CHaSPI: the Corporate Health and Safety Performance Index

Howard Fidderman tries out an impressive new tool to help investors choose where to put their money, and believes that it should help focus the minds of directors on health and safety.

The HSE is sponsoring a web-based "index" that aims to help investors assess the health and safety performance of organisations in which they are investing, or thinking of investing. The HSE hopes that this, in turn, will encourage directors and senior managers of organisations to take the issue seriously. The "Corporate Health and Safety Performance Index" (CHaSPI)¹ is undergoing a large-scale validation over the next few months and, after modification, should be open from early 2005 to all organisations with 250 or more employees.

The index asks organisations to supply health and safety information under nine "indicators", after which it generates reports and a score out of 10. Investors are not, however, the only group expected to use the index. It is also likely to help:

employers, to advertise their corporate social responsibility (CSR) performance, benchmark their health and safety performance against other organisations and monitor their own performance over time;

potential employees, to decide whether or not they wish to work for an organisation;

regulators, although the HSE stresses that it has no plans to use data from the index to inform its enforcement activities. If the index becomes widely accepted and used, the HSE says it could be another source of valuable information on the management of occupational health and safety. It expects local authorities to act in a similar way;

insurers, to help determine the terms that they offer an organisation; and

trade unions and pressure groups, to highlight the good and target the bad.

The online index has been produced by Greenstreet Berman and InfoMatrix, and it follows a two-stage pilot of a paper version (see below). The index itself was due to become live by mid-March, once an external validator has been selected. The validator, which will start work in early April, will work with around 100 organisations: some will be invited to participate; some will be selected from those registering with the index; and some will be "users" such as insurers.

The validation will continue through the summer, with a validator's research report published on the HSE website after September. The HSE will consider the report and the usefulness of the performance results in informing socially responsible investment decisions. The HSE will then modify the index in line with the report and launch the final version in early 2005.

Use of the index is free during the validation; there might be a small fee for registration or administration from 2005. During the validation phase participants remain anonymous. The HSE stresses that the index is still work in progress, and that it expects validation to lead to changes. Organisations may, therefore, find that its results are misleading, especially if they compare their score from this version against the final one.

Persuading directors

An HSE-commissioned research report published in 2002 (see box 1) highlighted the potential influence that investors, particularly larger institutions, can have on health and safety.

The report, produced by Claros Consulting, found an interest among investors in health and safety and an intuitive understanding of health and safety management issues, with some investors even asking questions about health and safety management. But it also revealed a need for a tool that enabled comparisons, and recommended that the health and safety "community" should assist in its development. The HSE therefore commissioned Greenstreet Berman to develop the tool, and CHaSPI is the result.

The development of CHaSPI reflects the increasing importance of CSR and fledgling attempts to include health and safety within the CSR agenda (see box 2). The index should help realise one of the main aims of the government and the HSC's Revitalising health and safety strategy (Employers face major health and safety at work shake-up), ie to find new ways to motivate senior managers and directors to take health and safety seriously. As such, it is part of a package that has seen advice for directors (Directors take safety on board), a renewed advocacy of the business case for health and safety and a "challenge" to the UK's top 350 companies to report publicly to a common standard on health and safety (Top 350 firms shun Government reporting "invitation").

CHaSPI

The Claros report recommended six indicators that would help investors. CHaSPI seeks information from participating organisations under nine indicators, which incorporate four of the Claros recommendations, although not necessarily as headline indicators, but exclude indicators on costs and lost-time accidents. The CHaSPI indicators also have far more depth than the Claros indicators.

Five of CHaSPI's nine indicators are quantitative and generate an overall score, which is the organisation's rating. They cover:

health and safety management;

injury rates (employees and contractors);

employee sickness absence rates;

a measurement of occupational health risk management; and

a major incident rate indicator.

Organisations can score up to 10 points on each of the five indicators. The scores are then weighted so the importance of the health and safety management indicator (50% of the total) is equal to the total of the other four (each of which rates 12.5%). The overall score is also given out of 10.

The four qualitative indicators, which do not count towards the rating, seek:

information on whether or not an organisation conducts highly regulated activities;

a declaration as to whether or not an organisation has suffered an "under-watch" event and what it is doing to rectify matters;

a directors' declaration that all health and safety risks are adequately assessed and managed; and

an acknowledgment as to whether or not the data used for the index have been verified by another organisation.

Each of the nine indicators links to at least one sub-indicator; one of the indicators has 11 sub-indicators, one has two, and the other seven have one each. Each of the sub-indicators links to a questionnaire comprising one to 15 "elements", most of which require a simple response of "yes", "no" or "some". Others also allow an answer of "non-applicable", and a few require statistics or further information. There is also a 10th indicator that some organisations are asked to answer in certain circumstances (see below).

The HSE says this combination of outcomes and management processes gives "a fair reflection of an organisation's overall performance in health and safety management". The five indicators and four additional sets of information are explored below.

Health and safety management indicator

The first indicator looks at how an organisation manages health and safety. This indicator draws on core HSE publications, notably Successful health and safety management (HSG65), to examine areas such as the organisation's policies, risk management strategy, monitoring and review procedures and involvement of its workforce. (The HSE stresses that the index is not supposed to provide guidance on effective health and safety management; for this, employers should look to HSG65 or standards such as OHSAS 18001 or BS8800.)

The indicator comprises 11 equally rated sub-indicators, which are set out in box 3. The sub-indicators link to a total of 83 elements; those for the first sub-indicator (targets) are set out in box 4 as an example. The first nine sub-indicators, according to the guidance to the index, give "strategic insight into the management approach of the organisation" and identify factors that should indicate an organisation's performance.

The health and safety management indicator, as noted earlier, is the most significant of the indicators and, consequently, takes more time to complete than the other eight indicators combined.

There is also a 10th qualitative indicator, covering supplementary information on the management of health and safety, but organisations complete this only if they do not have a recognised safety management system in place (sub-indicator no.10), or if there is any uncertainty regarding their system.

This indicator has five sub-indicators asking about an organisation's policy (12 elements), organisation (29 elements), planning (20 elements), performance measurement (28 elements) and auditing and review (10 elements). If organisations state that they have a management system, the information they provide feeds through to a publicly available page as "additional information".

Injury rate indicator

A report produced by Greenstreet Berman² on the pilot notes that the second indicator (injury rates) is "an adequate, if not perfect, outcome indicator of health and safety performance" for larger organisations. The guidance to the index adds that "it is also recognised that past injury rate performance is not necessarily a predictor of current and future performance".

After much discussion - for example, whether or not the indicator should be based on a combined fatal and major injury rate - the Greenstreet Berman researchers opted for injuries that lead to more than three days off work. This, in effect, means that organisations should include all RIDDOR-reportable injuries - ie fatal, major and over-three-day injuries. Although this does not allow compatibility with the Revitalising fatal and major injury rate target, the HSE says the choice of over-three-day injuries was driven by the need to match the form in which the majority of companies most reliably collate their data, and thereby to ensure consistency between organisations. Other than over-three-day injuries, companies use a myriad of reporting schemes, such as all injuries and first-aid injuries.

The researchers used annual RIDDOR returns to calculate a rate of 635 over-three-day injuries per 100,000 workers as the mid-point of a logarithmic one-to-10 scale. The upper and lower points of the scale are 100-fold higher and lower than the average.

The index asks organisations to provide the number of injuries to employees and contractors sustained during the previous financial year, and the total numbers of workers for the same period in each group. Recognising that supplying such information on contractors can be problematic, the index allows an organisation to supply the total number of contractors' hours worked or the percentage of its total output for which contractors are responsible. If an organisation is unable to supply any information on contractors, the index will calculate a rate based on employees alone.

The highlighting of the mere presence of contractors, rather than their injury rates, is interesting. The popular perception, due in no small part to recent railway accidents, is that the use of contractors is inherently riskier than the use of employees. But a research report published in late 2003 by the HSE found that subcontracting in the three industrial sectors examined had no negative impact on employee health and safety and, in some cases, had even improved it.

The HSE says the "last resort" question on contractors does not "imply or assume anything" about contractor injury rates. It said: "We are trying to give users of the index results an indication of whether the failure to provide contractor injury data is a significant concern or not. So, if the organisation can declare it makes little use of contractors, this may allay fears. If, on the other hand, the organisation makes extensive use of contractors this may prompt further examination by the user of the index results."

Sickness absence indicator

The researchers had wanted to use occupational health outcomes as the third indicator. The reliability of such outcomes is, however, problematic because of: long periods of latency; difficulties in establishing that an illness is related to work; and under-reporting and general paucity of official statistics. Consultation found a general acceptance that sickness rates "reveal something of the performance" of a company in relation to the general welfare of its employees (as well as work-related ill health), and so the researchers decided to use sickness absence data as a proxy for an indicator of health outcomes.

The benchmark rate reflects an average sickness rate per worker in the UK of eight days a year and a range of one to 15 days. On a linear scale of one to 10, sick days of up to two a year scores 10 points; anything above 15 days scores 0: an average of five days' absence would, for example, score 7.6 points.

Occupational health risk management

The fourth indicator looks at occupational health management and stresses the importance of seeking information proactively, rather than waiting for people to become sick. There is just one sub-indicator with 14, equally rated elements (see box 5).

Of the 14 elements, three cover rehabilitation and return-to-work services; there are two each on absence, repetitive injuries and lifting and handling, stress, and symptoms; and there is one each on targets, occupational health services, and individual health and capabilities. The prominence given to stress and musculoskeletal issues reflects their dominance of work-related ill-health tables.

Major incident risk indicator

The fifth and final indicator that counts as part of an organisation's index score covers major incident risk. The researchers included this indicator because they believed that the management of universal hazards such as slips, trips and falls does not necessarily reflect the management of major hazards.

The indicator is based on the number of major incidents in seven categories that an organisation has suffered in the previous three years (see box 6), rather than the fact that it is involved in activities that have the potential to cause a major incident. The seven categories range from incidents with more than 1,000 deaths to "anomalies" where an organisation's defences have remained intact. Organisations should note that this indicator could apply to many more sites than those covered by major hazards legislation - incidents that come under this umbrella include fire, for example.

An organisation has to consider if it has had any such incidents and enter types and numbers. The researchers developed the indicator by looking at the number of major incidents in the UK each year, calculating the rate per 100,000 employees and producing a logarithmic scale to cover the rates (again because the range is so extensive).

Flag waving

The sixth indicator seeks to ascertain if an organisation has suffered an "under-watch" event and what it is doing to rectify matters. The events are essentially those with the potential to cause a major loss in terms of money or image. If concerns remain, the index raises a flag.

Greenstreet Berman included the flag, originally described as a "code", because of the number of occasions when incidents or liabilities have resulted in significant problems for an organisation, for example the winding-up of Railtrack after one accident too many. The indicator is analogous to one used in the "Standard" and "Poor" index that applies to organisations where there is a major threat to an organisation's survival owing to an actual, pending or high probability of a major financial loss or business interruption.

The indicator requires organisations to confirm that they have read the guidance on the types of events (see box 7) that would put them "under watch" for health and safety risk management and declare whether or not they have suffered such an event and that it is currently affecting its performance or the perception of its performance. Organisations should note that this indicator can sometimes reflect loss of confidence rather than an actual inability to manage health and safety risks. It does not affect the overall score for the organisation.

The organisation has to provide a description of the event and has the option of outlining its response. An organisation is placed "under watch" until:

it becomes apparent that the potential event or its potential consequences have not, or will not, materialise;

it has demonstrated that it has recovered from the incident or has a recovery plan that will assure restoration of confidence in its safety performance; or

the consequences or actions arising from the event have ceased or been completed, and the organisation has effective health and safety management arrangements in place.

Organisations placed under watch may include those that have not "failed". In these cases, a flag is raised until further information on the event becomes available. Examples of serious incidents that do not reflect on an organisation include those caused by terrorist action, those that the organisation could neither foresee nor reasonably prevent, and crashes caused by isolated engineering or operational faults. But some no-fault events will still raise a flag if:

the direct costs and consequences of an accident are of such a scale that they have serious ramifications, for example the prohibition of an activity;

the remediation costs are prohibitive (some activities may be deemed to be inherently unsafe and hence unacceptable regardless of the competence of the organisation); or

the incident, though isolated, causes a loss of market position, for example where a product is temporarily withdrawn and consumers have found alternatives.

Greenstreet Berman acknowledges that this might seem "unfair" to a health and safety professional in some circumstances. But it stresses that investors need this type of information, which does not attempt to draw any conclusion about whether the circumstances are a misperception, or whether they may indicate a fundamental flaw in risk management.

Declaration of highly regulated activities

Indicator seven asks if the organisation carries out any activities that are subject to special Regulations "or their equivalent in other countries". The activities are those that are subject to or involve: the COMAH, railway safety case or offshore safety case Regulations; asbestos, nuclear site, CAA or FAA licensing; and mining, transport of hazardous cargoes, explosives and "others".

The purpose of this indicator is to show index users that hazards are present that have the potential to cause major accidents. It seems surprising that this indicator covers activities in "other countries", given that the index stresses that it is restricted to the UK. Greenstreet Berman says the indicator asks for this information because it is an issue of business risk management as a whole; a high-profile incident abroad will affect confidence in the organisation's ability to manage its risks in the UK.

Directors' declaration

The Greenstreet Berman report notes that investors with limited time and research resources will often require a declaration that organisations are managing their health and safety risks. The eighth indicator therefore seeks a statement to the effect that the board has assessed the health and safety hazards associated with its activities, and has implemented an appropriate set of risk management controls (see box 8). The information also includes the name of the person that has signed the organisation's declaration and the role of the director (or equivalent) who has signed the declaration.

External verification

The final indicator asks the organisation to declare whether or not the data it has used to complete the index has been verified by another organisation, ie a qualified and independent third party. A declaration of verification was not included in the pilot index, and the Greenstreet Berman report highlighted the need for it as crucial to the index.

Publishing the results

After an organisation has completed all nine indicators, it can "sign off" its entry but cannot make further changes without the agreement of the index validator. At any point during the process of completing its entry, an organisation can obtain a summary report of the entry and a CHaSPI indicator summary, which is viewable only by the organisation's registered users. When the entry is signed off, its details are included in three additional reports that are publicly available. These provide:

a list of all organisations' complete CHaSPI results, which can be sorted by overall scores and by specific organisations;

results by selection, using filtering criteria such as turnover, sector and FTSE or business class; and

same-sector comparisons relative to the highest, mean and lowest scores.

The pilots

The electronic version of the index now under validation is the result of a two-stage pilot carried out by Greenstreet Berman, using a paper version of the index with 20 different organisations (nine in stage one; 11 in stage two). Of these, seven and 10 completed an index and all returned a questionnaire. Over the two stages, six organisations thought the benefits outweighed the costs, four that they were balanced and five that the effort outweighed the benefits.

The majority thought the index was practical, clear, covered the right issues, and gave an adequate range of responses. Most said it was useful, although reactions were often qualified. Reasons for usefulness included health and safety improvements, benchmarking, setting key performance indicators and benefit for investors.

Although the first-phase feedback was positive -- only one organisation said that the index was not meaningful in its sector - there were concerns that scores were subjective and not verified externally and that there was little option to indicate partial compliance with an item. The second stage took account of most of the concerns; the researchers attribute the two dissenting organisations to reasons specific to the organisations. Overall, the responses were more positive, although only six out of the 10 thought that it would be a good idea to have their results published, which is a fundamental purpose of the index. Refinements made after the second phase included adding "explosives" to the declaration on high hazards, and a clear instruction in the management and occupational health indicators that "some" is an acceptable answer only where the organisation is on its way to achieving a "yes".

Comparing results

The results led the researchers to conclude that the index can allow comparisons of organisations, ie there was a sufficiently "wide variation" in scores, ranging from 4.09 to 7.99 out of 10, that allowed discrimination between companies. The researchers do admit, however, to "some clustering" that "may be explained by a bias in [the] respondents towards committed and well-performing organisations".

Closer examination reveals that of the 16 companies assigned a score in the pilot:

10 scored between 7.1 and 7.99;

three between 6.5 and 6.87; and

three between 4.09 and 5.

Although on first reading this does not appear to be a particularly wide range, both the HSE and Greenstreet Berman point out that the range is reasonable given the fact that they were dealing with "good organisations". While the researchers would have preferred to pilot the index on a wider range of firms, they were constrained by the voluntary nature of the project. One of the reasons for validating the index is to test the effectiveness of the scoring system across a wider group of organisations, and it would be expected that when the index is opened up, there would be many scores at the lower end of the scale.

Scores for the individual indicators ranged more widely, from 2.67 to 10. The HSE accepts that that the results of individual indicators may be more meaningful than the overall score, but the researchers found that many potential users were looking for a single indicator.

The correlations between the occupational health management scores and absence rates, and between health and safety management and the injury rates, were high. But those between health and safety management and health management were far lower, suggesting that firms with better health and safety management scores did not necessarily have better occupational health management. This is consistent with previous research that shows that organisations manage health and health and safety to different standards.

The HSB trial

In preparing for this article, we put a fictitious company - albeit based on a real-life model -through the index (subscribers can look at the reports by visiting the CHaSPI site and looking at organisation number 130).

The process was easy; organisations can save their entries at any point to return later, and there are colour-coded symbols to show which sections are completed, partially completed or unanswered. In addition to the three sets of downloadable general guidance, there are two types of on-screen guidance: an underlined word or phrase, which gives a brief explanation or example; and a link to a brief guide to each of the nine indicators. Both styles afford instantaneous and concise help.

The researchers say that the two-stage piloting of the index showed that it took between half and one day to complete the index. Factors affecting this include: who completes it; the availability of a current health and safety report; and familiarity with the index. The researchers would also expect firms to conduct wider internal consultation and take more care in their responses when the index is "live". The time required to fill in the index would, however, remain valid. The HSE is already planning further research into the time required to collate the information and to gain agreement within the organisation.

As the organisations in the pilot generally had a positive approach to health and safety, they are likely to have had much of the information to hand and so it might be expected that it would take other companies longer to complete the index. Our experience was that this would not necessarily be the case. In the HSB trial, it took around an hour to download and read the three guides that come with the index and two hours to input the data. Much of the information was either known or easily accessible to us and the process was intuitive.

This reduced completion time is likely to have arisen from the index being in electronic form, which makes it quicker to complete, and the improvements that have been made as a result of the pilot phases. Furthermore, our fictitious company was office-based and contained only low hazards; had it been engaged with hazardous activities, the index would have taken slightly longer to complete. Similarly, it would have taken longer to complete had it operated from more than one site, made extensive use of contractors, or was aware of its contractors' accident performance and health and safety management. Perversely, companies with advanced health and safety systems might find it takes longer to complete because they have more information.

The HSE points out that the completion time should reduce as organisations become familiar with the process and because the electronic version will allow more than one person in an organisation access at any one time.

From our experience, minor reservations include:

The index appears to be more relevant to high-hazard activities than to office-based work, although it is still relevant to all types of work. This makes sense, given that investors want to know about major risks.

The use of "over-three-days" and "fatalities" as the descriptor of injuries could be misleading to employers used to reporting injuries in terms of fatal, major or over-three-day. It might be simpler in the final version to say either injuries that "result in absence of more than three days" or by listing the three categories.

Many organisations with proactive safety management systems find the RIDDOR categories insufficient for analysing their organisation's injury rates because they have so few RIDDOR events. Instead, they use criteria such as lost-time accidents. It will be interesting to see whether the final version of the index looks to additional injury criteria, although again it might be argued that investors do not need information on such lesser incidents. Additional criteria may also help smaller firms, most of which are not likely to have experienced a RIDDOR event.

There are some inconsistencies in the availability of a "not applicable" option to some of the indicators (see box 9).

We tried the index on an AppleMac and a PC. Although it worked well and rapidly on both, a small number of the instructions do not display fully on the Mac, mainly the arrows for going back and forth between sub-indicators and indicators and the print options. This was only a minor problem and easily solved. Greenstreet Berman states that the software was built for PCs, but subsequently tested on a Mac, and it will raise the problems with the HSE. In terms of the print option, it suggests using the print icon in the browser toolbar.

Desirable but excluded

The index excludes two sets of indicators - on finance and compliance - that Greenstreet Berman considers "desirable" but not realistic, although it hints that they could be included in the future. Claros recommended both these sets in its list of six indicators.

Financial indicators are problematic because there are significant inconsistencies in the ways in which organisations calculate the amount they invest in health and safety. For example, the amount would probably represent the total spent per employee, but few organisations collect such details and many health and safety costs are subsumed within a more general budget. Furthermore, few organisations collate valid information on the cost of accidents or ill health.

There are also problems in deciding whether expenditure is positive or a negative; an increased spend on accident investigation could indicate either an increase in the number of accidents or improved investigations. Also, organisations operating higher-risk activities tend to spend more and incur higher accident costs.

Compliance indicators are outcomes of enforcing authority action, particularly enforcement notices and criminal convictions. The researchers point out that these affect a minority of organisations, may reflect enforcement policy as much as company performance, may skew the index and "distract from the use of the index to improve performance".

While Greenstreet Berman and the HSE are correct to fear that questions on enforcement might discourage voluntary participation in the index, it is equally the case that potential investors and other stakeholders would expect to see the inclusion of enforcement information. The fact that enforcement action affects only a minority of organisations - the HSE generally restricts its prosecutions to the most serious failures - means that such information would be highly relevant. Furthermore, it is not necessarily the case that the index would pick up a major event that had led to a prosecution through answers to other indicators; the indicator of major events is limited to the previous three years, and it is possible for a prosecution to occur more than three years after an event. If an enforcement indicator was included, however, it would also need to cover the organisation's response.

The HSE stresses that information on enforcement is available through the prosecutions and notices databases on its website: "We don't want to duplicate information that is available elsewhere, but we could consider adding a link to the databases if users indicate that would be helpful."

Providing a valid comparison

The Greenstreet Berman report concludes that the index can be valid, useful to stakeholders, likely to prove an incentive for larger organisations to improve their performance and is able to provide a valid comparison of companies within and across sectors. This is a conclusion that we support.

The report draws comfort from previous research that suggests that such indicators do not need to be perfect but should be sufficiently accurate to detect significant shifts in performance. The indicators should be designed with major stakeholders in mind; an index is of little value to an organisation unless it has targets in place. Measurement is not an end in itself, but exists to stimulate improvement in performance.

The report also recognises the need for an index that avoids duplication and is compatible with other measures, draws on other financial indices used by risk analysts and investors, and recognises that many organisations operate globally. While it appears to satisfy the first three of these criteria, it explicitly confines itself to the UK (except for the sub-indicator on major hazard activities).

The HSE counters that the index was designed with UK operations in mind because that is where its responsibilities lie. Organisations might encounter difficulties completing the index for operations outside the UK, for example because of legal requirements. But the HSE has no objection to organisations trialling CHaSPI abroad so long as they treat the results with caution. Use of CHaSPI abroad, says the HSE, "is one of a number of issues that we still need to consider".

Further developments

There are clearly some areas that require further development, most of which are recommended by the researchers:

Developing industry-specific indices. The HSE says there will be considerable stakeholder support, including from investors and pension funds, for the development of industry-specific indices that build on the generic corporate index. "Although we do not anticipate a leading role for the HSE in developing industry or sector-specific indices, we will lend our support to stakeholders to assist the process."

Developing measures of work-related ill health.

Developing an index for organisations with fewer than 250 employees. The HSE points out that this was "flagged up as a key deliverable" in the government's second-stage Employers Liability Insurance review report. It is currently under development, and Greenstreet Berman is due to send the final report to the HSE in June. The HSE "is expecting a validated SME index at this point and one that stakeholders consider feasible. Though their starting point has been the corporate index, the SME index could differ significantly. It has, for example, to reflect the ability of organisations to complete it and the needs of insurers. A key result we are seeking is for insurers of SMEs to take account of health and safety performance management when assessing risks to occupational health and safety and setting terms. We have yet to consider how the SME index should be administered; we need to do this before deciding how to launch it. But assuming all goes well, autumn is likely to be the earliest possible date for its launch."

Although the index is aimed at larger companies, the report notes a problem with holding companies, particularly where they have little operational control. The report recommends further analysis of the problem and guidance as to who should fill in the form. The HSE says that holding companies might find it useful to ask their individual constituents to complete the index. Their investors, however, "may want an overall picture. We need to be able to accommodate both. There is more work to be done with the web-based tool here. Our current intention is that the index will enable companies to use it in their component businesses for comparative and benchmarking purposes."