative judgments could be passed relative to radically different expectations. Respondents from different countries may have different expectations to the safety commitment of their managers and their colleagues, and the safety level of their businesses. It is on the basis of these (different) expectations that respondents make the assessments that the survey implicitly invites them to do (for example, “The drivers in my
company do everything they can to avoid unwanted incidents and accidents” and “The manager of my company focuses on safety”).
This may explain some of the surprising results we have seen in the small-scale survey, for instance when we compared CEE drivers and the Norwegian II sample. In this case, it is likely that respondents in the Norway II sample took the safety level and the safety measures of their own business for granted while answering (e.g. dangerous goods transport in Norway), while drivers from CEE took the safety level and the safety measures of their own business for granted while answering (e.g. CEE based transport company with border crossing transport). Although the absolute or the actual safety levels may differ substantially between these two contexts, the survey does not take this into account, as each group of drivers answer on the basis of their own perspective and knowledge.
If we should have had a real comparison across countries, the respondents would perhaps need full knowledge of, or experience with each other's national realities. Of course this is just a hypothetical example, but it illustrates that we cannot control for the safety level and the safety measures that respondents take for granted when they answer the survey. This illustrates the need for also collecting other data (e.g. interviews, accident data), in order to compensate for the potential weaknesses of survey methods.
One possible way to reduce the effect of the challenge with different baselines and relative answer alternatives could be to make the questions more specific, more focused on actual behavior and most importantly: distribute the answer alternatives on an absolute scale (for example: how often discuss managers and employees safety?, how often are safety issues reported in the company?, how often are assignments interrupted for the sake of safety?, how many procedures exist for safe work?, how often are safe job analyses performed?). These are questions we often ask in the qualitative interviews, and we thereby get important information that can be used to discuss possible interpretations of the results of quantitative data.
In a previous study of safety culture in three Norwegian goods transport companies, we concluded that our questionnaire worked well, as the results of the quantitative survey were in accordance with the results from our qualitative interviews with managers and employee representatives (Nævestad & Bjørnskau 2014). These three companies were all Norwegian, and we may perhaps assume that the respondents’ baselines and their expectations to managers, colleagues and businesses not were very different. In the present study, we may perhaps have to conclude that it appears to be difficult to assess safety culture, safety management and safety behaviours across national borders, because the reference points of respondents' answers are different.
Experience with and trust in surveys
Drivers from different nationalities or cultures may relate to surveys differently. Norwegian drivers are accustomed to being subjects of various tests and surveys. Drivers from other nationalities, however, may be less culturally attuned to these kinds of surveys, and react to them differently. This may have implications for how employees understand and answer such surveys. It is conceivable, for instance, that promises of anonymity are not trusted.
If Scandinavian employees are more accustomed to various tests and surveys, the result may be that Scandinavian employees provide more modest answers (generating lower scores). We see a general tendency, for instance, that CEE-drivers choose the extreme values of the scales to a much larger degree than Norwegian drivers, which seems to indicate that they relate to statements in a different manner. This is, however, mere speculation and more research is needed on this issue.
Awareness of comparison?
Drivers may be aware that they would be compared to other groups, and respond correspondingly. We intentionally omitted to inform the Norwegian sample that they would be compared to foreign drivers, as we believed this might compromise results. In the sample of foreign drivers, however, this was more complicated. In spite of the fact that they were not informed about the comparison, they would perhaps take this as a given, as they were approached in their capacity as foreign drivers in Norway. Since these two groups are competing in the same market, it is conceivable that this influenced responses. Although it is impossible to determine the potential influence of this mechanism, we should be aware of the possibility that this has influenced the answers of the foreign drivers and interpret results with caution.
It is also possible that the foreign drivers in the small-scale survey feared that the survey had some kind of relationship to Norwegian inspection authorities, or perhaps that their admissions of violations could lead to consequences of some sort. After all, the registration numbers of their HGVs were visible to Gunhild Levlin who
interviewed the foreign drivers, and she also saw company logos on the trailers. She stressed, however, that the she was not a representative of Norwegian authorities, and that the survey was anonymous. She experienced that the drivers believed her, but it is of course impossible to verify this impression.
One of the main strengths of our chosen way of recruiting foreign drivers in coincidental resting areas, and without going through their “superiors” is that we hoped to avoid “strategic answers” from the respondents. Another possible strategy of recruiting foreign drivers, could for instance be to go through foreign transport companies, transport centrals or forwarder companies. A potential weakness with such a strategy could be that the drivers’ relationship to these third parties could influence their answers, perhaps as they were fearing that negative results could lead to termination of their contracts. We hope that our recruiting strategy minimized such mechanisms, although we cannot rule out that the drivers were conscious about their reputation, or fearing consequences when they answered the survey.
The items are not good enough
When questionnaires generate results that are unexpected, and when actual objective differences between groups are not reflected in survey results, we should also
contexts of the groups we compare. A general rule here is probably that if the items have worked well in previous research in comparable groups, we would expect them to be “good enough” to register differences between groups. Nevertheless, the quality of questions must always be considered through for instance qualitative interviews with respondents and in pilot surveys.
Based on previous research, we may conclude that it seems that general and “abstract” questions are less suitable to register differences between groups than concrete and specific questions. “The manager of my company focuses on safety” is an example of a fairly general and abstract question which perhaps is unlikely to generate notable differences between groups. “In my company there are routines for reporting safety problems and safety violations”, is an example of a better survey question, because it refers to a more concrete aspect (routines) than “focus on safety”. It is therefore supposedly easier to answer for respondents.
National
National culture may not only influence safety behaviours, it may also influence answers given to questions (Tronpenaars & Hampden-Turner 1997 in Håvold 2005). Measuring safety culture and reporting culture by means of surveys (i.e. self-reports) is in one sense paradoxical, as giving straightforward answers requires a culture which encourages the communication of negative issues (i.e. a good reporting culture). A reporting culture is characterized by employees reporting incidents and near misses and taking actively part in evaluations of the organization’s safety performance. Moreover, it involves actively management encouragement of employees reporting incidents, who are confident that management treat reports and people involved in a just manner (Reason 1997).
If these preconditions are lacking; if employees fear that reports are not confidential, and that reports that may reflect negatively on themselves will lead to punishment, it is likely that they will avoid reporting more or less unfavourable conditions. Thus, perhaps good measurements of safety culture is reliant upon the existence of good safety culture; i.e. a reporting culture involving frank communication of both positive and negative issues.
Deference to authority
Guldenmund et al (2013) suggest that deference to authority may explain why foreign employees gave the answers that they did in a study of safety culture in construction in Denmark, UK and the Netherlands. This study found that groups of Eastern European migrant workers generally rated their managers more positively than employees who were born in the respective countries.
Drawing on Hofstede (2001), Håvold (2005: 453) stress that societies with high power distance teach obedience at school. In the same countries subordinates in work organizations are expected to be told what to do and not to question
hierarchies and authority. In cultures with low power distance on the other hand, it is opposite: teachers and pupils are equal, and in work organizations hierarchy must be justified and it have to serve a particular purpose in order to be accepted (Håvold 2005).
Sivesind’s (1997) doctoral thesis on culture in Norwegian and German companies shows that the Scandinavian labor market may hold a unique position in Europe when it comes to the egalitarian relationship between managers and employees.
While managers in Norway underlined that it is valuable that managers are friends with their employees, it was the opposite in German companies. This may also apply to other countries bordering Germany, for example, countries in Central and Eastern Europe.
It is conceivable that cultural norms concerning the relationships between employees and managers may influence how respondents answer surveys on working
environment and safety culture (cf. Håvold 2005). Such surveys often include a range of questions where respondents are invited to evaluate the efforts of their managers. Management commitment to safety is the most important dimension in studies of safety culture and climate (Flin et al 2000). If employees have great respect for (or perhaps even fear their managers) they may be reluctant to give negative answers. Thus, we may perhaps hypothesize that respondents from cultures with a high power distance and deference to authority are more likely to give survey answers that are in accordance with what they believe that “management wants to hear”. This could mean under reporting negative conditions and over reporting positive conditions. We may perhaps also expect that respondents from cultures with individualistic cultures with low power distance(e.g. Norway, UK) are less likely to report “what they believe management want to hear”. This is, however, mere speculation and more research is needed on this issue.
As noted in the method section, one of the strengths of the methodological
approach employed in the current study is that the foreign drivers not were recruited through their work places. They were also ensured that they were anonymous. We therefore hoped to reduce the consequences of “deference to authority” and answers shaped by deference to authority.
It is impossible to judge the extent to which deference to authority has influenced our survey answers. These issues should therefore be followed up in future qualitative and quantitative research. It is an important topic because it influences employees' willingness to report negative circumstances that may be related to safety; what we can call reporting culture. Moreover, the quality of reporting cultures also influences the quality of questionnaire data.
Reluctance to admitting mistakes and violations
Research indicates that people from different national cultures have different views on what it means to admit mistakes (Lamvik & Ravn 2004). Seafarers from some countries are reluctant to admit their own and others’ mistakes, as this may be conceived of as the same as saying that they are incapable of doing their job (and perhaps that they therefore should not have it) (Lamvik & Ravn 2004). This is a cultural trait that may inhibit reporting of incidents and mistakes, especially when it is