An Overview of the Arthurs Report, Part IV: Rate Groups and Experience Rating
By Kevin MacNeill
In this entry we will look at chapters 5 & 6 of the Arthurs Report, which respectively consider the WSIB’s system of rate groups (RG) and experience rating (ER).
Before getting into the details, Arthurs’ message overall is that the RG system is broken and needs to be fixed and the ER system may have to be abolished if it cannot prove itself up to the task of encouraging employers to reduce workplace accidents and to facilitate workers’ return to work.
Although a number of the points Arthurs raises on these topics may be valid, it is far from clear that all of the points he raises are. Further, it is also not clear that the radical changes he proposes will improve on the problems he invokes to justify them.
Perhaps most significantly, it is even less clear that many of the proposed changes will have any meaningful impact on the main issue concerned by the Funding Review, namely, the WSIB’s unfunded liability (UFL).
Let’s look at these issues in a little more detail.
Arthurs’ Proposed Changes to the Rate Group System
As brief background to this topic, it may be recalled that in Ontario 30% of the workforce is not covered under the WSIB scheme. 60 % of the workforce then is employed by Schedule 1 employers, who pay insurance premiums to the WSIB as a function of the RG into which they are classified and the size of their payroll. The final 10% of the workforce is employed by Schedule 2 employers who do not pay premiums but rather are self-insured, paying the direct costs of workplace accident benefits plus an administrative fee to the WSIB.
In support of his recommendation to reform the existing RG system, Arthurs starts by noting that, currently, Schedule 1 employers are classified into 9 “industry classes” (by regulation) and 154 RGs (under WSIB policy). We may also note that RGs are further sub-divided into Classification Units (CUs), which is useful for a more granular description of what exactly the nature of an employer’s business is but is insignificant in terms of impact on an employer’s premiums, which are determined by the RG.
For Arthurs, a shortcoming of the current RG system is that although it ostensibly groups employers together by similar business types, risk profiles and cost experiences, this commonality is merely assumed and not based on any empirical assessment of what the employers in question actually do. In fact, writes Arthurs, employers in a given RG may have many different business characteristics, which may impact on accident frequency and costs. These factors are apparently not taken into account in the initial design or RGs or their later modification. Arthurs continues that in some cases “firms in related businesses but with quite different risk profiles may be intentionally included in the same RG in order to ensure that it is large enough to maintain “statistical credibility””. Further, in some cases high and low cost firms “seem to be deliberately intermingled, with the apparent intention of producing a middling premium rate, in order to appease employers protesting against high rates or to accommodate a firm with a diverse workforce that can plausibly claim membership in more than one rate group”. On these points, however, Arthurs does not provide an idea of how often these sorts of examples are found under the current RG system.
Arthurs opines that:
The overall result is that the present RG system contains many gaps and overlaps, lacks a clear internal organizing logic, cannot easily be explained or defended, and provides employers with ample opportunities for “rate shopping,” assisted by a new sub-profession of advisors and advocates – often paid (I was told) on a contingency basis… rate shopping generate needless transaction costs and results in revenue leakage for the WSIB.
Arthurs further criticizes the existing RG system by suggesting that things would be better if employers were grouped together not only for purposes of accident cost allocation but also to improve the quality of accident prevention and re-employment initiatives. Arthurs theorizes that if “good firms” which invested in safety measures were grouped together with “bad firms” who did not, the former could exert a form of peer pressure over the latter to mend their ways. Further, because harm reduction often involves collective action, Arthurs sees synergies between employers who are grouped together for safety prevention reasons as much as for cost allocation ones. Simply put, employers with common business experiences can better share information about safety and also offer better re-employment opportunities than individual employers can.
The upshot of all this are Arthurs’ recommendations 5-1, 5-2 and 5-3, which are set out on this page.
There is no denying that some of the points Arthurs raises are fair. For example, he is right to say that industry classifications like “mimeographing” and “corset manufacturing” are outdated. However, it is far from clear how what Arthurs proposes will adequately address the evils he identifies under the current RG system.
Arthurs’ new classification system, however robust in comparison to the current system, will still not fully capture the immense complexity and diversity in business organizations which Arthurs himself raises as a criticism of the existing RG system. Arthurs’ own recommendations will still leave us with about 50 to 60 “sectoral groups”, which will be described by mere words, indeed likely a few paragraphs at most for each. As a result, there will be room for interpretation as under the current RG system. Accordingly, there will still be attempts to reclassify employers, “rate shopping” and the very transaction costs that Arthurs seeks to minimize. It is not clear at all how the current system of RGs makes any material contribution to the UFL or how what he proposes will have any significant impact in reducing that.
A more fundamental concern with what Arthurs proposes is this. He is recommending that we move toward fewer and larger sectoral groups. According to Arthurs’ own analysis, a move toward larger, more heterogeneous groupings of risk spreading tends to encourage a number of “controversial consequences”. He describes them in this way:
First, employers that pay more than they otherwise might may try to persuade the WSIB to transfer them to a group with greater homogeneity, less cross-subsidization of high-cost cohort members and consequently lower premium rate. Such “rate shopping” generates administrative costs for the WSIB and, to the extent that it succeeds, contributes to a loss of revenue in the short term. Second, employers that pay less than they should may be exposed to a “moral hazard”: they may become less assiduous in ensuring safe and healthy workplaces because they can slough off part of the benefit costs they generate onto group members that are providing the cross-subsidy. And third, conflicts of interest within the group may make it difficult to organize collective activities such as safety education programs.
In fairness, Arthurs does state that what is needed is to find an appropriate balance between homogeneity/heterogeneity and levels of cross-subsidization. He sees that balanced achieved through sectoral groups. However, it is not clear how what is proposed will do a better job than the existing RG system or a slightly updated version of it.
Indeed, although Arthurs says he had to tackle the issue of RGs as part of his mandate in respect of the UFL, he ultimately does “accept the point that the issue is too complex for me to do it full justice in this report”. He continues:
The design and implementation of a new and improved system of rate groups is a time-consuming process that requires extensive technical development, intensive consultation with stakeholders and the articulation of complex transitional arrangements, including the phasing-in of any new system.
In short, although Arthurs raises a number of concerns in respect of the existing RG system and recommends that it be reformed, his Recommendation 5-4 is for slow and thoughtful change.
Arthurs’ Proposals as to Who Pays What
Arthurs correctly notes that whether his recommendations about reforming RGs are accepted or not, the WSIB will still have to determine whether and how to spread among employers the burden of new and legacy claims costs, as well as other costs including legislated costs like safety promotion.
Arthurs addresses the issue of new claims costs in Recommendation 5-5.
As for the legacy costs in the UFL, it may be recalled that earlier in his Report, Arthurs recommended that employers be required to pay a fixed annual UFL charge are part of their premium costs.
For Arthurs, this UFL cost could be spread out in a number of ways including at the level of RGs (or Arthurs’ proposed sectoral groups), at the higher level of industry classes or at an even higher level of all Schedule 1 employers.
As he notes, there are several complicating factors that weigh against assigning legacy costs to the RGs that are seen as having created them. Among others, the composition of RGs has not remained static but has changed considerably over the past few decades. Moreover, the business realities of employers within certain RGs have changed in many ways including rates of profit, working conditions, technologies and safety practices such that it would be “impossible – and arguably inappropriate – to assign responsibility for legacy costs to the firms, or groups of firms, in whose workplaces those claims originated years or decades ago.”
For Arthurs, similar difficulties exist when considering assigning legacy costs at the level of industry classes. The relative size of industry classes in the overall economy has changed over the past few decades and, likewise, so have their contributions to WSIB claims costs. In some instances, industries that contributed significantly to the UFL now have diminished capacity to pay it down, noting their reduced profitability and payroll base.
These and other considerations lead Arthurs to the observation that “The UFL is an historical omelette that cannot be unscrambled”, a reality which points him in the direction of spreading the cost of the UFL among all Schedule 1 employers irrespective of their industry class, past or current safety records.
After reviewing these considerations, Arthurs makes Recommendation 5-6, which seeks to strike a balance among them. Under this recommendation, 5% of the UFL is passed on to Schedule 1 employers as a whole, 47.5% goes to industry classes on the basis of their historical contribution to the UFL and 47.5% goes to employers on the basis of their current claims record. Arthurs suggest that these amounts be assigned to rate groups within industry classes on the same basis as their estimated new claims costs.
Finally, Recommendations 5-7 through 5-9 address other cost issues such as the WSIB’s reimbursing OHIP for routine medical expenses and funding education, accident prevention and enforcement of occupational health and safety legislation under the auspices of the Ministry of Labour.
Of particular note under these recommendations, Arthurs asks that the government investigate the possibility of charging these latter expenses to all employers instead of just to Schedule 1 employers.
Further, Arthurs notes that there is broad consensus that experience rating programs, which have resulted in rebates exceeding surcharges by $2.5 billion over the years post 1995, a considerable portion of the UFL, should be revenue neutral.
Arthurs on Employer Incentives and Experience Rating
Arthurs indicated that of all the issues canvassed by the Funding Review, this was the most contentious as between worker and employer constituencies.
Workers submitted that experience rating incents employers to “engage in improper and illegal behavior to keep their records clean, their premiums low and their workers deprived of the benefits to which they are entitled.” By contrast, employers responded by saying that incentives do not have this effect, or to the extent they do, it is only among a minority of employers. Employers also claim that incentives have led to safer workplaces, a claim which worker groups contest.
Arthurs reviewed the evidence in support of these competing views before leaping to what, with the greatest of respect, are some astonishing and unbalanced conclusions. Before reviewing these conclusions and the recommendations that come out of them, let’s take a look at the key evidence Arthurs presents.
Arthurs tells us that, on the whole, the literature reviewed supports both the worker and employer views, but both inconclusively.
Arthurs also reviews “some 50 first- and second-hand accounts of workers victimized by employers intent on avoiding surcharges or claiming rebates” including a single instance where an employer actually admitted to making false representations to the WSIB in the course of aggressively managing an employee’s workplace injury. However, he concludes that while these accounts “had the ring of truth”, they are ultimately anecdotal, uncorroborated (except the one case where the employer admitted wrongdoing) and not statistical such that “there is no way of telling whether these incidents represent most or all cases of abuse (as employers contend) or merely the tip of the iceberg (as workers believe).”
Arthurs then considers that in 2009 and 2010 “the WSIB investigated just 153 employers for allegedly suppressing or misrepresenting claims”, ultimately laying 174 charges and getting 96 convictions against 49 employers. He further notes that in 2010 the WSIB imposed 4500 administrative penalties for such things as non-reporting, late reporting or misreporting claims, but without the provable deceptive intent required for charging these employers with quasi-criminal offences. Arthurs also notes that there have been recent years where the level of administrative penalties has been higher, such as in 2008 when 11,000 were imposed. To put this in perspective, Arthurs’ Report advises that experience rating programs cover about 120,000 employers, half of those registered with the WSIB.
The final piece of evidence Arthurs invokes is an unnamed “2002 study” which suggested that unreported workplace accidents may represent as much as 40% of the total.
Although Arthurs recognizes that accidents may go unreported for many reasons and employer pressure may not explain all of the non reporting he reviewed in his study, he ultimately concludes:
Nonetheless, the overall extent of non-reporting is so large that the WSIB must surely take prompt and proper steps to satisfy itself that, if injured workers are foregoing their benefits, the reason is something other than employer misconduct.
And further:
In my view, the WSIB is confronting something of a moral crisis. It maintains an experience rating system under which some employers have almost certainly been suppressing claims; it has been warned – not only by workers but by consultants and researchers – that abuses are likely occurring. But, despite these warnings, the WSIB has failed to take adequate steps to forestall or punish illegal claims suppression practices. In order to rectify the situation, the WSIB must now commit itself to remedial measures that might otherwise require more compelling justification. Unless the WSIB is prepared to aggressively use its existing powers – and hopefully new ones as well – to prevent and punish claims suppression, and unless it is able to vouch for the integrity and efficacy of its experience reading programs, it should not continue to operate them.
Flowing from these observations, Arthurs makes the extensive and, in some cases, draconian, Recommendations 6-1 and 6-2, which are set out below in full. Following that, we will briefly go over Recommendation 6-3 prior to returning to some concluding observations.
Redesigning the WSIB’s experience rating programs
Arthurs remarks that even if the WSIB adopted all of his recommendations above, which seek to stop claims suppression, it should also look seriously at redesigning its ER programs because, according to Arthurs, it is not really known whether they are effective in reducing accidents and encouraging the return to work of injured workers.
After reviewing a number of perceived design flaws in existing ER programs and making the point that ER should become an integral part of the rate setting process, Arthurs outlines a number of ways in which ER programs could be modified. A non exhaustive list of some of the proposed changes which stand to have a considerable impact on employers are:
- Making experience rating for larger employers prospective rather than retrospective. Under a prospective system, there would be no adjustment to an employer’s costs based on a hindsight look at what its actual claims cost experience was, as is the case under NEER currently.
- The review window for experience rating would be extended to 6 years (currently 4 under NEER).
- SIEF would be abolished or replaced with a program of wage subsidies for injured workers seeking to return to work with their original or another employer.
Arthurs also suggests that the WSIB should conduct a controlled experiment with a new ER model to determine if it achieves the goals of ER, in which case, Arthurs suggests the model should be rolled out across industry classes.
Arthurs’ final recommendation on experience rating is of particular note and hopefully will be abided by the WSIB: extensive consultation of stakeholders should occur before any redesign of experience rating or other incentive programs.
Concluding Observations on Experience Rating
Near the outset of his Report (at p. 20) Arthurs recognizes that a potential contributing factor to the UFL is that the WSIB may have “adopted excessively generous, even improvident, policies for determining entitlements as well as loose claims management practices.” He further recognizes (at p. 85) that “workers as well as employers may engage in fraudulent conduct in connection with the reporting of claims.” Near the end of his Report he also states that “many legitimate questions were raised about the WSIB’s expenditures” but he ends up stating that these sorts of questions did not fall within his mandate.
What an odd thing to say after recognizing the possible connection between these issues and the UFL, which was at the core of his mandate.
Against this background it is striking that Arthurs seems to focus most of his energies and recommendations on perceived employer abuses of the system.
Arthurs also appears to favor a punitive model for employers, as opposed to an educative, admonitory one. Yet he has not explored in detail which of these two models would be more effective in stemming perceived employer under-reporting or misreporting of claims and which would best encourage improvements in return to work. This is not to say that there is no place for punitive sanctions in appropriate cases. However, it is not demonstrated in Arthurs’ Report that emphasis on a punitive approach is warranted, especially one which appears to be focused on employer wrongdoing to the exclusion of worker wrongdoing.
Arthurs also has not explored the full potential consequences of what he proposes.
To say nothing of the fact that an aggressive enforcement model can itself lead to abuses of employer rights by over-zealous enforcers, a number of aspects of Recommendation 6-2.3 deserve special mention because they appear to be particularly harsh, and arguably unduly so.
The first of these is the recommendation that :
Employers found to have violated the WSIA or other occupational health and safety legislation should be automatically ineligible for favorable premium adjustments or rate rebates, for at least one year and for any additional period up to five years, as determined by a claims adjudicator, tribunal or convicting court that makes a finding that such a violated has occurred.
The second is the recommendation that the WSIB should be given enhanced power to impose administrative penalties for violation of workers’ rights to benefits or return to work, with such penalties being up to three times the employer’s annual premium rates and graduated to reflect mitigating or aggravating factors.
It may be noted first of all that, for many employers, the amount of such penalties could well go into the millions of dollars, a particularly heavy-handed outcome when considered in relation to the amounts imposed for violations of other workplace legal norms.
The recommendations with respect to having to reimburse the WSIB and/or worker for such things as legal costs , travel, relocation and medical expenses are also simply unheard of in other employment-related administrative legal proceedings such as those before the Human Rights Tribunal of Ontario, Ontario Labour Relations Board or even in prosecutions or s. 50 reprisal complaints under the Occupational Health and Safety Act.
Arthurs may have exposed some evidence of undesirable social results in relation to ER (claims suppression). But he also found evidence that ER may produce desirable outcomes (reduced accidents, increased return to work outcomes). He was not able to say that the evidence comes down more on one side than the other and ended up having to recommend simply that the issue be considered more closely.
If the jury is still out then on whether ER is working and whether employer abuses are as rampant as some say they are then certainly it is premature to leap to the extreme punitive measures which Arthurs also recommends.
This is all the more so when one recalls that there is another side of the equation which Arthurs notes but which he has not pursued with the same vim and vigour: that of worker abuses of the claims system and lax claims management by the WSIB.
The ultimate conclusions in relation to ER are thus somewhat unbalanced and not placed on the most solid footing.
It may well be that the old days of ER are quickly coming to an end. However, it is to be hoped that the WSIB will think these issues through and with the full consultation of the employer community before implementing any drastic changes to the existing ER system.
In our next and final entry in this series, we will review the Arthurs Report’s consideration of funding occupational disease claims and benefit indexation for partially disabled workers.