A multifaceted study of occupational safety culture and habits in the Irish construction industry.
This section draws together the evidence of this report addressing five questions:
- How well does
the sample represent the safety aspects of construction activity in
- What can be
concluded about the state of construction safety in Ireland?
- What factors
are associated with better safety compliance?
- What implications
can be drawn from the evidence?
- What recommendations should be made?
The sample was selected to represent both Northern Ireland and the Republic of Ireland and major dimensions of construction activity, which have been associated with fatal accidents. Thus it included both housing construction and general contracting, and construction in and around both the major metropolitan areas, Dublin and Belfast, and in regional towns in the Republic of Ireland. Both large and smaller sites were included, the former being categorized as IR£5 million or over, the latter, IR£1 million or over and under IR£5 million. While small companies are very highly represented in fatal accident statistics, these are often subcontractors working on larger sites. Thus it was anticipated that we would get a reasonable representation of those employed by these firms on the sites chosen. While laborers are the most highly represented occupational category in the fatality statistics, a wide range of occupations are involved in fatalities, and our sample of occupations was based on availability on the sites chosen. The high involvement of falling from heights in construction accident statistics led to a focus, in the site observations and operatives' questionnaire, on factors associated with falling from heights. This while the survey did not include very small sites or sites in purely rural areas, it has effectively included a wide range of construction activity, which includes the major dimensions associated with fatalities.
How representative is the sample of the construction industry in Ireland? There are two pieces of evidence that suggest that the sample may have a higher representation of "better" sites from the point of view of safety. First, the questionnaire results consistently showed that construction workers rated the prevalence of risky situations as being lower in the site on which they were interviewed that in other sites in which they worked. Secondly, it is possible to compare several dimensions recorded in this survey with the results of the surveys carried out by the HSA. These statistics show that less than five percent of sites inspected had safety representatives compared to forty percent of the sites in the Republic in this sample and one third overall (only one out of six in NI). Only fifty percent of the HSA inspected sites had safety statements whereas only one site (in NI) in this sample (of those whose documentation was inspected) did not have one.
While it is impossible to directly compare the criteria used in inspections with those used in this sample, it is interesting to note the relative level of compliance with safety requirements. The HSA inspections documented that 69 percent of sites had inadequate scaffolding, whereas, in this sample, 61 percent of all 18 sites had more than 15 percent of missing guardrails or edge protection. 24 percent of the HSA inspections found edges, openings or fencing to be inadequate, compared to 39 percent of sites in this sample having more than 15 percent of openings found unguarded. In the HSA sample, in eleven percent there was a recommendation for an improvement notice, and in sixteen percent a prohibition notice, compared to two sites (seven percent) in this sample reporting having received improvement notices. It is not clear from the HSA data how many notices were in fact issued.
Perhaps a reasonable conclusion from this would be as follows. Selection for inspection by the HSA is partly driven by internal and external complaints and so is likely to be biased towards sites that have a poor compliance record. The sites in this sample, mostly being drawn from the CIS database, may have a bias towards companies which are better established or better organized. However, the extent of this relative bias may not be very great. It is worth noting that sixty percent of the sites in this sample reported having had a recent inspection by the HSA or HSE. Thus it may be that the level of safety compliance in this sample will provide a modest overestimate of the level of compliance to safety among those working in construction in Ireland.
Nevertheless, even if there is a slight bias in this sample towards sites which may appear better on some indices of safety, the fundamental objective of this study is to investigate factors which are associated with compliance. For that it is necessary to have a good range of sites, which vary both on indices which may predict compliance and on the level of compliance itself. This sample fulfills these criteria quite satisfactorily.
What can be concluded about the state of construction safety in Ireland?
Bearing in mind the above comments, the following broad observations can be made about the state of the art of safety in the Irish construction industry.
Compliance with safety requirements
The level of safety compliance across the sites was quite variable. In relation to the items concerning prevention of falls from heights, compliance ranged from good to poor. Thus, only two sites had full compliance on edge protections and handrails, five sites on unguarded openings. There were ladders which were not tied in eight sites. Thus it can be concluded that there was much that could be improved in the majority of sites. Factor analysis of the fourteen items which had sufficient variability between sites indicated three factors - the first, indicating general noncompliance with safety, included ladders not tied, unguarded openings, unsafe storage, and deficiencies in scaffold boards, toe boards and guardrails. This factor was highly correlated with a smaller factor concerning specific fall prevention measures but rather less so with a factor concerning general housekeeping.
Perception of risk
How well are the risks associated with falling from heights perceived? In general the nine situations concerning working with ladders, scaffold and on roofs were perceived as high risk (the major exception being short ladders, which were perceived as medium risk). However a small minority of workers in each site saw many of these situations as low risk. The frequency of these situations was generally seen to be fairly low in the construction industry, though constructions sites are generally perceived to be dangerous places. Workers do not generally see difficulty in being aware of hazards, or their familiarity with work, or the demands of their work as significantly contributing to safety. Thus, for most workers, developing and maintaining awareness of risk is not perceived to be a significant problem.
The great majority of workers report that they would respond constructively to risky situations, either reporting the defect, fixing it, or stopping work (depending on the situation). However, a sizeable minority (between ten and twenty percent) would just continue working (and this was between twenty and thirty percent for ladders being too short and working on roofs in bad weather). Thus there is a small minority who may not perceive risks accurately and a larger minority who say they are prepared to continue working in risky situations.
Almost forty percent of operatives report that they have received no safety training from the main contractor. For a further fifty percent their safety training comprised the induction course (lasting between ten minutes and one and a half-hours). While managers consider this safety training to be necessary and indeed should be supplemented by further refresher courses, many believe the way it is currently carried out is primarily to "cover themselves" in case something goes wrong. Eleven percent of the sample has received safety training lasting more than one day. Very few sites claimed to have a systematic safety training schedule for their employees, and safety training was not usually a consideration in recruiting operatives, apart from the certificates required for scaffolders, crane and digger drivers, etc. In Northern Ireland there was much reliance on the mobile CITB unit, and in the Republic interest was expressed in the potential benefits of the Safe Pass Scheme. Thus it seems clear that for the majority of operatives the main way in which they achieve knowledge of the risks of their work is through their experience of work itself. Even where the main contractor provides induction training, this is often perceived as a formality, to protect the company, with little expectation that it would influence the knowledge and behavior of workers. The low level of provision of safety training is particularly worrying considering that twenty five percent of the sample report that they have worked in the industry for less than one year and fifty percent less than two years. Thus their level of experience of the risks associated with the job and of how to manage them may not be that great. Direct and vicarious experience of hazards may be the main way of gaining knowledge of the risks of construction and how to manage them. This is a reactive and potentially dangerous way of learning.
Both the management interviews and analysis of safety documentation indicated that whereas virtually all the sites had a safety plan, these were mostly generic documents that could be applied to any construction site. Thus, few contained a management structure with clear delegation of responsibilities. Few contained safety statements from subcontractors. Although half the sites had risk assessments which were site specific, in some of these not all the relevant site activities were addressed. These documents played little or no role in ongoing management activities and their function appeared solely to meet legislative requirements.
All sites conducted safety audits, mostly following a checklist method with a written audit report. However few sites documented improvement measures or actions taken to remedy audit deficiencies. Hazard reporting was rarely documented and the same was true of incidents and near misses. Communication about safety was most often informal and verbal. A quarter of the sites had regular dedicated safety meetings, though for others safety was a frequent item on site meetings.
In all but one site some or all of the management had undertaken CIF/CEF training courses. A small number of safety managers had also undertaken diploma courses in health and safety management. Six sites had a safety representative. Frequently these had been appointed by the management.
A factor analysis of the variables investigated in the safety management interviews indicated that a range of safety management activities, including those just discussed, were highly correlated in one general safety management factor. A second factor comprised two items, which concerned the effectiveness of audits and hazard reporting in redirecting organizational action to improve safety. These two factors were not significantly correlated indicating that commitment to a range of safety management activities does not necessarily imply an effective response to audits and hazard reports. In general the perception by workers of the commitment of management to safety was fairly positive.
Role of inspectors
While the majority of sites reported having had an inspection from the HSA or HSE, reactions to these inspections varied. Many reported no feedback or follow up from the inspectors to ensure implementation of recommendations. While many sites welcomed the inspections and would have wanted more in-depth coverage from them, inspectors do not see their role as doing a full site audit and their practice in conducting an examination of site documentation varies. From the inspectors' point of view, it is through their exercise of their legal function (improvement and prohibition notices) that they see themselves bringing pressure on the worst performing sites. Other comments were quite critical of the inspections. Negative comments did not seem to be systematically related to experience of inspection on that site. Some could reflect the fact that inspections of construction sites are sometimes done by inspectors with no specific qualifications in construction. The most common complaint was that smaller less safe sites were overlooked. However inspectors report that a large proportion of inspections are reactive to complaints or accident reports (overall, it is probably fair to say that there was little evidence of direct positive influence on the sites surveyed from the inspections reported).
In conclusion, the evidence suggests great variability in compliance with safety requirements, with plenty of room for improvement. The great majority of workers on construction sites report good awareness of risks associated with working at height and the majority would respond constructively to risky situations, though a significant minority would just continue working. Training for construction workers is patchy and seen to be inadequate, relying to a large degree on site induction training. Certification of specific trades (scaffolders, crane drivers, etc.) is monitored. The high proportion of relatively inexperienced workers gives rise to concern about their effective appraisal of and response to risk.
Although workers are fairly positive about the commitment of managers to safety, the evidence from management interviews and documentation is less positive. Safety plans and safety assessments are most often paper exercises with little role in day-to-day management. Only a minority of sites have safety representatives. Audits based on checklists are used in most sites but it is difficult to find documentary evidence of corrective action. While a range of core safety management activities is closely associated together, these do not correlate highly with effective response to audits and hazard reports. While the majority of sites have had an inspection from the HSA or HSE, the priorities of the inspectorate are reported to focus on using improvement and prohibition notices to influence compliance (other than that, there seems to be little opportunity for inspections to influence what happens on site).
What factors are associated with better safety compliance?
Perhaps the core question in this study concerns the factors that may be associated with safety behavior and compliance with safety requirements. Those factors that tend to be present in the sites with better safety outcomes may represent the most effective influences on safety, whereas factors, which do not discriminate between relatively safe and unsafe sites are unlikely to be critical in promoting safety. For the purposes of this analysis the critical outcome variables are the safety compliance factors, and the safety behaviors which workers report they would follow in risky situations - report it, fix it, stop working or just continue working.
The strongest relationship with the main safety compliance factor is with the presence or absence of a safety representative. A safety representative on site is associated with better compliance. The only other factor which is associated with compliance, is the second safety management factor - effectiveness of response to audits and hazard reports. The general safety management factor is not significantly associated with safety compliance, but does have a modest relationship with the presence of a safety representative (this is just short of statistical significance). The presence of a safety representative is the only factor which is significantly related to safety behaviors - safety representatives are associated with a greater likelihood of reporting risky situations and a lower likelihood of simply continuing working in such situations. The presence of a safety representative is also strongly related to the effectiveness of response to audits and reported hazards. This pattern of relationships suggests that safety representatives are the most important influence on the association between effectiveness of response to audits and hazards and safety compliance.
The presence of a safety representative on site shows the strongest relationship with safety compliance (the bivariate correlation of almost 0.80 is very high). Although there is a tendency (almost significant) for safety representatives to be present in sites with better general safety management performance, it would be a mistake to conclude that this general safety management factor accounts for the relationship between safety representatives and compliance. There is no significant relationship between the general safety management factor and compliance. This factor, at best, has an indirect relationship with compliance through the appointment of a safety representative. It is the presence of the safety representative which influences compliance.
How does the safety representative influence compliance? The pattern of correlation and the multivariate analysis suggest that the presence of a safety representative accounts for the great part of the relationship between the effective response to audits and hazards (Effectiveness of the management system or Management-Factor 2) and compliance with safety. Safety representatives strongly influence the response to audits and hazards (bivariate r = .60) but their influence on compliance is even stronger (bivariate r = .79). This suggests that they influence safety compliance not only through their influence on the response to audits and hazards but also through other means.
The safety representative variable is also the only variable that shows a significant relationship with reported safety behaviors-specifically reporting hazards and not continuing to work in hazardous situations. Thus safety representatives encourage the reporting of hazards and play the major role in ensuring that these reports lead to better safety compliance on site. Their presence also makes it significantly more likely that workers will not continue to work in hazardous situations.
It should be pointed out that these are not actual measures of workers' behaviors but reports of what they would do in specific hazardous situations. Thus, it is more accurate to say that the influence demonstrated here is on the expressed intentions of workers on site rather on their observed and recorded behavior. However, although we cannot assume that intentions will always be translated into actions, what has been demonstrated here is a strong and systematic relationship between the presence of a safety representative and intended behaviors in a range of specific safety situations which are generally perceived by workers as being risky.
Two aspects of the statistical analysis are worth a comment because of the, perhaps surprising, lack of statistically significant relationships. The first is the lack of association between de general Safety management factor and anything else except the appointment of safety representatives (almost significant). In particular, although the correlation between the general Safety management factor (Management-Factor 1) and Effectiveness of the management system factor (Management-Factor 2) is 0.36, this falls short of statistical significance and thus we cannot be confident that this is not a chance result. Likewise there is no relationship is which we can have confidence between the general safety management factor and either safety compliance or the variables assessing safety behaviors. These results must pose a fundamental question: why does so much activity which is undertaken in the name of safety apparently have so little influence on safety compliance and safety behaviors?
Secondly, there is the lack of any significant relationship between the attitudinal variables (safety attitudes, perception or risk and safety climate) and any of the safety outcome variables (compliance, safety behaviors). General safety attitudes were fairly high, though general attitudes towards risk were more ambivalent. The climate measure shows a generally positive perception of the management commitment to safety, and the specific risk perception measures show, amongst the great majority, an accurate perception of risk. These attitudes and perceptions do not appear to have an influence on safety behaviors and compliance. Thus it is possible to have fairly strong pro-safety attitudes and perceptions but poor levels of site safety compliance. This suggests that the difficulty of getting more consistent and higher standards of safety compliance may be less to do with attitudes and perceptions of workers and managers, and more to do with systemic factors -having mechanisms for reporting hazards, following up on hazard reports and audits, and doing what it takes to ensure that hazard reports and audits are translated into effective compliance with safety requirements.
What is remarkable about the findings reported here is how little safety management activity has a measurable positive influence on safety outcomes -suggesting that safety management systems as they exist very often do not provide a systemic basis for effective safety action. It is equally surprising that safety representatives who have little or no formal power, and some rights to be consulted and informed, have the strongest and most pervasive influence on safety compliance and safety behaviors.
It is worth noting that most of these safety representatives are, in no formal sense, representatives -as they were appointed by management rather than elected by their fellow workers. This in turn suggests the strong and important role of informal mechanisms of influence and persuasion in developing effective safety management. The role of safety representatives in ensuring the job goes smoothly, in facilitating communication, hazard reporting, and playing a strong informal disciplinary role was repeatedly highlighted in the interviews. Their ability to inspire trust and confidence amongst fellow workers on site is illustrated in the following quotes: "If you approach them in the right manner they will listen to you", "You need to communicate the message to young inexperienced staff". Although in this sample it was safety representatives who played this role (wherever they had been appointed), there is no reason in principle why these informal mechanisms of influence should not be exerted by others as well as safety representatives. This is clearly not being done in the current situation.
As the majority of sites had at some time been subject to an inspection by national authority inspectors, it seems sensible to pose the question -do these inspections have any direct influence on safety compliance on site? We are not here concerned with the possible influences of national activities of the authorities which are not quantifiable in this context. Setting aside, also, the role of improvement and prohibition notices, the effect of which could not be directly investigated in this study, the main sources of evidence which can be brought to bear on this question are:
- what site personnel
say about inspections,
- what inspectors
say about their role and objectives,
- whether there is any evidence that sites which have received inspections perform better than those that have not.
Thus, it could be concluded that while quite a lot of activity by inspectors was reported in relation to the sites in this survey, very little of this activity appears to have, or to be designed to have, a direct influence on the effective management of site safety. This may be a missed opportunity to create value (in terms of safety impact) out of a large scale national activity.
There was little evidence of any other direct influence on site safety management from other agencies external to the site, and the main and subcontractors concerned. This includes architects, the client of the building contract, or trades unions. External safety consultants were engaged on some sites in providing an auditing service. After handing the conclusions to the main contractor, these external consultants appear not to have an influence in the measures employed by the main contractor in relation to safety. The lack of apparent direct an identifiable influence by these external agencies does not, of course, imply that they have no important role in site safety. Our conclusion is simply that, in terms of the dimensions and measures which were employed in this study there was no evidence of a direct and identifiable influence. It would take a very different study to systematically investigate the safety role and influence of such agencies.
What are the implications of the evidence?
At its broadest level the main implication is that the system is not working. The core principle of the European safety management framework is the identification of responsibility for safety at the level of the enterprise and the consequent outlining of a range of safety management actions through which this responsibility is to be exercised. These duties of management are complemented by duties of employees to cooperate and to behave safely. The Irish safety regulations, including those specific to construction, derive from this model, which is based on the premise that, if management and workers fulfill their duties and obligations, then safety will ensue. Most starkly, the evidence from this study suggests that this model is not working or, at best, is only working in a patchy and fragmented manner.
- The formal
documentation of the safety management system - the safety statement
and risk assessments - are most often seen to be a paper exercise
whose role is to satisfy the legal requirement, rather than being
an effective guide to management action.
- The same can
be said about much of the operatives' safety training, which is undertaken.
There is a lack of serious effort to ensure that inexperienced workers
(a high proportion of the workforce) receive a thorough basic training
in safety. The study elicited less evidence about the quality of management
safety training, but none of the evidence suggests that it is having
any major impact on compliance.
- While nearly
all sites can provide evidence of the safety audits that they have
undertaken, few can show how and when the defects highlighted in the
audits have been corrected. Many safety managers express frustration
at their inability to ensure that recommendations are translated into
corrective action. When it comes to executive authority it is often
too easy to marginalize the advisory role of the safety officer.
- The primary
role of inspectors from the regulatory authorities is to inspect the
site itself. There is no systematic approach to auditing or holding
accountable the safety management system of the main contractor and
its subcontractors. Other than through the formal channels of improvement
and prohibition notices (which were not investigated in this study)
there is little follow-up and little evidence of significant influence
of the regulatory authorities on safety behavior or compliance.
- Only a minority of sites have a safety representative.
- The way in
which safety representatives can exploit their very informal role
to help to influence what both workers and management do.
- The positive
attitudes towards safety exhibited across the sample
- The high awareness
of the riskiness of situations associated with working at heights
and the tendency of the majority to act in a pro-safety manner in
- The level of commitment and activity by many safety officers and management to take their safety responsibilities seriously and fulfill them wholeheartedly.
the safety management process
- Defining more
clearly the measurable outputs of safety management
- Developing competence
standards for training
- Ensuring that performance criteria & targets are auditable.
Finally it is appropriate to make some recommendations arising from the findings of this study.
This study has demonstrated the potentially strong role which safety representatives can play in influencing both behavior and compliance with safety requirements, and ensuring that both audits and hazard reports are effectively dealt with. All sites should have safety representatives and their role and functions should be reinforced as part of the safety management system.
Training and certification
The most plausible interpretation of the findings in relation to safety representatives is that their effectiveness is largely due to their ability to exert influence and persuasion through informal interpersonal methods. The social and interpersonal skills which this requires should not be the prerogative of safety representatives alone. There is enormous scope for improving the ability of those who have management and safety responsibilities, and indeed all those who work in construction, to manage the human relations of safety more effectively. These skills are trainable and susceptible to systematic development. They need to be clearly and systematically addressed in all training related to safety in construction.
As far as possible, training should not only seek to foster awareness of hazard and risk, but it should strengthen knowledge and skills in managing risky situations effectively. This should include the communications and interpersonal skills, which are necessary at every level to ensure that the correct influences on behavior are consistently reinforced. Transfer of these skills to the working environment needs to be carefully fostered and monitored. All levels of training should be addressed including
- Site induction
and refresher training
to the industry and specific crafts / trades training
- Management training. Particular attention should be paid to fostering participation in in-depth professional safety management training at diploma and masters level.
Clearly the requirements for safety management systems need to be radically reviewed and overhauled. It is too easy to comply with the law through having a paper system, which does not effectively operate in practice. This review should address:
stronger criteria for active and effective safety management systems.
These should include design and planning, day-to-day management and
monitoring and auditing practices.
- These criteria
should be developed in new ways of auditing safety management systems,
which can routinely and reliably assess the activity and effectiveness
of the system.
- The role of
safety officer should be strengthened to ensure that it is less easy
to marginalize what is essentially an advisory role.
- The accountability of operational management needs to be made clearer and firmer, and this accountability needs to be tied to measurable outputs of the safety management system, including the demonstration of effective action to address identified defects and hazards.
The regulatory authorities and the social partners who collaborate in rule making are responsible to ensure the effectiveness of policy. Perhaps the clearest conclusion from this study is that established policy is not working well. New initiatives are currently underway in the Construction Safety Partnership Plan (2000), and new codes of practice in scaffolding and other areas have recently been published. At the time that the field work for this study was collected (Nov. 2000 - Feb. 2001), the Construction Safety Partnership Plan was in the early stages of its implementation. Two measures in the CSPP were most apparent in the survey: the safety representatives' scheme which was evident in a number of sites in the Republic, and safety training for managers (involving the great majority of sites). The evidence suggests that the safety representatives' scheme is highly successful. On the other hand safety training for managers does not seem to be delivering a higher standard of safety compliance on many sites.
Other initiatives do not seem to have been implemented widely enough to have been apparent in the survey, including the recommendation for inspectors to leave a report on their inspections (the reported inspections were at a time earlier than the survey). The Safe Pass Scheme and Construction Skills Certification Scheme were known about but not widely implemented at that time.
In the light of the findings, critical issues for the CSPP include the following:
- Extending the
safety representative scheme to all sites. There was a tendency for
the sites with safety representatives to already have better general
safety management and many of the representatives were, in fact, appointed
by the main contractor. Developing effective safety representation
on all sites will need more effective support from the social partners
to make this work where management commitment is lower.
- An effective
methodology for routine site audits must be developed. Auditing will
not be effective unless it includes an effective system for monitoring
the implementation of safety measures and response to hazard reports.
The CSPP only states that the current auditing arrangements will be
reviewed by early 2001. This should be strengthened to require more
stringent criteria for auditing.
- The recommendations
for the introduction of a Safety Management System by the Construction
Industry Federation should urgently be reviewed in the light of the
evidence of this report. Will the proposals for an SMS effectively
overcome the limitations in safety management identified in this study?
The evidence suggests that a lot of safety management activity is
relatively ineffective from the point of view of safety performance.
Recommendations for Safety Management Systems must address the problem
of translating a paper demonstration that there is a management system
into clear evidence that that system is delivering improvements in
safety on the ground.
- The wide range
of training recommendations in the CSPP are to be welcomed. However
it is important to ensure and to demonstrate that this training is
effective and that the safety messages are transferred to the site
and result in safer behavior and more effective safety management.
The management training program should be urgently reviewed in the
light of the lack of evidence that it is having an impact. The other
training initiatives should also be reviewed with respect to how well
they address the social and interpersonal processes which are essential
to ensuring that safety is effectively addressed on a day-to-day basis.
- The role and
activities of Inspectors should be reviewed with a view to maximizing
their impact on site safety in a cost-effective manner. The advantages
of developing additional methods of influence which do not require
legal sanction should be explored. Attention should be given to strengthening
the oversight of inspectors in relation to the safety management systems
of the contractors concerned; to improving their communication and
feedback to encourage better safety activity; and to arrangements
for follow up activities to monitor improvements.
- The CSPP is a major social initiative with considerable resource implications for all the social partners. Yet there is no proposal to monitor whether or not this initiative will be effective in improving safety in the construction industry, beyond noting the implementation of the various measures. If it is partially successful, there will be no way to systematically evaluate the strengths and weakness of various measures and initiatives and adjust them with a reasonable expectation of improving the overall program. A systematic evaluation strategy needs to be urgently developed. This current survey was conducted at an early stage of the implementation of the CSPP. It could form part of a baseline measure against which improvements could be judged. However evaluation should not be done piecemeal, but as part of a clearly thought out strategy in relation to the objectives of the program. Good public policy would seem to require that such a strategy be developed as a matter of urgency.