BENDING THE RULES II

Why do people break rules or fail to follow procedures? and What can you do about it?

The Vtolatlon Manual

P T W Hudson W L G Verschuur
Leaden UnwersltV

R Lawton D Parker J T Reason

Manchester Unwerslp

Vlolauon Handbook

Vewon 1 2

Contents
0 Part 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 3

Management Summary

I Introduction Sclentlfic Approaches Remedies The Theory The Behavloural Cause Model Remedlatlon Dlagnoas-Remedy Matrices Conclusion

5 5 5 6 6 8 9 10 13 14 14 14 15 16 17 19 20 21 22 24 24 24 24 27 27 28 30 30 31 31 32 34 36 38

Part II 2 0 Detatled Introduction 2 0 1 The Nature of the Problem 2 0 2 The Am of the Manual 2 0 3 Procedures 2 0 4 Natural Born Violators Wolves m Sheeps Clothmg 2 2 2 2 1 Why IS Vlolatlon a Problem7 - Human Error 1 1 Human Error and Danger 1 2 The Types of Vlolatlon 1 3 Why IS Vlolatlon so Dangerous? What Should You Do3 Proactive and Reactwe approaches What can be done to make thmgs better? The Behavloural Cause Model Testing the Model An Altematwe Model - Detection and Pumshment The Manual - Structure

2 2 22 1 2 2 2 2 2 3 224 2 25 2 3

Part Ill 3 0 The Proactive Approach - The Diagnosis-Remedy Handbook 3 1 Diagnosis Section 3 1 1 Vlolatmg Behavlour 3 1 2 lntention 3 1 3 Expectation 3 1 4 Plannmg 3 1 5 Attitudes

Vlolauon Handbook

Version 1 2

316Not-m 3 1 7 Feelmg of Control 3 1 8 Oppoltumty 3 1 9 Poweriklness 3 1 10 Powerlessness 3 3 3 3 3 3 3 3 3 3 3 3 3 3 2 The Remedies 2 1 Dlvlslon I Remedies 2 2 1 Supervlslon 2 3 2 Analysis of Exlstmg Vlolatlons 2 4 3 Analysis of Pre-condltlons for Vlolatlon 2 5 4 Structured DIscussIons 2 6 5 Procedures 2 7 6 Du-ect Management 2 2 Dlvlslon II Remedies 2 2 1 7 Selection 2 2 2 8 Trammg 2 2 3 9 DetectIon 2 2 4 10 Reporting 2 2 5 11 Incentives

40 43 44 46 48 49 50 50 51 52 53 55 56 57 57 58 59 60 61 62 62 63 64 66 68 69 70 71 72 72 72 73 75 77 79 81 83 85 86

1 r-

3 3 Remedies and Problem Areas 3 3 1 Vlolatmg Behavlour 3 3 2 Intentton 3 3 3 Expectation 3 3 4 Plannmg 3 3 5 Attitudes 336Noms 3 3 7 Feehng of Control 3 4 Part 4 0 4 1 42 43 44 4 5 46 4 7 Summary of Dlagnosls-Remedy Approach IV Introduction to the Reactive Approach Identlfymg the Problem Unmtentlonal Violations Routme Vlolatlons Situational Vlolatlons Optimismg Violations ExceptIonal Vlolatlons Summary of the Reactive Approach

I
F1

5 0 References 6 0 Appendix Reactive Test Items

Vlolatlon Handbook

Version 1 2

Bending the Rules II

Why do people break rules or fail to follow procedures? and, What can you do about It?

Management Summary
Why do people bend the rules, dehberately fall to follow procedures? There IS an assumption that people ~111 follow the guldelmes and procedures as laid down m the Safety Management System When this assumption 1s broken the whole basis of the Safety Management System 1s temporanly put at nsk The mam reasons for vlolatlon are 1 Expectation that the rules w111 have to be bent to get the work done, 2 Powerfulness, the feelmg that one has the ablllty and experience to do the Job without slavishly followmg the procedures, 3 Seemg the Opportunities that present themselves for short cuts or to do thmgs better, 4 Inadequate Work Planning and advance preparation, leading to workmg on the fly and solvmg problems as they anse If you know how someone ~111 answer questions about those four factors, you can make a very good predlctlon of whether they are likely to bend the rules or not On the other hand, having good mtentlons, bemg closely watched over for rule-breakmg and having the threat of pumshrnent when caught are not good predictors The reason for ths 1s that, wlthm Shell, these problems are solved, at least they have been solved to the pomt that extra effort here ~111 not help a lot In the North Sea, Shell offshore workers have good mtentlons about not breakmg rules, but end up domg so anyway because of the other factors One factor that helps IS havmg strong personal norms, values, that may be translated as behevmg m good engmeenng practice Thus manual provides two ways to combat vlolatlon, proactzve and reactzve The reactzve approach mvolves finding out what rules are bemg broken and fixmg the reasons why that IS happening
l

TheproactIve approach concentrates upon finding out whether the scene IS set for vlolatlons, even If they are not happemng nght now, and ensunng that problems dont anse
l

Because most people do not find themselves breaking the rules that frequently, It IS often better to concentrate upon the proactive approach The field studies have found that the majority of Shell workers offshore m the North Sea can be described as Natural Born Vzolators or WoZves However, at the same time, most people dont report that much vlolatlon, and the majority have broken few rules m the previous SIX months, so the Wolves are Wolves tn Sheeps Clothing Wolves are opportunists, lookmg for advantage and takmg
3

Vlolauon Handbook

Version

1 2

it, usually for what they see as the benefit of the company rather than Just personally Managers are most likely to be Wolves and may often condone vlolatlon, if it IS success&l The approach taken here IS that advantages are there to be taken, but rule-breakmg or bending IS so dangerous that the process has to be carefUlly managed Most vlolatlons are, m fact, preventable Thus can be done by either changing the rules or avoiding the problem What remains must be managed by recogmsmg the types of people who are performmg the work - using the theory - and by understanding why violations are so dangerous The recommended remedies for violation problems are
l

l l

Supervision - watch more closely over people, plan their work, provide an example Analysis of existing violations - find the background behmd vlolatlon behavlour Analysis of vrolatzon potential - use the theory to see If people will find vlolatmg easy Structured discussions - discover what IS going on and get people to agree on actions Procedures - improve the quality of procedures so vlolatlon 1s not seen as necessary Direct management - a techmque to vary the procedures as circumstances demand

Other possible remedies are also covered, but these are not recommended because they may be less effective or have slgmficant negative side-effects Any one remedy will only be effective for some causes A Diagnosis-remedy matrix IS provlded showmg whch approaches are effective and what side-effects may be expected This manual starts m Part I with a simple descnptlon of the problem and the solutions proposed The diagnosis-remedy matnx forms part of this bnef overview The remainder of the manual 1s optional, wntten for people who want to go finther Readers intending to apply the diagnosis and remedy techniques are recommended to read those sections Part II provides a more detailed mtroductlon to the theory for understanding vlolatlon and the proposed remedies Part III documents the proactlve approach There IS a section detailing each contnbutmg factor to help diagnose whether there 1s a problem with e g , expectation or powerfulness, and a section with associated remedies These form the background to the diagnosis-remedy matnx Because of the complexity these sectlons are internally crossreferenced Part IV documents the reactive approach To enable readers to restrict themselves to single sections, there 1s some degree of redundancy m the text to allow each section to be stand-alone The knowledge m thus manual 1s based upon field studies carried out by Leaden and Manchester Umveratles for SIEP, m co-operation with NAM and Expro In thus study a theory was developed to understand deliberate failures to follow procedures The theory has been tested on offshore operators, supervisors and contractors m the North Sea and found to be accurate to predicting more than 2/3rds of vlolatmg behavlour Alternative models, slmllar to the beliefs of managers and supervisors, have been found only to predict 20% of vlolatmg behavlour The results apply to the North Sea operations and, probably, m slrmlar cultures elsewhere m the world While the theory 1s general, the relative importance of specific factors will depend upon the local culture

Vlolatlon Handbook.

Version 1 2

Part I Bending the Rules II

Why do people break rules or fail to follow procedures? and, What can you do about it? 1.0 Introduction Why do people bend the rules, dehberately fall to follow procedures7 There 1san assumption that people ~111 follow the guldelmes and procedures as laid down m the Safety Management System When this assumption IS broken the whole basis of the SMS IS put at risk The mam reasons for violation are 1 Expectutzon that the rules will have to be bent to get the work done, 2 Powerfulness, the feeling that one has the ability and experience to do the Job without slavishly followmg the procedures, 3 Seeing the Opportunztzes that present themselves for short cuts or to do thmgs better, 4 Inadequate Work Piannq and advance preparation, leading to workmg on the fly and solvmg problems as they anse If you know how someone will answer questions about those four factors, you can make a very good prediction of whether they are likely to bend the rules or not On the other hand, havmg good intentions, being closely supervised and having the threat of pumshment are not good predictors The reason for this ISthat, wlthm Shell, these problems have been solved, at least so far that extra effort here ~111 help a lot In the North Sea, Shell offshore not workers have good mtentlons about not breakmg rules, but end up doing so anyway because of the other factors One factor that helps 1shavmg strong personal no?mA,values, that may be translated as behevmg m good engineering practice 1.1 The Scientific Background This knowledge 1sbased upon field studies camed out by Leaden and Manchester Umversltles for SIEP, m co-operation with NAM and Expro Fom those studies a model, ongmally proposed by Qen and Flshbem (1984) as the Theory of Planned Behaviour, was developed and greatly extended to understand deliberate failures to follou procedures The theory, now called the BehavIoural Cause Model, has been tested on offshore operators, supervisors and contractors m the North Sea and found to be accurate to predicting more than 2/3rds of vlolatmg behaviour The results apply to the North Sea operations and, probably, m smular cultures elsewhere m the world While the model IS general, the relative importance of the specific factors WIII depend upon the local culture

Vlolatlon Handbook

Version I 2

1.2 Approaches to Violation This manual provides two ways to combat vlolatlon, proactwe and reactwe
l

The reactzve approach mvolves findmg out what rules are being broken and fixmg the problem The proactwe approach concentrates upon finding out whether the scene 1sset for vlolatlons, even if they are not happenmg nght now, and ensuring that problems dont anse

Because most people are not breakmg the rules that frequently, It 1soften better to concentrate upon the proactive approach The studies found that the majority of Shell workers offshore m the North Sea can be described as Natural Born Vzolators or Wolves However most people dont report that much vlolatlon, so the Wolves are Wolves m Sheeps Clothmg Wolves are opporturusts, lokmg for advantage and takmg it. usually for the benefit of the company rather than Just personally The approach taken here to vlolatlon ISthat advantages are there to be taken, but rulebreakmg or bending 1sso dangerous that the process has to be carefully managed Most vlolatlons are, m fact, preventable either by changing the rules or avoiding the problem What remams must be managed by recogmsmg the types of people who are performmg the work - using the theory - and by understanding why vlolatlons are so dangerous 1.3 Remedies The recommended remedies for vlolatlon problems, called Dlvlslon I remedies, are
l

Supervision - watch more closely over people, plan their work, provide an example Analjsis of existing violations - find the background behmd vlolatlon behaviour Ana@sis of violation potential - use the theory to see if people will find vlolatmg easy Strucfured discussions - discover what IS gomg on and get people to agree on actions Procedures - improve the quahty of procedures so vlolatlon 1snot seen as necessary Direct management - a techmque to vary the procedures as circumstances demand

Less effective, and therefore less recommended remedies, called Dlvlslon II remedies, are
l

Selection - pick people who wont be tempted vlolate to start with Training - tram people more on the real nsks and the possible consequences - sharpen the lookout for vlolatlons Reporting - encourage people to report when they or others bend the rules Incentives - pay for compliance or purush people who break the rules
Detection

The reason why remedies are placed in Dlvlslon II are because they are less effective In combatting vlolatlons and they often have side-effects that negate the posltlve benefits

Vlolatlon Handbook

Version 1 2

Selection of a compliant workforce, for Instance, means that workers are more likely to follow incorrect procedures Studies in the US nuclear Industry have found that the majority of procedural problems arose from people followmg bad procedures, not for falling to follow good ones The survivors from Plper Alpha were those who did not follow the procedure Those who went to their muster statlon and waited, died Selectmg comphant workers therefore places more pressure on the plannmg process to ensure that problems do not anse m the first place Trarnmg 1sessential m competence-based workforces. but possesslon of extra knowledge 1sone of the major factors behind powerfulness, one of the most Important causes of vlolatlon DetectIon and reporting can create strains and require a strong and blame-free safety culture While these strategies do not create problems, they are less effective than the Dlvlslon I remedies at tackhng the same problems Fmally mcentlves are never very effective on their own and can have serious negative side effects when handled mcorrectly

Uolatlon Handbook 1.4 The Theory: The Behavioural Cause Model

Verston1 2

Figure 1 shows the Behavroural Cause Model (BCM), which was developed dunng the research programme to explam why people vlolate l People behave the way they do because they plan to do so l Plans are made up to take account of 1) external requirements, the work to be done, rewards and supervIsIon 2) the Intention to do thmgs according to the book. 3) the expectations that yet agam comers will or will not have to be cut, 4) the existence of opporturutles to get things done quicker or better l IntentIons and expectations are determined by attitudes to work and to the vlolatlon of procedures, by past experience, by social norms and by feelings of being rn control or of being out of control of ones own actlons l The loss of a feehng of control can mean that previously well thought out plans are shortcircuited by events and condltlons l Motlvatlon IS seen as a general factor, makmg things better or worse, a well motivated person IS more hkely to see an opportumty, and bend the rules to take It, less highly motivated people are less hkely to be affected by norms and attitudes to do thmgs properly

U
$< @q

oweriessness

Fgure I The Behawoural Came Model

Vlolatlon Handbook

Version 1 2

1.5 Remediation Vlolatlons arent all bad, they are the exercise of mltlatlve when they are successful, but they are dangerous because violators always assume everyone else IS keeping strictly to procedures If someone else makes an error or somethmg goes unexpectedly wrong, then the vlolatlon can become extremely dangerous, otherwise it can be very advantageous This means that simply trying to stop rule-bending ISto fall to understand a complex phenomenon The pay-off for successwhen rule-breaking, taken together wth the perceived chance of success, 1smuch greater than the possible bad consequences, especially when people strongly believe that It wont happen to them A number of different approaches to stoppmg or managing vlolatlon are avallable The solution to tackling the diseaseof vlolatlon 1sto seek out the appropriate remedies and apply them rigorously The remedies can be dlstmgulshed mto Fzrst and SecondDwzszonremedies First Dlvlslon remedies are the most effective Second Dlvlslon remedies may be effective but certainly have nasty side-effects Most people tend to prefer a lot of Second Dlvlslon remedies, but the field studies have shown that these are less effective One of the reasons why vlolatlons m the workplace are so hard to manage ISthe tendency for managers and supervisors to beheve that the Dlvlslon II remedies are the most appropnate, based upon folk psychology rather than sclentlfic data The behef that mdlvldual charactenstlcs are the mam cause of rule-breakmg behaviour leads to searching for remedies m areas such as selection, trammg and incentives The facts are that the real causes predommantly he outslde the mdlvldual, m poor planning, fatlures to tmprove procedures, the existence of oportumtles The major mdlvldual characterlstlc that 1simportant, powerfulness, 1sJust the one those managers and supervisors probably possessmost strongly - whch IS why, when the opportumty arises, it 1sJust those people who exercise lmtlatlve, by bendmg the rules

Vlolatlon Handbook

Version 1 2

1.6 The Diagnosis-Remedy Matnces Dlvlslon I

l

Supervision - watch more closely over people, plan their work, provide an example Analysis of existing violations - find the background behmd vlolatlon behaviour, and fix it Analysis of violation potential - use the BCM to see If people ~111 find vlolatmg easy,& fix the problems found Structured discussions - discover what ISgomg on and get people to agree on actions that get carried out Procedures - improve the quahty of procedures so vlolatlon IS not seen as necessary Direct management - a techmque to vary the procedures as circumstances demand Dlvlslon II

Selection - pick people who dont vlolate to start wtth Training - tram people more on the real nsks and the posstble consequences Detection - sharpen the lookout for vlolatlons Reporting - encourage people to report when they or others bend the rules Incentives - pay for comphance or purush people who break the rules

Of these Selection and Trammg can work on Attitudes and Incentives can affect Personal Norms However these can all have nasty side-effects, as selectmg more comphant workers requires that the plannmg process be considerably improved Another side-effect of tranung 1sthat It may increase the feeling of powerfilness, makmg vlolatlon more rather than less likely Analysis of exlstmg vlolatlons IS effective, but vlolatlons are actually not so common, so analysis of the condltlons IS more hkely to prevent problems m the future Structured dlscusslon groups can find out what needs fixed first Fmally, once the most common routme vlolatlons have been swept away, a rigorous management approach to unusual sltuatlons can be apphed instead of having people takmg dangerous lmtatlves

The matrlces below represent notes on the apphcablhty, or otherwlse, of the different possible types of remedies (rows) as applted to the different causes of problems (columns) These notes are very summary, and are Intended to be taken m conJunctIon with the more detailed mformatlon available In Part 3

10

V~olmon Handbook

Vcrsloll

Expectation Supervision
SupervIsors need to walk the talk about Improvements Fmd out what IS wrong, fix It and tell people Primary mechanism for fixmg thrs problem Fmd out what IS wrong Identifies which other factors need to be addressed

Powerfulness

Division I Remedies Opportunity

Can be recogmsed earlier by good supervIsors, who should take them up in the planning

Planniug

Personal Norms
Set an example of professionalism to the workforce Teach young workers by example

Contmenls

Analysis of existing violations Analysis of violation potential Structured discussions

Requires bemg seen as a problem Supervisors are often the biggest problem because of their experience Can uncover whether feelings of powerfulness IS a root cause Clear ldentlficatlon of problems and mdlvlduals m advance

Supervrsors major task, rather than watching over people A good supervisor manages Opportunltles and Are vlolatlons a improvements once response to poor Identified can be planning? Is this procedurahsed common practice? Opportunltles need Dlstmgulsh which to be sought for and part of the taken up m the planning process needs to be fixed planning process

Procedure improvement

Effective If carried through fully Requires action after the talk Crucial to the whole Needs people to do approach Can solve It the problems m one Crucial and effective Allows powerfulness Crucial

Sets social group norms m public

Direct management

Plans and procedures need to be matched to each other Allows mltlatlve Alters planmng process on-line

Requires active work by supervisors, not JUSt standmg back and pohcmg Less effective if there are few active violations Requires followmg the proactive manual Can set prlorltles and achieve ownership Needs to be seen to be bemg done Must be very carefully Introduced

II

Vlolauon Handbook

Version 1 2

1.7 Conclusion The message IS clear, management of violation requires understandmg of the slgmficant preclpltatmg factors rather than being based upon preconceptions untested m practice Concentrating upon detectIon and supervIsIon will only produce margmal improvements, concentrating upon the major factors m the BehavIoural Cause Model can produce mayor improvements The effectiveness of different remedial measures can be assessed,to allow prlontisation on the basis of data rather than impression and prejudice If you want to know if people are gomg to wolate, dont ask what they intend to do, we know that answer Ask them what they expect they ~111 have to do If they say they may well have to break the rules, ask why and ask why nothmg has been done about it so far The remammg ingredients of the lethal cocktail are a feehng of absolute competence (m the face of impossible procedures and situations), poor plannmg of the work to be performed, and the recogmtion, usually by well motivated personnel, of opporturutles Do not expect that purushment ~111 an effective solution, most of the violators m the study were tqmg to get be the Job done, on time or faster, m the face of procedures that seem to them to be impossible to follow m the real world

13

Vlolauon Handbook

Version 1 2

Part II 2.0 Detailed Introduction 2.0.1 The nature of the problem Vlolatlons are devlatlons from the rules, procedures, mstructlons and regulations developed for the safe and efflclent operation (or maintenance) of equipment, plant etc Deviations from good practice, even when not laid down formally, may also be regarded as vrolatlons Breaches m these rules can be either unmtentlonal or dehberateThe Importance of vlolatlons m mdustnal safety was brought to the fore after the Chernobyl accident Here, of the 7 human actions that led dn-ectly to the accrdent, 5 were deliberate devlatlons from wmten rules and mstructtons rather than shps, lapses or mistakes (Reason, 1987) An exammatlon of railway accidents m Bntam between 1989 and 1992 (Free, 1994) revealed that vlolatlons play a considerable role m accidents to staff, e g personal 1nJut-y fatalities Many and accidents wthm Shell Group are caused, at least m part, by vlolatlons cormmtted by one or more people Vlolatlon IS one of the maJor causes of accidents m Industry The deliberate failure to follow known procedures 1scalled a vlolatlon, when it goes wrong, but may also be called the exercise of lmtlatlve, especially when It succeeds Because of tlus paradox, vlolatlon can not be stopped simply by telling people not to do It Vlolatlons m whrch a negative outcome 1s mtended, such as wllfbl acts of sabotage or vandalism, are beyond the scope of this manual Research mto understanding the nature of vlolatlon has confirmed how much vlolatlon IS done by those who are highly motivated Vlolatlons m the workplace occur for many reasons, but, contrary to expectations, are usually the result of well mtentloned staff, attempting to get theJob done, rather than, as rmght orlgmally be expected, by those who are only interested m their own comfort (Verschuur et al , 1996) The mtroductlon of Safety Management Systems, now extended to HSE Management, has led to a slgruficant increase m the assuredsafety of workers and guaranteed asset mtewty Many of the controls put m place wthm the context of the SMS are adrmmstratlve (see sectlon 4 7), that IS they rely upon procedures, trammg and management processes There IS, necessarily, an assumption that people WIII follow the guldehnes and procedures as laid down m the SMS When tlus assumption 1sbroken the whole basis of the SMS IS momentarily put at nsk This manual IS Intended to help manage that nsk

14

VlolauonHandbook 2.0.2 The aim of the manual

Version1 2

The aim of this manual IS to provide managers wth step-by-step guides to understanding the grounds for vlolatlon and to reducing ldentlfied types of vlolatlons m the workplace Two approachesare taken aproactwe approach and a reactwe approach
l

The proactive approach IS basedupon a theory of why people behave the way they do and why, therefore, they sometlmes fall to follow the rules This allows the dlagnosls of problem areas and the defimtlon of appropriate remedial plans, whch can be seen m a matnx structure The reactive approach IS based upon the existence of vlolatlons and 1sset m a five step approach Thus five step approach to lmprovmg safety by encouraging procedural comphance 1sbased on a classlficatlon of vlolatlons outlined m a previous Shell report (Reason et al, 1994) and described again below This classlficatton was first put forward by Reason ( 1990) and was supported and extendedby the work of Free (1994) Zt IS a particularly usehI classlficatlon m terms of remedlatlon, becauseIt 1scausally based

This detailed mtroductlon forms Part II of the manual It introduces vlolatlon and sets vlolatlons mto the context of Human Error The concept of Natural Born Vloiators or Wolves servesto introduce the idea that vlolatlon IS a corn wth two faces, the other of whch 1suutlatlve Part II explains why vlolatlons are dangerous, sets out vlolatmg behaviour m the context of tnherent and normally safe operation and, finally, presents a theory - The Behavioural Cause Model - which illustrates the different psychologlcal factors, and the external factors, that lead to people either vlolatmg or, m many cases,not bending the rules despite pressure to do so

In Part III the proactive approach 1sfollowed There 1sa dragnostzc and a remedzal section Each of the components from the The Behavioural Cause Model 1spresented as a short descnptlon, so defined that a manager, supervlsor or operator, reading the defimtlon section, can decide whether thts forms a problem m their worksite A checkhst for dlagnosls 1s embedded mto the text for each component If there IS a potential problem then they are led to the root causes The problems so identified can then be looked up m the appropnate remedial section

The reactive approach 1sfollowed m Part IV Here a checklist IS also embedded mto the descrlptlve text whch allows the ldentlficatron of the types of vlolatron causing most concern, followed by appropnate suggestions for remedy There IS no extra support or dlagnostlc software This manual representsboth the documentation on vtolatlon m the workplace, and the means for dlagnosls and sets of suggestlons for remedy or management As such It takes the 80 20 approach, trading off accuracy against easeof use All the proposed remedies form best practice, so whle overeager apphcatlon of the dlagnosls sections may create work, It can never lead to harm

15

Uolatlon Handbook

Version 1 2

2.0.3 Procedures Without procedures and clearly defined guldelmes, there would be no vlolatlons lmprovmg the quality of the procedures IS one of the essential solutions to the problem of vlolatlons What the field studies have shown, however, IS that peoplesperceptlons of procedure quality are a poor predictor of whether there are problems or not Nevertheless no procedure IS perfect, not all situations have been thought of m advance But, either thmkmg that there are problems or feeling that the procedures are good both turn out to be rmsleadmg This has to be seen m the context of the current high level of quality of procedures m the North Sea In any operating umt where much work remams to be done ttis restnctlon on perceptlons may not apply There are three procedure problems that have to be identified 1 Poor procedures that are not followed - Routine Violations 2 Special situations that are not or poorly covered m the procedures - Situational Violations 3 Sltuatlons that are not covered by any procedures - Exceptional Violations The reasonswhy the perceptions and the actual quality are not well correlated are complex In the case of many routine vlolatlons they may have become so routme that not followmg the procedures no longer contributes to the feeling that rules are being broken With exceptional and sltuatlonal vlolatlons, perceptlons of the adequacy of the procedures are determmed by remembermg when one last broke the rules oneself, not by the quality of those rules Finally the cellmg effect, with procedures generally very good, means that the reasonswhy people feel happy or unhappy about their procedures, m general, are very ldlosyncratlc There IS one other procedure problem, not leading to vlolatlons, but still dangerous This IS when people persist m followmg incorrect procedures Studies m the Amencan Nuclear Industry have found that the maJonty of procedure problems are of ths sort A typical example IS a procedure that requires testing a valve wth the mstructlon Open and close the valve, people often stop at this point, without returning the valve to the open position which was the mtentlon of the procedure wnter (the open/close operatron was a quick check that the valve works, thmgs should be left as they were afterwards) Procedures can be seenas operating at one of three levels 1 Guldelmes - requirements for safe and efflclent operation 2 Work Descnptlons - definition of work to be camed out wrthout the specific requirements as embodled m work rnstructlons 3 Work InstructIons - step for step defimtlons of work to be performed, usually obligatory, sometimes legally required People often co&se levels, regarding detailed mstructlons (level 3) as descnptlons (level 2), and followmg descriptions without also refemng to the general guidelines (level 1)

16

VlolaoonHandbook

Verslon 1 2

The most common reasons for changmg procedures are reactions to mcldents This almost mvarlably results m more comphcated procedures, with extra checks and balancesbuilt m Many procedure problems anse becausethere IS not enough time to perform all the actlons required, checks are redundant or seen as unnecessary.or people feel that they know of a better way of doing the task The solutions to procedure problems mvolve systematic mvestlgatlon and reahstlc improvement As improvements often led to the mltlal problems, procedures have to be made so that they are capable of bemg compiled with In many cases procedures have acquired extra steps wlthout conslderatlon of the time neededto perform the procedure as a whole Optional steps may have acquired an obhgatory status that IS mappropnate for the probablllty of the outcome that 1ssupposed to be prevented Procedure problems arise when the official procedure, defined off-site, loses touch with the workmg reality Effective supervision and direct management of vlolatlons form a partial, short-term solution to such problems Long term solutions require systemattc analysis and, where necessary,re-wntmg of the procedure so that comphance IS easier than non-comphance

2.0.4 Natural Born violators - Wolves in Sheeps Clothing While it rmght seem obvious to start by lookmg for vlolatlons, fixing the causesand therefore feelmg satisfied that the problem has been solved, a study offshore (Hudson & Verschuur, 1995) suggested that even m the absenceof wolatlons the ground may still be ripe for vlolatlon behaviour to stnke as a bolt from the blue Askmg people offshore about then- actual hlstory of wolatlon, and analysmg the data mto dlmenslons which suggest whether mdivlduals are hkely to violate if the circumstances arise, led to the discovery of two dimensions One dtmenslon was Sheep KS Wolves Sheep are, approximately, those who do not like vlolatmg, they feel unhappy wth their own behaviour even if they feel compelled by circumstances to bend or break the rules (see Types of Vlolatlon below) Wolves, m contrast, have no such problems The second dlmenslon was Sheep s Clothq vs Wolves Clothrrtg, how do people look to others Thus second dlmenslon reflects whether people have actually found themselves rule-breakmg or not m the recent past People who vlolate look hke wolves, those who do not look like sheep, appearancescan be deceptive Another way of expressing ths 1sto thmk m terms of Natural Born Violators, people who are naturally opportumstlc and who have no great problem with bendmg or breakmg (unnecessary)rules if It semes their purposes

17

Vlolatlon Handbooh

Vernon 1 2

The study (Hudson & Verschuur, 1995) found four groups

l

The first group, Sheep m Sheeps Clothing (up-front sheep) were 22 5% of the workforce responding Thusfirst group represents the guardians of standards l The largest group, with 33 8% are Wolves m Sheeps Clothmg, they have not violated, yet, but will not have great problems with It when they do l The smallest group were Sheep m Wolves Clothmg (I 4 l%), who had violated, but werent happy with this themselves l 29 6% were obvious Wolves, who reported vlolatmg and had the characteristics of violators These results mean that m the study 56% of respondents did not report a major hstory of vlolatlon, but that 63 4% of the population studied were Wolves, whAe only 43 7% reported having violated recently

Every orgamsatlon needs both Sheep and Wolves Sheep are the guardians of high standards, people who manage cntlcal actlvltles and who are not afraid to shut down whatever pressures rmght be brought to bear Wolves see opporturutles and grasp them, mterpretmg procedures more as guidance than as rules The end Justifies the means, for Wolves, but not for Sheep An orgamsatlon composed totally of Sheep may go bankrupt, while an organisation of Wolves IS likely to go bang! Both are needed, the art of the manager IS to balance the two, to ensurethat the need for actual vlolatlon never arises and to create condltlons m which mmatlve becomes productive and not dangerous (reallang that probably all managers are Wolves, or wont adnut to anythmg else) The existence of all the Wolves means that even wlthout obvtous rule breakmg, the grounds for vlolatlon may well be present The fact that they form the majority of employees m the 011and Gas industry m the North Sea only highlights the need for timely and effective management The proactive approach 1sIntended to allow the diagnosis of those grounds and the constructlon of an effective plan of actlon

18

Vlolauon Handbook

Vernon 1 2

2.1 Why is violation a problem? Human Error

Vtolatlon IS one of the four major forms of human error It IS dlstmgulshed Corn the more ordmary forms of error becausethere IS the intent not to follow the rules, both the action and the specific behaviours are intended, unhke rmstakes where the actlon may be Intended, but the behaviour IS unmtended m the hght of the possible outcomes

A vzolatzon 1sa sort of mistake, m that few people actually Intend thrngs to come to harm (In our settings at least), so what actuaily happens IS also not intended The behaviour IS qurte delrberate, and the actrons that make up that behaviour are also Intended Behlnd most vrolatlons, however, the?e IS strll a good wrll Crmunal behaviour IS characterised by a lack of good wrll, as well as Intended behaviour and actlonJ Most forms of human error are not only unmtentlonal, they can also be detected and recovered from T~K means that someonewho makes a simple slip or even a lapse can often detect that they have done so and take appropnate actlon When you tahe a wrong turnmg off a road, ths 1susually quxkly obvious and the route to recovery simple Mstakes are harder, but even here It may be possible to put thmgs nght once someone discovers that they are actmg nustakenly A nustaken choice of route to avold a trafficJam may turn out eventually to be Just that, a horrible rmstake as you sit m an even larger queue, but even then some recovery IS possible even if detectlon, that all IS not well, often takes longer Vlolatlons, on the other hand, are Intended, although not becausethe violator intends harm In such caseswe would talk of cnmmal actlvlty

19

Vlolauon Handbook

Version 1 2

2.1.1 Human Error and Danger There 1sa certain correlation between the types of error, the existence and effectiveness of techmques for their avoidance and the opportumtles for harm
l

Most shps are bemgn and, becausethey are often quickly detected, frequently do not lead to damage or injury Any system that IS so deslgned that a shp, such as selectlon of the wrong one of two identical-lookmg buttons on a control panel, leads to a disastrous outcome, IS unacceptable Such systems should have been identified and rectified by the Safety Management System As slips are caused by factors such as haste and dlvlded attention, good work plannmg 1salso a highly effective remedy Lapses are more dangerous becausethey may be rmssed, it 1sharder to spot that you havent done somethmg Forgettmg to replace a gasket, falmg to torque up a flange, nussmg a vital lsolatlon check are all examples of lapses The problem may be compounded by the persons thmkmg that they have done the Job completely, so an extra round of checks 1sno longer necessary Lapses are especially dangerous m mamtenance, where a problem may he unnoticed until it IS too late Many aviation accidents are caused by forgetting to replace 0-nngs Safety Management Systems can require checks and balances for safety-cntlcal actlvltles, wluch forms one type of defence agamst the lapse Mistakes are even more dangerous than lapses becausethose making a rmstake thmk they are doing the right thmg They can be so sure of themselves that evidence telling them they are wrong IS ignored A nustake, such as falling to understand a pattern of alarms and decldmg what to do on the wrong hypothesis, leads to performmg the wrong corrective actlons, it may be the first explosion that signals that a nustake was underway Safety Management Systems will find rmstakes harder to combat, but good trammg, supervlslon and support certainly help reduce the chances of makmg rmstakes Vlolatlons are most dangerous of all They often represent a quite dehberatemtentlon not to follow safety or other procedures whtch put everyone at penl Safety Management Systems are not constructed with vlolatlon m mmd and only truly Inherently-safe systems could be automatically expected to survive all sorts of vlolatlons

r 1 F

1 I

2.1.2 The Types of Violation There are five mam vlolatlon types that cause problems for orgamsatlons attemptmg to control behaviour These are listed below m the order m whch the five step guidance (Part III) recommends that they should be tackled Unintentional violations Erroneous or unmtentlonal vlolatlons occur for two mam reasons First, they arise from procedures which are wrrtten m an attempt to control behaviour that It 1slmposslble for the employee to control e g do not shp or remam m control of your vehicle at all times (an example from the Bntlsh l3ghway Code) Second, umntentlonal vlolatlons may occur when employees do not know or understand

20

Vlolatlon Handbook

Version 1 2

the rules Thus may be particularly relevant to new employees or when completmg tasks that require adherenceto a large number of rules These vlolatlons, but for the existence of a rule, would be considered errors Strrctly speakmg, the defimtlon of vlolatlon requires that devlatlon IS deliberate, I e mtentlonal However, from an orgamsatlonal perspective, It 1simportant that the unmtentlonal vlolatlon of formal procedures be avoided Routine violations. Vlolatlons of ths kmd, as their name Implies, are common practice They often occur wth such regularity that they become automatic and unconscious behaviours Such devlatlons from formal workmg practice are often perceived by employees to mvolve httle nsk and are accepted by the particular work group as the normal way of domg theJob In this case,vlolatmg the rule has become the group norm Situational violations. These vlolatlons occur as a result of factors dictated by the employees unmedtate work space or environment, which make it difficult for the employee not to commlt a vlolatlon Factors such as time pressure, lack of supervlslon, unavallablhty of equipment and msufficlent staff all have tmphcatlons m terms of sltuatlonal vlolatlons For example, when an operator improvises becausethe equipment specified m the procedure IS not available Optimising violations. Thuscategory of vlolatlons IS related to the nature of theJob or the task itself, as optlrmsmg vlolatlons ii-equently occur m an attempt by the employee to make aJob more exciting or mterestmg These vtolatlons are related to the nonfimctlonal aspects of work e g a desire to impress or to relieve boredom These vlolatlons are also assocratedwrth staff testing the safety boundaries of the system In such cases staff may actively search for ways of lmprovmg production These vlolatlons are more common when employees are mvolved m long penods of monotonous work, (such as monttonng work) or Jobs where the rules are overly restnctlve or seen as out of date Exceptional violations. As their title suggests,these vlolatlons are rare and tend to happen only m very unusual circumstances e g an emergency, or where somethmg goes wrong e g equipment failure Exceptional vlolatlons can be the result of either conscious decision makmg or mstmctlve reactlons An example of this type of vlolatlon nught mvolve an employee entenng a vessel to assist an unconclous colleague who has been overcome by fumes, despite rules that forbid such rescue attempts Before embarkmg on a nusslon to ensure comphance with all exlstmg rules and procedures, it IS necessaryfor managers to asks themselves a number of questions # Do employees know and understand the procedures7 + Do we need all of theseprocedures7 ? Are there xtuatlons when rf IS rmposslble to apply procedures7 + Does thejob itself encourage vlolatrons7 4 Is It possible to have a procedure for every srtuatlon 7 + Are there alternatives to procedures7

21

Violation Handbook

Version 1 2

We ~111 return to these questions m Part III, but it 1sworth remembenng when reading ths guide that deviance from procedures 1snot always bad - sometlmes It saves hves, and so it follows that comphance with procedures 1snot always good - sometlmes It lulls Ed Punchard (1989), a sur-vlvor of the Ptper Alpha disaster describes the behaviour of people on the platform dunng the disaster All over the ng, people were mstmctlvely followmg thetr tranung and emergency mstructlons In the absenceof any form of announcement, most were trymg to make their way to the galley to muster, have a head count, and take mstructlons After all, that was what they were tramed to do (p 128) Tragtcally the accornmodatlon was m the line of the fireball when It erupted, mearung that the majority of people who had comphed with the emergency procedures did not suave Those who did suave were those who disobeyed mstructlons andJumped 2.1.3 Why is violation so dangerous? There are several reasons why vlolatlons Increase the lrkehhood of an acctdent
l

Vlolatlons take people outslde the boundanes of safe workmg practice, makmg the environment less forglvmg to errors In other words, vlolatlons circumvent one layer of defence, the rule book, whch alms to ensure predlctable and safe workmg practrce Vlolatlons can themselves be errors when the mdlvldual does not know or understand the rule Thus lack of understanding IS dangerous m Itself, becausewhle not appreclatmg the risk, people often fall to protect themselves Vlolatlons can take people mto new or unpracticed sltuatlons, m whtch the person IS more hkely to make an error

Vlolatlons mvolve knowz~.@ydomg somethmg or, conversely, knowzngiyfazlzng to do somethmg, such as a full electncal tsolatlon Vlolatlons are worse than mistakes because they mvanably put the system out on the edge A system that, before any maintenance work starts, should be electrically Isolated and tested as hydrocarbon free, will be much more open to disaster if one or both of those steps are sbpped The real problem can be understood when we reahse that the vzolator almost always assumes that everyone else wzll do the rzght thong But this assumption 1sno guarantee that someone else ~11 not comnxt an error, such as dropping a ferrous hammer and creatmg a spark, or turning on a piece of electrical equipment when under the rmstaken ImpressIon that this 1sJust what 1srequired at the time Other peoples slips and nustakes form the second half of the equation (Free, 1994) that shows how seenungiy safe rule-bendmg can turn Into disaster Violation + Error = Death/Doom/Disaster

22

Vlolatlon Handbook

Version 1 2

i+iizigy
The Edge Fpre 3 7heoperatzng envelope Specrjc events such as A and B eachzntroduce some perturbatron In the caseof A, wlthzn the Area of Normally w Operatzons,posszble outcomes range between znherently safe and may approach the edge Event B, however, has many outcomes that go over the edge, even If there are also some that move the systenr rn to normally safe operation In thus representatzonan error may be regarded as rncreaszngthe radrus of outcomes around the event porn t Speeding m the desert 1snot a problem until the dnver has to cope wth dnvmg through the wmdrow, where a turn m the wrong dIrectIon can mean a hgh speed roll-over and. all too frequently, death (especially when the occupants are not weanng thetr seat belts either) All of a sudden a simple and apparently quite safe actlvlty, Just drlvmg a bit faster than the official speedhrrut, can turn mto a lethal factor Other evidence for the role of vlolatlons m accident causation also comes from work m the field of dnver behaviour Parker, Reason, Manstead & Stradhng (1995) surveyed 1656 dnvers and usmg the Dnver Behaviour Questronnalreexammed the relatlonshlp between drlvmg behaviour and accident mvolvement They found that self reported vlolatlons, those behaviours that mvolve dellberate deviations from safe drlvmg practice, were found to be predlctlve of accident hablhty The self reported tendency to make errors (e g misludge the speed of an oncommg velucle) or to have lapses (e g attempt to drive away from a traffic light m third gear) did not predict accident hablhty Vlolatlons were found to be a statlstlcally slgmficant, posmve predlctor of accident tnvolvement, even after the effects of exposure, age and gender had been partialled out

23

Vlolauon Handbook

Version 1 2

2.2 What Should You Do? 2.2.1 Proactive and Reactive Approaches There are two approaches taken in all the Tnpod tools The reactrve approach involves waiting until something goes wrong, and then finding out what the problems are and why they happenedThe proactrve approach requires some analysis before thmgs go wrong, followed by fixmg problems before they turn mto accidents So accident mvestlgatlon Tnpod Beta - and subsequent analysesare pnmanly reactive, but strong becausesomethmg really has gone wrong for an accident to happen Tnpod DELTA IS proactlve, nothmg has gone wrong yet, and always open to the argument that what hasnt gone wrong yet wont necessarily go wrong m the fbture For vlolatlons there are also two equivalent approaches The reactive approach mvolves lookmg to seewhat vlolatlons people are actually performing, so that they can be stopped and managed The proactive approach requires lookmg at the grounds which allow vlolatlons to take place Part III of the Vlolatlons Manual concentrates upon the proactlve approach, ldent@mg the grounds known to lead to vlolatlon wlthm Shell Group compames and proposmg specific solutions to problems when they are identified Part IV of the Vlolatlons Manual IS reactive, almed at ldentlfjrmg vlolatlons as they currently occur and proposmg ways of managing and reducing vlolatmg behaviour 2.2.2 What Can be Done to Make Things Better? There are a vanety of techmques wlthm both proactlve and reactive approaches Because the types of vlolatton and the reasonsfor vlolatlon are so many, It 1snecessaryto first find out what and where the problem rmght be, and then to select the remedial approachesthat are most likely to be effective There are many perfectly effective steps that can be taken What should be avoided, however, IStakmg unnecessaryaction for non-problems, ones that have already been solved The most specific lesson that has been learned IS that m Shell Operatmg Umts m the North Sea environment, the problem of IntentIon 1snot really a problem people have good mtentlons The steps needed to solve intention problems are, therefore, only necessaryto ensure that mtentlon does not become a new problem If condltlons ever change Ic however, mtentlon IS not a problem, IS there anythmg that 1~7 Because people are still vlolatmg, the answer to ths question 1salmost certainly yes The reasons for vlolatlon are bound up m the ways we do the work, the people we hke to hre and the ways we expect them to go about their business 2.2.3 The Behavioural Cause Model The Behavioural Cause Model IS a way of lookmg at why people behave the way they do It reflects the components leading to vlolatmg behaviour that have been collected together as a result of a number of studies on offshore operators, supervisors and managers rn Shells UK and Netherlands North Sea operations

:
b

F-

24

Vlolauon Handbook

Vernon 1 2

r Igure 4 The BehavIoural Cause Model The sltuatron IS added to the model and the lndrvlduals Internalfactors are brought together with the srtuatlonal variables m the plannmng component prior to behaviour Powerfulness and powerlessness,feehngs of control or of loss of personal control over one s actlons, can both shot t-cncult planrung, as can the effect of habits ,whxh generate a human error route The model attempts to Integrate a number of different psychologlcal processes All of them are considered to lead to an mdlvtduals behavmg m a particular way, but they are generally separatedfor the purposes of study For a practical theory, however, It ISnecessaryto Integrate the pnnclpal causal factors m an attempt to maxlmlse the proportlon of behaviour that can be explained and, therefore, predlcted In the figure an arrow IS Intended to show the causal chain So Intentions are constructed using Attitudes, Social Norms and Feehngs of Control Consequences,the negative outcomes possible if thmgs go wrong, are seen as formmg an mput to Attitudes, and thence to Intentlon and Expectation, rather than formmg a direct input Plans are constructed usmg the external goals (the work that needs to be done etc ), a persons IntentIons and Expectations plus the opportunltles that are recogmsed at the time Actual behaviour, m this sensedefined as vlolatmg behaviour, IS seen as determmed by plans (see the section on Human Error on Plans and Intentions), but can be drastrcally altered by either feelings of loss of control or the sudden appearanceof unexpected opportumtles

25

ViolationHandbook

Version1 2

Motlvatlon IS seen as a general driver, set at least partially by externalgoals,andmfluenctng basic inputs such as Attitudes and Norms Motivation almost certainly also influences the

ablhtyto recogruse oppotturutles whenthey anse Motlvatlon ISa two-edgedsword The

unmotivated do httle, neither active nor m vlolatlon The highly motivated can see opportumtles and suddenly decide to vlolate, for the best of reasons Below are hsted m more detail the actual factors mvestlgated m the BehavIoural Cause Model
Behavlouxal Athtude Element Defuution

The mcbvlduals posihve or negahve

evaluation of performmg the parhcular

behaviour of interest Habits Sod Noms a) Comphance Soual Noms b) Conformity

Sltuatlon-specific sequences that have become automahc, routnuzed to the extent that there 1s no consaous deaslon to act A consaous desve to follow the rules, even
when the me&ate necessity or sense 1s not always obvious The desve to be the same as other members of the group

Soual Norms c) PersonalNorm

Feelmg of Control Feehng of Control a) Powerfulness b)

lntemahsed moral norms & anhapated regret reflechng expected emohonal

response If tlungs go wrong Powerlessness

The feelmg of supenonty, competence and sk~.U based on experience The feelmg of not bemg able to resist the temptahon to perform the behaviour, the feelmg of mabtity to alter events The degree to wluch a person has formulated consaous plans to perform or not perform some speclhed behaviour The m&vlduals eshmahon of the hkelihood that they will perform some speahed behaviour The quabty and efhuency of the plannmg process that precedes work The access/posslbtihes an m&vldual has to do the work III another (better, more efhaent) way and their sublechve eshmate of the consequences of workmg III that way Tune, rewards and the speclhc Job requvements Non-psychologcal factors wluch dnve much of the planmng process The possible negative outcomes and thev probab&tles that can be assoaated with fdures of planned achons TIus separates out a knowledge component from attitude

Intention Expectahon

External Goals Consequences

Mohvatlon
Behaviour

The mtemal drive to act m a gWenway Vlolatlon behaviour

26

Vlolauon Handbook

Version 1 2

2.2.4 Testing the Model The BehavIoural Cause Model was tested on 182 operators and supervisors offshore m the North Sea (Verschuur et al , 1996 ) This study had two goals The first was to test the Behavioural Cause Model, It was found that the model allowed successtil prediction of 64% of the variance m vlolatmg behaviour with Just four maJor factors that provided the dommant dnve to violate, Expectation, Opportunity, Powerfulness and Preparation Planning Knowledge of an mdlvlduals scores on these factors allows the prediction of whether they are likely to vlolate or not The prediction 1s2/3 of the variance m vlolatmg behaviour Predicting or explammg the variance IS a measure of accuracy, and also mdlcates how much other factors not considered could also be effective A proportion of variance explained m excess of 60% means that any other factors will have a mmonty effect To put this m context most such predictions of behaviour are about 20-30% accurate and even massive oplmon polls before electlons can only reach such accuracy wth samples m excess of 1000 and with very simple voting behaviour This study allowed an empmcal assessment of the importance of the different factors and the extent to which changesm contributory factors rmght be expected to produce changes m the probability of vlolatmg 2.2.5 An Alternative Model - Supervision and Punishment A number of alternative models, explammg why and how people vlolate, can also be tested The most obvious IS one which assumesthat people are basically bad and lazy, vlolatmg unless they are forced to do the nght thing The most effective methods of controlhng vlolatlons m ths SupervzseCC Punzsh model are detection of vlolatlon by supervlsors and strong purushment by both supervisors and managers whenever vtolatlons are detected Violators are seen as poorly motivated, especmlly with respect to safety Wfule the Behavioural Cause Model explamed m excess of 60% of vlolatmg behaviour, ths Supervise & Purush model could also be tested and was found to explain only 20% When the two are combined, by adding the BCM factors and the Supervise and Pumsh factors into the equation, the total variance explained only rose from 62% to 67% Thussmall increase of only 3% IS partly due to the fact that many of the effective elements m the Supervise and Pumsh management approach have already been taken up by factors such as planmng and expectation The message1sclear management of violation requires understandmg of the slgmficant drlvmg factors rather than bemg basednutlal preconceptions that are untested m practice Concentrating upon detectlon and supervisIon will only produce marginal improvements, whle concentratmg upon the stgmficant factors m the BehavIoural Cause Model ~111 produce major improvements The efficacy of different remedial measures can be assessed, with associatedtolerances, and the values allow prlontlsatlon on the basis of data rather than impression and prejudice

27

VlolauonHandbook 2.3 The Manual - Structure

Version1 2

The manual IS deslgned to help solve the dlagnosls problem and present suggestions for remedlatlon In Part III each of the slgmficant factors of the BehavIoural Cause Model IS described m detail The questlons used to discover whether there IS a problem are set out m the defirutlon section If the likely answers to these questlons, m a given operational urut, are felt to reflect a sign&ant problem then the relevant section on the underlymg components gives more detail The reader can choose to move on to the component sections before selectmg the appropnate remedy sectlons As the remedies are mostly defined m terms of the components this 1sexpected to be the usual approach Each section m the dlagnosls part of Part III consists of the followmg elements
l l l l

What isX? - e g Intention, Planmng, Powe&lness How do you define it ? - The questions that you can ask to detect problems The Components - Pomtmg to more detailed underlymg causes Pointer to Remedy sections - What to do about the problem

2.3.1 The Remedies There are a wide range of remedies avallable No smgle solution IS sufflclent This manual IS Intended to help you select the appropnate remedies In this section a general overview IS given of the possible approachesto remedlatlon Each one IS evaluated with Its advantages and disadvantages and a simple ratmg IS gven of each approach As such evaluations are crude It IS only a three-pomt scale The mtentron ISto force people at least to thmk about the rated charactenstlc and not to Imagine that every approach ~11 be equally effective, easy to use or will take effect lmmedlately Remedies are rated m terms of Good
l l l l

Ease of use Easy Counter-productrwty No side-effects Effectweness Low effectiveness Time scale to success Immediate impact

Medium Some Side-effects Effective Months

Difficult Major side-effects I+gh effectiveness More than a year

The mdlcatlons on these scales are given tn bold typeface

28

Vlolatlon Handbook

Version1 2

There are a wide vanety of possible ways to remedy vlolatlon or potential vrolatmg behaviour Some of these are really effective, some only appear to be so Some may be effective but have such bad side-effects that they are best not pursued or only chosen and implemented wth considerable care The dlstmctlon between recommended, effective remedies and the less effective 1sgiven m terms of Dlvlslon I and Dlvlslon II remedies The second dlvlslon remedies are only introduced and dlscussed here m order to show that the full range of posslblhtles has been consldered and to prevent people from proposing what looks attracttve but doesnt work The possible approaches to remedlatlon are Division I 8 Supenwion - watch more closely over people, plan their work, provide an example Analysis of misting violations - fmd the background behmd vtolatlon behaviour Analysis of violation potential - use the theory to see If people ~11 find vlolatmg easy Structured discussions - discover what 1sgoing on and get people to agree on actlons Procedures - Improve the qualrty of procedures so vlolatlon 1snot seen as necessary Direct management - a techruque to vary the procedures as circumstances demand
l l l l l

Division II
l

- pick people who dont violate to start with Training - tram people more on the real nsks and the possible consequences Detection - sharpen the lookout for vlolatlons Reporting - encourage people to report when they or others bend the rules Incentives - pay for compliance or pumsh people who break the rules
Selection

29

Vlolatlon Handbook PART III 3.0 The Proactive Approach - The Diagnosis-Remedy Handbook

Version 1 2

Thus part of the Vlolatlon Manual 1s mtended to allow the reader to dlscover whether there are grounds for vlolatmg behavlour wlthm the workforce Each section can be read Independently, although reading one may lead to another For each of the important factors uncovered and used m the Behavloural Cause Model there IS a dlagnosls section, of one or two pages If the specific factor 1s recogmsed as a problem, or a potential problem, the reader can turn to the appropnate remedy sectlon Do not expect that everythmg WIII be a problem If you read a medlcal dlctlonary and feel you have every disease you read about you are either already dead or are suffenng from hypochondria The same applies here If you find a problem with every sectron then you either have a very senous problem Indeed (wlzlch IS most unlikely) or you are reading too much For instance, Intention 1s unlikely to be a serious problem wtile expectation IS often a major problem To calibrate yourself go out and ask people the questions m the sectlons (the scales are given next to the questlons) The basic scales used for answermg questlons or agreemg mth statements are Likely Very unlikely Desirable Very desirable AgreementTotally agree Frequency Never Approval Strongly approve - unlikely - desirable - neutral - likely - very hkely - very undesirable - totally disagree - almost always - strongly disapprove I4 r l-

- n e u t r a l - undesirable

- agree

- neutral

- disagree

- hardly ever - approve

- sometimes - not approve/ not disapprove

-often - disapprove

30

VIolaDon Handbook

Version 1 2

3.1 Diagnosis Section 3.1.1 violating Behaviour What is Violating Behaviour? Behavlour 1s what you do In the context of vlolatlon Behavlour IS set more narrowly as do people bend or break the rules or procedures Behavlour 1s seen as the performance of a plan of action If the plan 1s a good one we expect the behavlour to be good as well Errors arise when local factors, such as haste, overload or dlstractlon allow other actions to be performed (slips) or essential actions to be slupped (lapses) Poor plans are rmstakes, whch result In intended but faulty behavlour Two factors can be seen to shortclrcult the plannmg process and hijack behavlour l The sudden dIscovery of highly attractive opportumtles l The reahsatlon (nustaken or not) that a specific course of action has to be followed, there IS a feeling of having no control over ones own behavlour How do you define violating behaviour? The followmg statements served to define peoples behavlour How often, m the last [Frequency]
SIX

months, have you found yourself not followmg the rules7

Posltwe answers to either of these Sometimes 1 deviate from the rules I sometimes bend the rules What are the components3

(Agreement] [Agreement]

The Behavloural Cause Model defines the effective components leadmg to vlolatmg behavlour The field studies showed that four components are the most important l Expectation l Powefilness l Planning l Opportumty Because the actual mcldence of vlolatlon may be quite low, certainly what IS reported, It IS Important to remember the the maJorlty of the workforce will probably be Wolves, so It 1s still worth followmg back to these four components before stopping Remember, If you thmk that there isnt a problem, you will be all the more surprised when somethmg does happen1 Do you have a problem? If you think so after reading this section go to the remedy_ section.

31

Vlolauon Handbook

Version 1 2

3.1.2 Intention What is Intention? Intenttons are what you have before you actually do somethmg Intentrons define what you want to do If you behave m a certam way we assume that there 1s some mtentron behmd the behavtour In general we do what we intend to do, bamng error, rmshap or ctrcumstances If we behave wzthout an obvrous and explamable mtentton, we feel that that 1s a sign of somethmg wrong More generally we do thmgs we know we should not do because we have the mtentton to do them, so the mtention IS bad When people do not have good mtentions, there IS little that can be done except to either 1) fix then mtentrons, or 2) ensure that they have no choice m how they behave, regardless of their intentions How do you define intention? Intention can be gauged by the followmg Items
l l

I intend to follow the procedures It 1s obvious I wrll follow the procedures

[Agreement] [Agreement]

There IS an mtentron problem rf either one or both of these IS likely to be answered at all negatrvely The first 1s obvrous and failure here is crrttcal (or even crnnmal) The second is more subtle and tells us about the local culture Is It obvrous that people will, m general, intend to follow procedures or is even this assumptron unwarranted? If there IS an ldentrfiable mtentron problem, read on and assess which one or more of the component factors discussed below nnght be the root cause of the problem The field studies m the North Sea suggested that Intentron, on Its own, IS probably not a problem If rt ts, thrs IS where you have to start If It IS not, then exhortation and more trammg are unhkely to make any difference and would be a waste of resources better employed m solvmg the problems that still cause people wrth good mtentrons to vrolate

32

Vlolauon Handbook

Version 1 2

What are the components of Intention? The theory behmd the Behavloural Cause Model assumes three major components to Intention 1 Attitudes Attitudes to work determme what 1s felt to be more or less Important There IS a certain hnk to Personal Norm m that what IS felt to be less important IS more hkely to be dlsregarded If attttudes to work are poor thts can appear as (1) Incorrect assignments of what IS and IS not Important and (3) Incorrect beliefs about what can happen 2 Norms Personal and GrouD Norms Group norms (my mates/colleagues) may conflict with desirable behavlour Personal norms may not be strong enough to overcome temptation 3 Feeling of Control Feelings of control over the work and the sltuatlon contrast mth loss of control Is there a can-do culture or are there sltuatlons where people feel they have no choice? These all contribute to the formulation of Intentions to act or, m ths case, to violate and bend rules

Feeling of Control I/

Do you have a problem? If you think so after reading this section go to the remedy section.

33

Violation Handbook

Version 1 2

3.1.3 Expectation What is Expectation? When you expect to find yourself m a posItIon that ~111 require bendmg or breakmg rules, tlus makes It much more likely that. when that sltuatlon arises, you ~111 not hesitate to do Just that Expectation turns out to be the biggest single factor leading to rule-breakmg Even If a persons intentions are good, and there 1s no great pressure, expectations dnve behavlour Such negative expectations are built over a penod of time Previous experiences, when vlolattons took place, are very strong m determmmg what you will do the next time unless somethmg has clearly changed m the mean time People will usually explam why they behaved the way they did by referrmg to the circumstances, such as time pressure, lack of alternatives, or poor procedures These attrrbutzons are not always acceptable to someone outslde, such as their boss, but they are certamly what drives the person themself How do you deJine expectation? The followmg statements serve to define peoples expectations Posltlve agreement to any of these
l l

l l

I may not always follow the rules In future (Agreement] In the near fLture I do not expect that I ~111 find myself always workmg accordmg to the rules [Agreement] I am sure that thmgs will still go wrong m the future [Agreement] Wle I intend to follow the procedures in future, I expect that there will be times when I will not [Agreement]

and ratmg as likely L.

l

How likely do you thmk it IS that a sltuatlon WIII arlse In the f&ure m which you will choose not to follow a procedure7 [Llkehhood] All thmgs considered, how likely 1s it that you will actually always follow the procedures? (Llkehhood]

r .

suggest that there 1s an expectation problem

Vlolabon Handbook

Version 1 2

What are the components of Expectation? Indirect Attitudes

Personal Norm

Powerlessness

Past Behavlour Past behavlour, generated m sltuatlons which led to vlolatlon m the first place, creates a VICIOUS circle Unless this cycle IS broken the best of mtentlons will never get off the ground Attitudes Attitudes to work determme what 1s felt to be more or less Important There 1s a certam hnk to Personal Norm m that what 1s felt to be less Important IS more hkely to be disregarded If attitudes to work are poor thus can appear as (1) incorrect assignments of what 1s and IS not Important and (2) Incorrect beliefs about what can happen Powerlessness Powerlessness appears when sltuatlons arise m which there appears to be no choice TIE 1s charactenstlc of sltuatlonal vlolatlons Are there such sltuatlons which are regularly confronted? Indications of this are frequent complamts from different people about the same work, often when different contractors are required to do thelob This can be hldden when routine vlolatlon has become the standard way of solvmg the problem Comphance Whether or not one wishes to Ident@, comphance implies that one 1s wllhng to go along Weanng ID passes, even when one IS well known, or filling tn reports may seem unnecessary, but can prove vital Is there an atmosphere of non-compliance or even defiance7 Note that ths IS dlstmgulshed from conforrmtv Personal Norm Personal norms form a primary defence against poor expectations Are there dlstmct boundanes over whxh people will not go? If this IS not the case, then this IS a serious problem

Do you have a problem? If you think so after reading this section go to the remedy section.

35

Vlolatlon Handbook 3.1.4 Planning What is Planning?

Version 1 2

Plannmg 1s a well known phenomenon, everyone does It The problem IS that plannmg IS most notlceable when it goes wrong Planmng mvolves settmg goals and definmg the processes needed to attam those goals, asslgnmg resources to perform the tasks, settmg measures of success and, finally, placmg the whole m a time-based framework Much plannmg assumes that the details will be filled m by those present at the time, ths 1s where thtngs go wrong Plannmg failures can produce vlolatlons m a number of ways, whch 1s why we dlstmgutsh three different sorts of plannmg process *Preparation planning - the lmtlal overall defimtlon of the work to be done *Job Planning - the local details of how the Job 1s to be done, by whom etc aAction Plannmg - when the unexpected arises and plans may have to be changed The fadure of plannmg 1s to be found m the sudden appearance of an unexpected (1 e unplanned) sltuatlon which needs to be fixed Vlolatlons appear when people respond to that failure If they decide to deviate from known and acceptable procedures to fix the problem, then ths becomes a vlolatton, especially If It goes wrong If thmgs go well it may be regarded as the appropnate exercise of mltlatlve and rewarded accordingly The tools may not be there, the extra person fals to appear, having not been recogntsed as necessary and therefore asslgned elsewhere, the time alloted for theJob may not be long enough, the appearance of a problem may create a totally new sltuatlon A vlolatlon occurs when a person deades, on the spot, to do It hs own way With all the time m the world one can possibly go back and start agam, but somettmes a situation has started and can not be left unfimshed The toothpaste wont go back m the tube If there IS one thmg people feel they are tired for It 1s to exercise their mmatlve and specialist knowledge and experience A major solution to vlolatlon IS the exercise of planrung, off the Job, and the recogmtlon of plannmg failures dunng a Job coupled with a vlolatlon management approach How do you measure the quality of the planning? Use the followmg questlons Preparation plannlnz Before we start aJob The Job IS always thought through m detail We are always aware of all the problem which rmght anse Jobs are planned together with supervisors Work IS planned m an ad hoc manner bob by Job)

r 1

r-

[Agreement] [Agreement] [Agreement] [Agreement]

36

Vlolauon Handbook

Version 1 2

When work IS carried out The tools and equipment are always available Job plannq

(Apeement]

When work IS carried out [Agreement] Thmgs rarely go wrong (Agreement] We are left to fimsh the work we have started [Agreement] Plans are often changed before work IS stat-ted [Ag eement] Work 1s often rescheduled at short notice [Agreement] The feedback on the plannmgls always sufficient [Agreement] It 1s not often clear when the standards, set by the plannmg, are met [Agreement] Planmng and scheduling often provide too little time [Agreement] The commumcatlon of plans IS often lackmg Actlon plannmng In my Job, situations often arise m which I have to adjust my plan of work [Agreement] Doing this Job means that you frequently have to change your work plan [Agreement] I have to change my planned work actlvltles almost every day [Agreement] We often find out that the plannmg 1s nmot feasible/runs out of time [Agreement] I sometimes come to reahse that theJob cannot be done as planned [Agreement] what are the components?
l l l

Preparation Plannmg, - The nutlal defimtlon of the work to be done Job Plannmg - the local details of how the Job 1s to be done, by whom etc Action Planmng - The situation when the unexpected arises and plans have to be changed

Do you have a problem? If you think so after reading this section go to the remedy section.

37

Vlolahon Handbook

Version 1 2

3.1.5 Attitudes What are Attitudes? Attitudes are the way people thmk (beliefs) and feel (values) about thmgs hke the work they perform What they thmk of as important, what they feel 1s not necessary etc If you feel that somethmg IS particularly Important, then you are more likely to try harder If you beheve that fallmg to attend can have serious consequences, and also beheve these consequences are quite hkelj to happen, then you are more likely to attend If, on the other hand, somethmg IS seen as tnvlal, unlikely to be dangerous or go wrong, then less attention ~11 be pald When It comes to setting pnontles, attitudes determme what comes first and what IS left to last Attitudes consist of beliefs and values Belzefs What someone knows to be true This 1s determmed by what they have been taught, mmally, but more importantly by what they have personally experienced over the years Twenty years worth of experience 1s a strong personal counter to someone wth a hgher Job group simply saying that your experience IS wrong Consrderatlon of negative consequences has to be welghed m with the expected probablhtles to arrive at personal Judgements of nsk Poor attitudes m the workforce may be confUsed with different behefs m the set of possible outcomes Vahes What IS seen and more or less Important or desirable Values are attached to outcomes, actlons and behavlours

Attitudes can be either positive or negative The dont grve a damn attitude contrasts with attitudes whch express ttis IS really important When someone decides to slup part of a procedure, their attitudes form a major factor m decldmg what gets slupped If we know what they consider important, and what they believe about outcomes, we can predict pretty accurately whch steps m a procedure they will always adhere to, and which parts they ~11 shP Knowledge of the potential negative consequences provides one of the inputs to the formatIon of attitudes, but these can be filtered by many factors, such as the feelmg of powefilness, which reduce a persons assessment of the (known) probablhtles hsk IS seen as probability x consequence, once someone has decided that the bad consequences are not what ~111 happen to them personally, their attrtudes to their own and others behavlour may be altered Attitudes can be altered by outside mfluence If figures whom someone respects appears to take some thmgs seriously, the effects of norms may affect the construction of attitudes

T P

38

Vlolahon Handbook

Version 1 2.

How do you define attitudes? Agreement with the followmg Items People cut corners In ON orgamsatlon because They can get the Job done faster They can get the Job done more easily They can meet the performance targets The Job becomes more mterestmgiless bonng Rating as likely or very likely the foliowmg If I were to cut corners then It would save time It would make theJob easier It would allow me to meet performance targets It would make the Job more interesting and rating as desirable Savmg time is Makmg theJob easier IS meeting performance targets is makmg the Job more interesting 1s What are the components?
l

[Agreement] [Agreemeni] [Areement] [Agreement]

[Llkelrhood] [L&elrhoodJ [Llkelrhood] [Llkehhood]

fDesrrab&yJ [Desn-abrlrtyJ [DeslrabrlrtyJ [Deslrab&yJ

Beliefs m the importance to the orgamsatlon of getting work done quickly so that the next Job can get started The belief that rapid work, tigh production rates, IS work of high value and that speed itself 1s important Beliefs rn the rmportance to the mdlvldual of having an easy life, of not doing more than 1s necessary (note the attitudinal component of what 1s to be defined as necessary) Beliefs m the importance to the mdlvldual and the orgarusatlon of meeting preset performance targets T~.E may be seen to condone occasional corner cutting as the penalties of not meeting targets (hgh probablhty unless comers are cut) are felt to outweigh the nsks of something going wrong (low probabdlty given the mdlvlduals own competence - see powerfulness) Behevmg that vanatlon in workmg methods enables the worker to keep doing the Job and reducing boredom TIE factor 1s probably the least important of the four components

Do you have a problem? If you think so after reading this section go to the remedy section.

39

Vlolauon Handbook

Version 1 2

3.1.6 Norms what are Norms? There are two types of norm l Personal Norms define the lines past whch an mdlvldual will not go, come what may l Group Norms are the values expressed by the group as a whole Someone who wants to be seen and accepted as a member of a group will not wish to be seen as followmg a different set of norms from the other group members This can lead to behavlour which someone would regret if there were not a broader posltlve payoff m terms of acceptance by the group Bendmg or breakmg a rule, wth little chance of being caught or of thmgs going wrong, can be traded off against being allowed mto the group Group norms arent necessarily bad If the values are high people are as likely to raise their standards as they are to lower their standards when the group values are less acceptable Personal norms define what I am and how I see myself, group norms define what I would hke to be An Important factor 1s the feehng of antlapated regret d somethmg should go wrong How do you define norms? 1 Personal Norms Agreement with the followmg statements mdlcates hrgh Personal Norms
l l

l l l

In my opnuon a professional always follows the procedures [Agreement] If damage should result from my failure to follow the rules, I would feel bad about It afterwards [Agreement] Id find It hard to Just@ bending the rules to my colleagues [Agreement] Personally I strongly disapprove when rules are bent [Agreement] If I bent the rules I would feel bad about It afterwards because it 1s unprofessional

[Agreement]
l

I would find it hard to face my colleagues afterwards, If I had not followed the rules

[Agreement]
l

If 1 were to cut corners I would feel bad about it afterwards, because It 1s n-responsrble It 1s n-responsible to your colleagues not to follow -he rules I disapprove when someone IS not followmg the rules One should always follow the rules

l l l

[Agreement] [Agreement] [Agreement] [Agreemenf]

40

Vlolatlon Handbook 2 Group Norms

Version 1 2

Generally the group norms at different levels are measured by approval or disapproval of comer cutting using approval of the followmg statement When people cut comers m our orgarzlsatlon they expect l Their colleagues/peers to [APProveI l Their supervlsors to rAPProve l Their managers to rAPProve Compliance by mdlvlduals
l

IS

mdlcated by agreement with the followmg statements

If I were to cut comers then [Agreement] l My supervrsor would disapprove [Agreement] l The company would disapprove [AFeement] l My work mates would disapprove [Agreement] l Others (techmclans) would disapprove Exactly whch groups norms are being complied wrth depends upon which group 1s felt would disapprove Agreement mth the followmg general statements mdlcates mdlvldual compliance to group norms
l l l l

l l l

l l

l l

l l

l l

[Agreement] I follow the rules, whether I agree with them or not [Agreement] I only bend rules if others allow me to [Agreement] You had better follow the procedures, Its m your own interest When somethmg goes wrong others wrll cover themselves by using the rules to blame me [Agreement] You better stick to the rules because others will cover themselves tf you dont [Agreement] [Agreement] If you dont follow the rules you should expect to be putushed People who disobey the rules are Just bemg awkward and makmg hfe hard for themselves [Agreement] [Agreement] I would not question the rules If they were wntten by experts I am m favour of very strict enforcement of all laws no matter what the consequences [Agreement] An experienced worker 1s Justified m not domg the Job exactly to the book [Agreement] A really slulled worker IS one who knows which rules to bend so as to get theJob done more [Agreement] efficiently People who rnmdlessly follow the rules are Just covermg themselves [Agreement] I would not obey rules enforced by a person who does not deserve their posltlon of power [Ageement] (Agreement] I dont obey rules unless I agree with them If a manager asked me to do something whch I did not agree with, I would not do It [Agreement]

41

Vlolauon Handbook

Version 1 2

When my colleagues thmk one should follow the procedures, I usually resign myself to It

[Agreement]
l l l l l

1 deal mth rules no differently from the way my colleagues do [Agreement] [Agreement] I usually do the same as my colleagues as It comes to rules Its obvious you handle procedures the way your colleagues expect you to [Agreement] [Agreement] When It comes to rules I agree with my colleagues [Agreement] I agree with colleagues who thmk there are too many procedures

What are the components?

l l l l

Anticipated Regret Unthmkmg comphance Pragmatic rule-bending Perceived legltlmacy see powefilness

Do you have a problem? If you think so after reading this section go to the remedy section.

42

Vlolauon Handbook

Version 1 2

3.1.7 Feeling of Control Wat is Feeling of Control? When you are m control you can choose what actions you perform, when you perform them and how you do It When one or more of these IS constrained, you may feel that you have lost control over your own actions We have defined feehngs of control m terms of whether you feel that you have power over your actions, or that events have power over you, so that you have no choice Powerlessness IS a very senous problem because it can short-circuit the best of mtentlons If you feel that you really have no choice, It takes a very strong counter-force to overcome the dnve to do what you know you shouldnt Powerfulness IS ofienJustlfied, but allows people to overlook what they know to be Important problems on the grounds that that wont happen to me Both powefilness and powerlessness are mtenslfied by strong motlvatlon While someone who IS little motivated to achieve an end will not be too affected by ablhty or constraint, a highly motivated person will be very frustrated when they experience powerlessness, and will be even more likely to see their own ablhtles as a Justlficatlon for doing what they are about to do How do you define it? See Powerlessness See Powerfblness What are the components?

Powerfulness Powerlessness Do you have a problem? If you think so after reading this section go to the remedv section.

43

Vlolatlon Handbook

Version 1 2

3.1.8 Opportunity What is Opportunity? In a hghly planned and regulated environment there are few surprises Sometimes, however, It IS possible to see a better way This may be faster, cheaper or less effortfbl Opportumtles exist at a personal and at a Job level Opportumty appears to be a major factor m determmmg whether people will violate or not Are there many opporturutles or few? Do people get rewarded and looked up to If they recogmse and seize opportumtles that anse? Are the stones m the bar about what people saw and did, despite bemg supposed to do somethmg else, or are they about how careiil people were when lgnonng the opportumtles that arose? This posltlve aspect of opportumty should be dlstmgulshed Corn the luck of opportumty that 1s better covered by powerlessness under loss of control Here the best and proper way 1s blocked - a lack of opportumty People expect to find themselves cutting comers because The procedures take too long to follow There are too many procedures The procedures are too comphcated and The procedures change too often The procedures need to be updated Many sltuatlons arent covered by procedcures How do you define opportunity? Sl tuational Opbortumty The exrstence of sgnlficant Srtuatronal Opportunrty IS rndrcated by agreement wrth the following statements l We could really get the Job done a lot faster /Agreement] l There are opportumtles to take shortcuts which would slgmficantly speed up the work [Agreement] l There are opportumtles to take shortcuts which would make hfe easier [Agreement] l There are sltuatlons m whch there IS a better way of domg theJob than the one m the procedure [Agreement] l There are srtuatlons m which It 1s possible to do the Job safely m ways other than that defined m the procedure [Agreement] l In ths Job ctrcumstances anse In which there are several ways to do the Job [Agreement]

44

Violation Handbook

Version 1 2

Personal 0pportunrQ
l l l l l

SometImes there are sltuatlons m which I could get the Job done a lot faster [AgreementJ [Agreement] Im able to take shortcuts to speed up the work [Agreement] If I take shortcuts It would make hfe easier In some sltuatlons I can do theJob better than the one In the procedure [Agreement] In some sltuatlons I can do the work safely m ways other that defined m the procedure [Agreement]

FKhat are the Components? The major factor IS the workplace If there are alternatives available then people may know of them and, even, use them However, motivation plays an important role Unmotivated people are likely only to see opportunmes for their own ease, the highly motivated, on the other hand, may see room for Improvements and grab them before It 1s too late Do you have a problem? If you think so after reading this section go to the remedy section.

45

Violation Handbook 3.1.9 Powerfulness What is Powerfulness?

Version 1 2

Powerfulness 1s the feelmg that you can control events and outcomes You know what you are domg, you have been tramed for the task and are tughly experienced This may not always be true, m whch case the feeling of control 1s illusory Illusory control 1s what amateurs have, or young drivers when they have Just passed their drlvmg tests In the oil and gas industry powerf3ness reflects real ablhty, although feelmg powetil may suddenly become a problem when the situation 1s not exactly as usual PowerWness reflects proven competence, but no one mdlvlduals experience 1s ever gomg to be so wide that they will have personally tested themselves agamst everything that could be requu-ed This means that the downside of competence 1s over-confidence m ones own abilities - the feeling of poweri%lness How do you define Powerfulness? Agreement wrth the followvg statements It 1s mamly up to me whether or not I cut comers
l l l l l l l l

l l l l

l l

It IS OK for somebody with my experience to cut comers [Agreement] When you are slulled you know how to deviate from the rules safely [Agreement] My experience tells me when bending the rules ~111 not be harti to me [Agreement] You can bend the rules if you know what you are doing [Agreement] I dont follow rules if I know how to manage without them [Agreement] My experience enables me to devlate from the procedures safely [Agreement] If you are slulled you know when it IS OK to break or bend the rules [Agreement] My professlonal knowledge tells me whether certain rules are appropriate or not [Agreement] People wth a lot of experience are expected to cover shortcommgs m procedures [Agreement] I dont need rules to tell me what to do /Agreement] I dont follow rules that are time-consummg fAgreementJ I wouldnt run mto problems even if I didnt follow the rules /Ag cement] Sometimes experience means that you know a better way of domg the job than the one defined m the rules /AFeementJ I know when there IS a better way to do theJob /Agreement] There are situations m which I use my own Judgement, whether that tnvolves followmg a procedure or not fAgreementJ I know from experience when I should follow the procedures or not /Agreement]

46

Vlolauon Handbook

Version 1 2

What are the Components7 l Trainmg l Experience l Attitudes l Feehng of control

Do you have a problem? If you think so after reading this section go to the remedy section.

47

Violation Handbook

Version 1 2

3.1.10 Powerlessness What is Powerlessness? Powerlessness IS the feehng you have when you no longer have control over events and people Thus feeling, m combmatlon wth a high degree of motlvatlon, can be extremely fi-ustratmg A feeling of loss of control, powerlessness, can result m rule-breakmg despzte the best mtentlons and reasonable expectations Powerlessness 1s recogmseable when you watch someone else doing aJob you know you can do as well (or better - e g powerfulness) and you know they arent domg It well, but you cant do anything to do it better How do you define Powerlessness? Agreement with some or all of the followmg
l l l

Some procedures are so time-consummg they cant be followed anyhow [Agreement] [Agreement] Sltuatrons sometlmes anse which force you to break the rules In my work environment sltuatlons can anse m whch procedures are very hard to follow

[Agreement]
l

In my work environment I am often confronted with circumstances m whch I feel compelled to deviate from the procedure [Agreement]

U/hat are the Components?

l l l

Trammg Attitudes Motlvatlon

Do you have a problem? If you think so after reading this section go to the remedy section.
T

48

Vviatlon Handbook

Version 1 2

3.2 The Remedies There are a wide range of remedies available No one solution 1s sufficient Tl~s manual 1s intended to help you select the appropnate remedies In this section a general overview 1s given of the possible approaches to remedlatlon Each one IS evaluated wth Its advantages and disadvantages and a sample ratmg 1s given of each approach As such evaluations are crude rt IS only a three-point scale The mtentlon IS at least to force people to thmk about the rated charactenstlc and not to lmagme that every approach ~111 be equally effective, easy to use or will take effect nnrnedlately Remedies are rated m terms of Good Ease of use Easy Counter-productwq No side-effects Effectiveness Low effectiveness Time scale to success Immediate impact Bad D&lt Major side-effects figh effectiveness More than a year

l l l l

Medium Some Side-effects Effective Months

The mdlcatlons on these scales are given m bold typeface There are a wide variety of possible ways to remedy vlolatlon or potential vlolatmg behavlour Some of these are really effective, some only appear to be so Some may be effective but have such bad wde-effects that they are best not pursued or only chosen and implemented with condlferable care The dlstmctlon between recommended, effective remedies and the less effective 1s given m terms of Dlvlslon I and Dlvlslon II remedies The second dlvlslon remedies are only Introduced and dtscussed here m order to show that the fi~ll range of posslblhtles has been considered and to prevent people from proposrng what looks attractive but doesnt work The possible approaches to remedlatlon are Division I Supervision - watch more closely over people, plan then- work, provide an example Analysis of existing violations - find the background behind vlolatlon behavlour Analysis of violation potential - use the theory to see If people will find vlolatmg easy Structured discussions - discover what IS going on and get people to agree on actions Procedures - improve the quality of procedures so olatlon IS not seen as necessary Direct management - a techmque to vary the procedures as cn-cumstances demand Division II Selection - pick people who dont vlolate to start with Tratning - tram people more on the real risks and the possible consequences Detection - sharpen the lookout for vlolatrons Reporfing - encourage people to report when they or others bend the rules Incentives - pay for compliance or pumsh people who break the rules

49

Violation Handbook

Version 1 2

3.2.1 Division I Remedies 3.2.1.1 1. Supervision

SupervIsIon can both set an example and serve to correct bad hablts before they get out of hand Supervisors can actively search for problems that lead to vlolatlon, such as poor procedures and inadequate planrung, they can iden@ who are sheep and who are wolves and allocate them appropnately They are the people who should perform or check the plannmg process and have the experience to recogruse opportuxutres earlier Supervisors should not. however, merely act as policemen. watclung others work m the hope of detecting vlolatlons Advantages SupervIsors are close to the action If there IS any rule-breakmg going on, they should be aware of It They are also m a position to do somethmg about the work plannmg and should have the experience to see opportumtles early and, therefore plan for them Disadvantages Supervisors may be the biggest violators, they know better, have seen It all and say Do as I say, not as I do or Dont you do It this way unttl you are as experienced as me The problems with work plannmg may be being caused by supemsors (the solutron 1s to tram them to plan better, see trammg) Also too close personal supervision may cause resentment, m competent workers, at being watched and felt not to be capable of doing the work on therr own
l l l l

Ease of use Easy Counter-productlmty No side-effects Effectiveness Low effectiveness Time scale to success Immediate impact

Medium side-effects effective months

Difficult Major side-effects I-hgh effectiveness More than a year

50

Vlolauon Handbook

Version 1 2

3.2.1.2

2. Analysis of existing violations

Consider m a direct approach which vlolatlons are takmg place and why - see Part IV

Advantages
This approach gets straight to the heart of the matter, supposmg that vlolatlons are not hidden and are capable of being made open to dlscusslon

Dmdvantages
One of the North Sea studies found that while the maJonty of people were wolves, they also found that the malorlty did not have a recent history of vlolatlon, sheeps clothmg If there are no vlolattons to analyse, the lmpresslon can easily be gamed that there IS nothing to worry about
l l l l

Ease of use Easy Counter-productlvlty No side-effects Effectiveness Low effectiveness Time scale to success Immediate impact

Medium side-effects effective months

Dlficult Major side-effects High effectiveness More than a year

51

Vlolauon Handbook

Version 1 2

3.2.1.3

3 Analysis of p-e-conditions for violation

Take the proactive route, analyse the make-up of the workforce, the attitudes of operators, supervisors and managers with a view to uncovering the grounds for vlolatlon Look for whether opportumttes are to be found, and people praised and rewarded for takmg them (successfilly) Examme whether the quality of the plannmg process IS hable to let people down Just at the moment that they feel they need to contmue m order to get theJob done Analyse the current expectations and find out why they arlse Use the dlagnosls sections of this handbook to ascertam m whch areas there are likely to be problems Use the remedy sectlons to select the appropnate remedy approaches Impacts may not be unmedlate, but are hkely to be long-lastmg Advantages Gets at even more than the direct approach, as It allows analysis before anythmg has required an mvestlgatton of why a vlolatlon took place Drsadvantages Requires slttmg down and bemg proactlve when there appears to be no reason Justlfymg proactlve work IS always hard as nothmg IS wrong Requires people to use the mformatlon here systematlcally, when they may take one look at the names of the factors and unmed~ately draw their own conclusions without having the drsclphne to actually follow through
l l l l Ease

Easy Counter-productrvrty No side-effects EfSectrveness Low effectiveness Tune scale IO success lmmedlate Impact
of use

Medium side-effects effective months

Dlfiicult Major side-effects High effectiveness More than a year

52

Vlolauon Handbook 3.2.1.4 4 Structured Discussions

Version 1 2

One major method that can be used 1s a structured dlscusslon session with the workforce In small umts or at the managenaUsupervlsor levels, all mvolved can and should be present In larger groups a selection has to be made, but with more than one session more people can be directly and personally mvolved Unless there 1s a specific structuring rule a dlscusslon session IS hkely to turn mto a bitch-session, with the air cleared but no progress Thus IsJust the kmd of lack of progress that creates negative expectations The structured approach IS related to Tnpod DELTA generation and dlagnosls sessions The task IS to discover the reasons for failure, to see If there are specific underlying causes that can be ldentlfied for treatment (cure the disease, not Just the symptoms), then to find remedial actions and carry them out Method z Select a dlscusslon group => Start with a general dlscusslon of vlolatlon q Use the techmque of askmg each mdlvldual to describe a vlolatlon they personally perform (dont offer the chance to say IF they nolate, but HOW they violate everyone does It and m such a setting it 1s easier for everyone to become more honest) => Agree that vlolatlons are present and need to be fixed 3 Go round the group m search of routme vlolatlons Collect at least one from each member Group leader goes first to break Ice If necessary 3 DISCUSS the reasons for these routme vlolatlons Use the Diagnosis Sections to help 3 Go round the group m search of sltuatlonal vlolatlons Collect at least one from each member If at all possible 3 Inventonse the sltuatlons s Pnontlse problems m terms of danger and ease of solution =B Make an actron plan with no more than 10 action Items to be fixed m l-3 months Advantages Capable of uncovering all sorts of problems and, simultaneously, commg up wth remedies and Improvements Potentially good fnr group morale Dlsadvantages Unless It IS well run and held m an open and blame-free environment, hable to turn mto unproductive argument, dommated by people who want to get their personal hobby horse fixed and run by managers who want to blame everyone but refuse to accept their own role m the sltuatlon

53

Vlolatlon Handbook
l l l l

Version 1 2 Medium side-effects effectwe months Difficult Major side-effects High effectweness More than a year

Ease of use Easy Counter-productwty No side-effects Effectiveness Low effectweness Time scale to success Imrnedlate Impact

54

Vmlauon Handbook

Version 1 2

3.2.1.5

5. Procedures

Vlolatlon IS all about deliberately not followmg procedures The solution IS to remove unnecessary procedures, improve those that are found to Induce vlolatlon because they are too long, dlffkult or n-relevant Routme vlolatlons are almost mvanably caused by Ill-thought out and incorrect procedures, so sensible behavlour becomes vlolatlon and, as a result, vlolatlon becomes a habit Sltuatlonal vlolatlons are caused by havmg procedures that have not been well enough thought through for all the sltuatlons m which they should apply Improvmg those procedures can only be beneficial Advantages Getting the procedures nght removes all the routme vlolatlons and should remove all the sItuatIona ones as well With a considerably reduced frequency of vlolatlon the culture will shfi of Its own accord At this point any vlolatlon should be much more obvious and, If felt to be necessary, the direct management approach (See 11) with a procedure for vanatlon due to circumstances will be easier to operate Getting the procedures nght ~11 almost certainly have considerable benefits m terms of operating efficiency Disadvantages Re-exammmg and rewntmg procedures IS a lengthy and potentially expensive process The main reasons why procedures have been changed m the past IS the result of a single incident (Shuttmg the stable door after the horse has bolted), but the effect IS to overlay a procedure with controls for the highly Improbable and to lose all sense of what 1s really vital Because of the mdlvldual mcrdents it can be hard to argue against removing what seems a sensible precaution, the overall view IS hard to mamtam when m the middle of reviewing an old but suspect procedure
l l l l

Ease of use Easy Counter -productrwty No side-effects Effectrveness Low effectiveness ike scale to success Immediate Impact

Medium side-effects effective months

Dtfficult Major side-effects High effectiveness More than a year

55

Uolatlon Handbook

Version 1 2

3.2.1.6

6. Direct Management

If you accept that vlolatrons are gomg to happen, and that the majority of your workforce may be Wolves, born violators, then one solution to ensuring safety and effectiveness IS actrve and direct management of vlolatlon behavlour I e Allow vlolatlon by creating a procedure that makes vanatlon pemssable under strictly defined precondltlons

2 Ensure all commumcatlon devices are avallable and operatmg (telephones, handheld radios, line of sight) = Get people to recogtuse when they are vlolatmg => Recogmse a vlolatlon-provokmg situation 3 Apply Rule of Three to identify extra sources of error z Inform all tn vlcunty that rules are bemg broken, put up flags, flashmg hghts etc = Note the reasons why vlolatlon appeared necessary s Routme - IdentlfL and consider changing the rule a SItuatIonal - identify and remedy sltuatlons => Optnmsmg - consider for Improvement of procedures *FIX necessity for next time and before next tzme 3 Hangmg offence IS NOT followmg ths procedure when vlolatlon occurs Advantages Takmg this approach allows work to continue, opporturuttes to be taken and improvements to be Implemented Dzsadvantages
l

This IS such a drastic approach It can really only apply once most of the reasons for vlolatlon have already been fixed Because it generates a fi.111 paruc mode people may be tempted to sktp It (Just ths once), which IS why failure to use ths procedure has to be a full hanging offence
l l l l

Ease of use Easy Counter-productlwty No side-effects Effectiveness Low effectiveness Time scale to success Immediate impact

Medium side-effects effective months

Difficult MaJor side-effects High effectweness More than a year

56

Vmlauon Handbook

Version

1 2

2.3.3 Division ZZ Remedies 2.3.3.1 7. Selection

One approach to managrng vlolatlons mvolves selectmg a workforce that ~111 not violate to start with Selection would mvolve creating a comphant workforce that did not question the procedures and wanted to follow them to the letter Advantages A docile workforce ~111 do exactly as It IS told There are some safety cntlcal tasks that real& require sheep rn sheeps clothmg To the extent that these mdlvlduals can be identified, as by usmg the questions here m a selection test, this route IS possible Disadvantages The workforce IS hired, m the first place, to perform work that has not or can not be automated (I e fi~lly procedurahsed) A docile workforce wtll do exactly as It 1s told, so the quality of the plannmg has to be even better than when shortcommgs can be covered by people taking direct mltlatlve As plannmg failures form a much more potent cause of vlolatlon than simple failures of intention, selecting a totally obedient workforce may prove costly and even dangerous In the US nuclear mdustry it was found that the maJorlty of procedure causes tn incidents were caused by people followmg the procedures, not by deliberate farlures to obey1 So, tf you want to go ths way. remember that the money still has to go mto optmusmg plannmg and procedures, which may prove expensive when we now require perfect procedures and plans, rather than those Just good enough for competent workers to do the nght thmg
l l l l

Ease of use Easy Counter-productlvlty No side-effects Eflectrveness Low effectiveness Tzme scale to success Immediate Impact

Medmm side-effects effective months

Difficult Major side-effects figh effectiveness More than a year

57

Vlolatlon Handbook

Version 1 2

2.3.3.2

8. Training

Trammg can mvolve safety-specific trainmg or more general techmcal or professional trammg Trammg may be class-room or on-the-Job and ~111 spread across a persons career The expectation IS that people can be persuaded not to violate The reahty looks more complex Trammg can mform people about potential consequences of actlons, help them to thmk complex sltuatlons through The down side IS that trammg increases peoples feeling of competence and therefore their confidence m theu- own actions (powerfulness) When someone deviates from a procedure, clamung that they wrote the procedure m the first place, you have an example of both why they were hired and what damage they might do F Advantages Trammg operates to mfluence attitudes by provldmg people with mformatlon about consequences and sendrng messages about acceptable norms The lessons learned can be applied to mfluence attitudes, mtentlons and expectations and, therefore, behavlour Disadvantages The evidence IS that it IS bemg well tramed, or at least thmkmg you are, that forms one of the contributing factors to vtolatlon Competence 1s seen as an mvltatlon to excerclse ttutlatrve and that 1s what people are bred for

l l l l

Ease of use Easy Counter-productlvlty No ade-effects Effectiveness Low effectiveness Time scale to success Immediate impact

Medium side-effects effective months

Difficult Major side-effects I+gh effectiveness More than a year

Vlolahon Handbook 2.3.3.3 9. Detection

Version 1 2

Detectmg vlolatlons can be done after the event, usually m an accident report, or by direct supervision, as on walk-rounds Advantages A vlolatlon that has been detected can be analysed (see remedy 2) Really dangerous mdlvlduals, those with poor Intentions, can be detected and dealt with Other people can help to uncover the grounds for their vlolatlon The mam reason for using detection appears to be as a preventative measure, I e supervisors are there to be able to detect violation (See remedy 1) Drsadvantages Does everyone do th& How many people will own up to deliberately breakmg rules? A system than rehes upon the detection of violation. except as an mcldental. will be very ineffective and probably much too late For self-detection see remedy 10 When using supervlsors as detectors It IS all too often not clear what will happen when they do find them This problem IS often exacerbated by the knowledge that it IS sometimes the supervisor who IS the pnme violator (see 1)
l l l l

Ease of use Easy Counter-productrwty No side-effects Effectrveness Low effectiveness Time scale to success Immediate impact

Medium side-effects effective months

Difficult Major side-effects tigh effectiveness More than a year

59

Vlolatlon Handbook 2.3.3.4 10. Reporting

Version 1 2

Reporting IS a form of self-detection that 1s like a near-nuss system Such systems are mdlcatlve but haphazard Reportmg on others can be dangerous m most cultures While reporting should be encouraged, this will only occur if there IS no blame culture In the aviation industry there are confidential reporting systems, such as CHIRP, but even here the legal consequences mean that not all such mformatlon can be truly confidential Advantages Makes people aware of what they are supposed to report Similar to Unsafe Act Auditing Disadvantages Requires a very open company culture, which IS rare and difficult to put mto place from outside among some contractors Unsystematic as It requires mdlvlduals to have a sufficiently deep understandmg to recogmse what It 1s they have to report Could be nusused to get other results, such as cntlclsmg supervisors and managers for forcing violations
l l l l

t-

Ease of use Easy Counter-productwrty No side-effects Effectiveness Low effectiveness Tzme scale to success Inunedlate Impact

Medium side-effects effective months

Difkult Major side-effects figh effectrveness More than a year

60

Vlolabon Handbook 2.3.3.5

Version 1 2

11. Incentives

Paymg people to behave properly, or at least finmg them if they break the rules, 1s a classIcal techmque Smularly within orgamsatlons careers can be influenced by the extent to whch people do or do not comply As a techmque rt seems simple and effective The problem with mcentlves IS that they are applied by people who thmk that they ~111 work for them, but they dont break the rules or the system doesnt apply, wtule they are untended for other people (e g , Prison IS felt to be an effective threat for the law-abldmg, who assume it ~111 be equally effective as a deterrent for the cnmmal The size of the prison population seems to disprove the Idea) The use of mcentlves m a simple form Implies that mtentlon IS the determmmg factor, as mcentlves will work on the mtentlons As the studies have shown that wthm Shell companies offshore mtentlon IS already a solved problem, and the causes for vlolatlon he elsewhere, mcentlves are much less likely to be effective than naive outsiders would imagine A more subtle approach mvolves maupulatmg incentives for compliance, as suggested m the ongmal Bendmg the Rules report (Reason et al, 1994) Here again the problem IS that compliance appears to be of little effect m determmmg whether someone ~11 vlolate, as the Important factors (Plannmg, Opportumty, Power-filness and Expectation) can serve to overrule all but the most compliant Incentives can work, very indirectly, on the development of Personal Norms, but this will be a slow and not very sure process Advantages Appears to be what 1s needed Acceptable to outside bodies as an mdtcatlon of takmg thmgs seriously Disadvantages Not very effective and potentially counter-productive, leadmg to mls-reporting m order to obtain the benefits Thus will make thmgs worse rather than better as what httle mformatlon there IS may be suppressed to obtam the short term rewards

l l l l

Ease

Easy Counter-productmty No ade-effects Effectiveness Low effectrveness Tune scale to success Immediate Impact

of use

Medium
side-effects

effective months

Difficult Major side-effects tigh effectiveness More than a year

Vlolauon Handbook

Version 1 2

3.3 Remedies and Problem Areas All of the remedies described above can be applied Some are less effective than others Trammg and SelectIon have probably little margmal benefit, what advantage there was has already been obtained Incentives are superficially effective. but suspect SupervIsIon, DetectIon and Reporting should be employed, but with recogmtlon of the disadvantages The three analysis techmques, Direct Analysis, ProactIve Analysis and Structured Dlscusslons should be apphed to see exactly what problems there are Finally lmplementlng Improvements m Procedures should have a maJor effect Once these have been started and shown to have effects, the Direct Management approach can be star-ted The next sections define briefly the appropnate remedies for problems Identified The most usehI remedies are tighllghed m bold 3.3.1 Violating Behaviour Tell them not to This IS not very effective on its own Try gomg back toward the root causes The message management has to send 1s that wtule mttlatlve may be desired, vlolatlon 1s not A vlolatlon management approach mvolves 1 Recogmsmg that someone (oneself) 1s about to break or bend a rule This becomes easier when vlolatlon 1s no longer routine and the rules are well known and not impossible 2 Callmg a time-out to signal a situation 3 Dlstmgulsh between sltuatlons offenng no choice, an obvious improvement or a new problem 4 Either create enough time and resources locally so as not to violate or, 5 Flag a variance and mform everyone what IS gomg on The most effective defence agamst an ongoing vlolatlon IS the knowledge that it IS happenmg, who, where and when Radios, special signals etc all sewe thus purpose See Direct Management 6 Reduce the chances of errors ocumng elsewhere that have any chance of mteractmg 7 Recogmse that most potentially lethal combmatlons have been Ignored as too unlikely, the specific combmatlon of vlolatlon and error IS therefore hard to imagine m advance and certainly defined as below the ALARP threshold In the SMS a SelectIon l Trammg l Supervision l Detection l Reporting l Analysis of existing violations l Analysis of violation potential l Structured dlscussions l Procedures l Incentives l Direct management

T t-

62

Vlolauon Handbook

Version 1 2

3.3.2 Intention
The problem with intentions lies m one or more of the three components 1 Attitudes 2 Norms 3 Feeling of control 1 Attitudes - These are defined by the the thmgs people find important and what they feel are ummportant So we need to discover If people feel that it 1s safety that IS Important, followmg procedures, not hurting colleagues 2 Norms are group and personal - If people have poor Intentions because they thmk their mates/colleagues would have them behave m a certain way7 Does a person have a personal hnut beyond which they are not prepared to go7 3 Feelmg of Control - Poor mtentlons anse when people feel that they lack control over their actlons Why 1s ths so7 Where does this feelmg anse7 The primary solutions are selectlon, trammg and perceived rewards 1 Select people with value systems that are right to start with - dont expect to take people and then try to make them better overnight Unless there 1s no choice choose wellmtentloned people This selects on Norms and Attitudes 2 Trammq tells people about what IS important and predonunantly Influences attitudes by mformmg people about the consequences when thmgs go wrong 3 Rewards shape the way people want to behave Those who wish to get on , to start with If all else falls Fn-e them
l l l l l l l l l l l

Selection Training Supervision DetectIon Reporting Analysis of exlstmg violations Analysis of vlolatlon potential Structured dIscussIons Procedures Incentives Direct management

63

Vlolauon Handbook,

Version I Z

3.3.3 Expectation Expectatron has been found to be the smgle largest factor determmmg whether people violate or not We expect to find ourselves vtolatmg because the srtuatron m the past seemed to requn-e It and now because nothmg has changed The old srtuatron still exrsts and the same pressures will arise again Unless somethmg IS changed and seen to have been changed fol the better, expectatrons wtll remam as they are and vrolatrons wtll remam likely A manager or supervisor faced with an expectatron problem can expect an uphrll struggle because they are mstrumental m creatmg the condrtrons for vtolattonl Tlus IS what commrtment m Enhanced Safety Management IS all about

Expectatron 1s made up of a number of different components Each of these requires a different solutton Two approaches can be taken 1 Analyse the srtuatron on your own Draw up a one year plan and Implement tt The best way to do thrs IS to have the plan as one of your personal tasks and set measureable targets for success Make sure that the targets are feasible and that the changes are vrstble to those liable to have the poor expectatrons, otherwise they wont believe anythmg has been done and the expectatrons wtll be made, tf anythmg, worse rather than better Wlule some problems may be qurckly solved once identified for what they are (we all know what they are, the problem IS we fall to understand Just how tmportant they are), some will take time because fixmg expectation requires both fixmg the root cause and then having those affected change then behavtour as a result of dtscovenng that real change has taken place For instance, haste IS a major problem The solutron for haste lies m preparation planrunq and manpower allocatton Until a full cycle of work has been gone through people may not believe that real change has taken place With a 12 month productron- maintenance cycle tlus places natural hnuts on getting instant results 2 Get the Workgroup, or a representative selectron, together + Have a structured discussion armed at rdent@mg specific examples, make a dragnosrs of the underlymg and nnmedrate problems Start with having each mdrvrdual put forward an example of a bent or broken rule, a procedure that they, personally, have knowmgly failed to follow fully Once the group has a full hst,start to search systematrcally for the reasons why the rules were bemg broken Use the questrons m the defimtron of expectation and then work backwards through the different components as a way of structurmg the dragnosrs drscussron + Go through the exercise m Part IV to tdenttfjr routme rule-breaking and srtuatlons whrch seem to require vrolatron for success Revrsrt the procedures + Discover the Personal Norms

64

Violation Handbook + Check on IntentIons to make sure there Isnt an Intention problem* SelectIon l Trammg l SupervisIon l DetectIon l Reportmg l Analysis of existing violations l Analysis of violation potential l Structured discussions l Procedures l Incentwes l Direct management

Version 1 2

65

Vlolauon Handbook 3.3.4 Planning

Version 1

Fnung poor plannmg requires first ldent@ng the problem m more detail Use the breakdown Preparatron planning - The lmtlal defirutlon of the work to be done Job Planmng - The local details of how the Job IS to be done, by whom etc Actlon Plannmg - The situation when the unexpected arlses and plans have to be changed Examme the last (n?) tunes that thmgs went unexpectedly and attempt to iden@ whether 1 The work actually needed to be done 2 The original planning was camed out with hands-on knowledge 3 Previous experiences had been taken into account + Was enough time allocated? + Were enough people allocated7 + Were the nght people allocated7 + Were the nght tools and equipment available? + Were possible problems considered m the plan Was plannmg ad hoc, opporturustlc, top-down or hnear What tt.lls means 1s Is the plannmg process Itself planned? Ad hoc and hnear plannmg processes are to be avolded Top-down plannmg needs to be tempered by reality + Ad hoc planrung Plannmg takes place on a moment-to-moment basis The work plan 1s very rough and ready and there 1s conslderable rehance on experience + Oppofturustlc planmng Plans are easily set aside when opportumtles are discovered Not necessarily bad, but needs to be controlled + Top-down plannmg Planning starts with the most general activity and IS broken down m increasing detail Problems arlse when the level of detail IS either too little or, sometlmes, too much Plans have to be flexible enough to cope with unexpected situations and top-down planmng 1s ngld + Linear planning Plannmg starts at the beginning and goes on till the end This approach 1s liable to get into temble problems If It IS found it must be replaced
l l l l l l

* 1 r-

Selection Trammg
Supervision

Detection Reporting Analysis of existing violations

66

Uolauon Handbook

Version 1 2

l l l l l

Analysis of vlolatlon potential Structured dlscusslons Procedures Incentwes Direct management

67

Uolauon Handbook 3.3.5 Attitudes Flxmg poor attitudes requires bnngmg people to understand what have to correct emotional response to situations
IS

Version 1 2

really important and to

1 What is really important Ttus 1s trammg, especially much safety-related trammg, whch concentrates upon the consequences First people have to know what those consequences are 2 How should you feel. Thus IS much harder to fix Our emotional responses are the product of our genes and our early environment What you can do IS to try and apply the desired emotional response to specific sltuattons byframrng situations to appear For instance, situations can be described m positive or negative terms A simple medical analogy IS found In an expenmental task where people must choose treatments If they are framed in terms of 65% survive then they make quite different choices than when the task IS framed with 35% die Logically there 1s no difference, m practice people are really affected by the use of the word survive vs die SelectIon Training Supervision Detection Reporting Analysis of existing violations Analysis of violation potential Structured discussions Procedures Incentives Direct management

68

Uolauon Handbook

Version 1 2

3.3.6 Noms There are two types of Norm, Group and Personal Group norms require careful selectlon If the workforce operates with norms that are dlstrnct from those required for safe and comphant behavlour, one should consider breakmg groups up and selecting figures For Instance people are more likely to change their dnvmg behavlour for a well-known racmg dnver (Damon Hill or Michael Schumaker) than for an anonymous pohceman Personal norms are one of the most Important defences against vlolatlon These will have been set early on Younger people may still be In a more labile stage, whch means that they pick up and retam the values by reference to status figures Umversltles and professional orgarusatlons represent slull norms Engmeers for instance, should have lnmts past which they will not knowmgly go. defined by Good Engmeenng Practice Do not expect supervisors to set group norms They can, however, set personal ones Selection Training Supervision Detection Reportmg Analysis of exlstmg vlolatlons Analysis of violation potential Structured discussions Procedures Incentives Direct management

69

Vlolauon Handbook

Vernon 1 2

3.3.7 Feeling of Control Powerlessness Powerlessness means that people feel they have no control over then- actlons, the external environment IS doing the dnvmg The fixes ~11 either mvolve altenng the setting so that alternatives are obvious, or changing the person by letting them take control when the situation arises
l

1 u- 1 r

Create situations in which the feeling doesnt have to mse by plannmg m advance At the workface level techruques like JSA (Job Safety Analysis) can help by lookmg at what has to be done Just before doing it FIX the situations so that altematlves are always avalable Thus IS a task for supervisors and planners Change the procedures appropnately Learn to recogmse the feeling, so that It 1s possible to inform others that an lmposslble situation has ansen Tram people to reahse what IS going on If you feel powerless, thus may be a time to involve the Direct Management approach Report situations where lack of choice becomes a problem If supervisors arent told, they may never know any other way Have a procedure for stopping or tellmg others, 1 e the Direct Management techmque needs to be known about and, ideally, practiced In advance

Powerfulness The nature of powerfulness means that people have to learn that, no matter how good they know they are, the feeling of powefilness and control IS itself potentially dangerous
l

l l

Learn to recogmse ones shortcommgs and recogmse the feeling that one IS takmg control when it may be unwarrented Learn to use the Direct Management techmque Hold Structured Discussions to discover who feels powerful, when they feel powe&l and where they are likely to feel that they can get away with clearly dangerous short cuts Tram people by putting them In situations where they create their own problems and learn to understand that powetilness has a downside Selection Training Supervision Detection Reporting Analysis of existing violations Analysis of violation potential Structured discussions Procedures
70

1 h

Violation Handbook

Vernon 1 2

Part IV
4.0 Introduction to Reactive Approach Before embarkmg on a rmsslon to ensure compliance with all existing rules and procedures, It may be necessary for managers to asks themselves a number of questIons
l l l l l l

Do employees know and understand the procedures? Do we need all of these procedures? Are there sltuatlons when it IS Impossible to apply procedures7 Does the job itself encourage vlolatlons7 Is It possible to have a procedure for every sltuatlonv Are there altematlves to procedures?

We will return to these questions m the summary. but it IS worth remembermg when reading thts guide that deviance from procedures 1snot always bad - somettmes It saves lives, and so It follows that comphance wth procedures IS not always good - sometimes It k11ls Ed Punchard (1989), a survivor of the Piper Alpha disaster describes the behaviour of people on the ng durmg the disaster All over the ng, people were mstmctlvely followmg their tratnmg and emergency mstructions In the absence of any form of announcement, most were trying to make their way to the galley to muster, have a head count, and take mstructlons After all, that was what they were tramed to do (p 128) Tragically the accommodation was m the hne of the fireball erupted, meanmg that the majority of people who had comphed with the emergency procedures did not survive 4.1 Identifying the problem

The prevalence of the five types of vloiatlons can be identified by asktng employees to indicate their level of agreement with a number of pertinent questions Each violation type IS consIdered below and a list of statements ISpresented to allow the orgarusatlon to assessthe extent of the problem at a particular work site or locatIon A blank question sheet IS provided m Appendix 1 All Items are scored on an 5 point agreement scale rangmg from 1, strongly disagree to 5, strongly agree A score for each vlolatlon type can then be obtained as follows Sum respondents ratings and dlvlde by the number of respondents to arrive at a mean rating for each Item Sum the Item means across the 10 Items to arrive at an mdex of the extent to whch each vlolatlon type IS a problem Step 5, dealing with exceptional vlolatlons contains only 5 statements and as such a mean score IS obtained by dlvldmg the overall mean score by 5, rather than 10 Possible strategies for promoting procedural compliance are offered m relation to each of the vlolatlon types

72

Violation Handbook

Version 1 2

4.2 Unintentional

violations

The first step m ensurmg that safety IS promoted by rules and procedures IS to estabhsh that eMsting rules are correct, comprehensible and not overly burdensome Moreover, unless staff are aware of the rules and have been tramed m their use, violations are likely to arise simply through lack of awareness and understanding 1 The rules are not written m simple language 2 Some rules are very difficult to understand 3 Some rules are so complex that I lose track 4 Rules commonly refer to other rules 5 Some rules are factually incorrect 6 No system eusts to check people understand procedures before they are used 7 I sometimes have difficulty getting hold of written rules and procedures 8 I sometImes come across a rule I did not know about 9 I sometimes fall to fully understand whch rules to apply 10 I have rules for tasks I will never have to do A mean agreement rating with these items m excess of 3 0 suggests that the orgamsation nught benefit from a consideration of the rules themselves and/or perhaps the trammg process The orgamsatlon mght also want to consider the way m wluch rules are dlstnbuted and used Are rules and procedures to be kept on the shelf and referred to only when a problem ansesv Does the employee need to know all the rules m detail, m which case personal copies may be essential? figh scores on questIons I to 5 rather than 6 to 10 suggests that the problem IS with the rules themselves It 1sobviously important that rules are written with the audience m mmd and also take the best form for the task On some occasions where a series of actIons are required, checklists are the best means for ensuring the task IS can-led out correctly For other sltuatlons a diagram may better represent the mformation figh scores on questIons 6 to 10, but not 1 to 5 suggests that it 1snot the rules per se that are at fault, but the distnbutlon and trammg with respect to the rules Developing and pubhshmg good rules 1sessential, but If employees cannot get hold of these rules or are not aware of their eMstence, vtolations will occur simply through lack of awareness It may be helpfL1 when tackling Step 1 to consider what makes a good rule Twmmg and mers (199 1) define a rule as a genera! norm mandating or guJdmg conduct or actJon m a given type of sJtuatJon (p 13 1) In their book How to do thmgs with rules, they examme rules fi-om a legal point of view and discuss the features of a good rule A techmcally perfect rule (p 200) 1 has a clear and acceptable aim 2 JS clearly and precisely expressed, leavmg no room for doubt about Its application m any possible case (loopholes) 3 has a scope whch IS coexistent with its purpose 4 IS certam to aclueve the purpose, without undesirable side effects

73

Vlolauon Handbook

Version 1 2

If rules were constructed wlthm this framework there would be httle reason for noncomphance, as there would be no rational grounds for makmg exceptions to such rules That rules are violated lmphes that the rules themselves are producmg mfnngements by then- very eustence Reducing unmtentlonal vlolatlons can be achieved both reactively- by rewntm-g and redeslgnmg eMstIng rules and ensurmg that these are effectively dlstnbuted and a hgh level of awareness 1sencouraged. and proactIvely by seriously consjdermg the need for new procedures Procedures are written to increase the predlctablhty of employees behaviour They are a means by whch orgamsatlons can attempt to control behaviour There are altematlves to ths form of control (see summary sectlon)

74

Violauon Handbook 4.3 Routine violations

Vernon 1 2

Whale It IS important that procedures are user-i%endly, this IS not sufficient to ensure compliance Zeltlm (I 994) wanting to understand why people failed to follow safety procedures, tested whether it was due to faulty commurucation or r&y decisions Zelthn found that even when mstructions were well commumcated and understood, risks were taken He concluded that If a worker places hgh value on hs or her time, convetuence, self-image, status among peers, and other factors, and If he or she estimates the probability of being injured by disregarding an mstruction as sufficiently low, then tt IS likely that the mstruction will be ignored despite its clarity of presentation Vloiations are likely to become routine when the perceived costs of rule followmg outweigh the apparent benefits These violations may then become the normal way of workmg for some, if not all of the work force, and are ofien perceived to be associated with little risk, particularly if petiormed by a skJled person Routine vlolattons allow the job or task to be performed tn a way that 1scaster for the mdlvidual and therefore gets the job done more quickly These vlolattons are seen to be a better or more efficient way of domg the job and so employees feel they are justified m adopting these workmg practices The next step, therefore, havmg ensured that employees are aware and understand emsting rules, ISto be cntlcal of the rules Violations that occur routinely are often a good indicator of rules that are beheved to be unnecessary or overly restrictive It may be the case that these beliefs are unfounded and therefore hazard awareness training nught be necessary but, more ofien than not, some rules emst that can be flouted with very little, if any, risk of any negative outcome Often groups of staff develop alternative and sometimes more efficient practices that the orgarusatlon mght want to recogmse as correct practice The followmg items can be used to ident 11 12 13 14 15 16 17 18 19 20 the prevalence of routine violations

The rules do not always describe the best way of working There are some rules that make the job less safe/efficient Infringements of some rules occur all the time I have found better ways of doing my job than those given m the rules Short cuts are acceptable when they mvolve little or no risk I can get the Job done quicker by lgnonng some rules Some rules are only for mexpenenced workers Some rules do not need to be followed to get the job done safely Supervisors turn a bhnd eye to some rule violations Some rules are only of value to protect managements back

Step 2 m the process of improvmg safety through procedural comphance mvolves rat&ng emsting rules to ensure that none are superfluous to requirements TUB filtering process has two very important effects First, it reduces the number of rules, which by default reduces

75

Vlolauon Handbook

Version 1 2

the number of violations Second, the orgamsatlon fimshes with a set of rules that it genumely wants employees to follow Thus comttment to the rules by the orgarusation IS hkely to have a posltlve effect on comphance Rules that the orgamsatlon has m place, but 1s less than keen to enforce (perhaps because of eficiency drives) can have a slgmficant effect on the perceptIons of staff m relation to rules generally Rules may be perceived as being m place only to protect managements back and such beliefs may be endorsed by supemsory and management actions that actually seem to put pressure on employees to break rules It IS obviously important to ascertain which rules m particular employees are refemng to when they agree wtth these statements Thus 1seasy to a&eve by askmg respondents to give an example of such a rule Group discussions, m which small teams of staff convene to discuss the rules and any changes that rmght improve current practice are one way of movmg fomard with ths stage of the process Work to rule situations are symbolic of the dysfimct~onmg of mdustnes wluch apply rules mdlscnmmantly &gorously applied, rules can reduce efFiciency and ensure that deadlines or productivity targets cannot be met As such they are worse than useless, because they bnng the whole system of rules mto disrepute, Hale and Glendon (1987, p 298) Many mdustnes, particularly those wth a long hstory have dracoman systems of safety rules whch ongmated m the nmeteenth century, m response to the occupational safety laws These rules did not offer guidance or ~tiormat~on about what should be done tn a certam situation, but tended to represent behawoural dos and do nots, contravention of whch was pumshable Without enforcement, compliance wth such rules may remam low Preventing routme violations mvolves considering the necessity for each new procedure before it IS Introduced Procedures are contmually being updated to cover changed workmg condltlons and new equipment They are also reused to prohibit actlons lmphcated m some past accident or incident Over time, these procedural changes can become mcreasmgly restrictive The actions necessary to get the Job done often extend beyond these prescriptions In ths situation more procedures ~111 actually mean more violations Some of the alternative means of organisational control are outhned m the summaT at the end of this document

76

Uolauon Handbook

Vernon I 2

4.4 Situational Violations Assuming that Step I and Step 2 have been completed and the rules are now m good shape. Step 3 is to provide employees with the resources necessary to adhere to these rules A large number of violations m many mdustries arise because the situations m which people are asked to work are far from ideal Commonly there are discrepancies between what IS required by the rules and what is possible grven the imuted resources Vrolations of rules occur m an attempt to complete thelob under these constramts Agreement wnh the items below will provide evidence for such vrolations Rule enforcement commonly shps under these conditions, when the supervrsor may feel that acmevmg targets or completmg the lob conflicts with safety consrderatrons 2I 22 23 24 25 26 27 28 29 30 Schedules seldom allow enough time to do the lob accordmg to the rules I sometimes cant get the eqmpment necessary to work to the rules Sometimes condmons at the workplace stop me working to the rules There are mcentives to ignore some rules There are circumstances where managers will support rules being broken StafF shortages sometimes result m rules being broken to get the job done Some rules are impossible or extremely dn?icult to apply Supervisors recognise that m some snuations deviations from rules are unavoidable It IS necessary to bend some rules to achieve a target At certam times rules have to be bent to keep thelob gomg

The problem of situational violattons need to be tackled at a local level Thrs ISbecause factors promotmg these violatrons are locally determmed For example, it may be that a replacement welder has not been found and now a shrfi is operating one man short or n may be that a piece of important equipment has not been fixed or replaced Srtuational factors that promote violations m&de time pressure, high workload, unworkable rules, Inadequate equipment, bad conditions, short staffing and poor supemision Many rules and procedures specify the use of certain equipment or the need for a certam number of people to be present before commencement of the operation Difficultres often arise because, m practice, what should be available is m fact very drfficult to obtain This results m the adoption of alternative/improvised methods of workmg that may or may not be safe Management may not be aware of these local problems, but once hrghhghted they may be able to ensure that the necessary support IS returned and Introduce momtot-mg systems to ensure that this support remains available In addition to providmg support for rule compliance managers need to be aware of the effect of workmg conditions on mdividuals attempting to do their job according to the rules Working conditions mclude a multitude of factors from noise and hghtmg to underfoot conditions and available space The important condmons m relation to rule compliance vaq from Job to lob For example excessive notse may cause problems for people attemptmg to comply with rules requn-mg the use of radros to communicate mstructrons

77

Violation Handbooh

Vernon I 2

Finally, It ISworth consldermg some of the underlymg reasons why vlolatlons ofien result not from a lazy or careless staff, but from the well intentioned loyalty of staff whose pnonty 1sto keep the Job runnmg Workplaces differ, but actIon or state rules are wntten to control behaMour across sltuatlons These are rules that define the exact action or the required state of the system (e g , safety boots must be worn at all times m the shed) Hale (I 993) pomts out that these rules remove almost all freedom of chonze (p 4) There IS often a failure to recogmse that wlule a rule may be essential and totally relevant m one sltuatlon, It IS lmposslble to work m another Thus lack of flex~b~htym a system where actlonstate rules predonunate, means that the focus 1son compliance rather safe behaviour If the rules do not work wlthm a particular sltuatlon. then people ~111 violate m order to do the job There may an unw&ngness to find alternatIve ways of domg the job safely, probably because no dlstmctlon IS made by those m authonty between the safe violation of a prescnptlve rule and unsafe violations Other sItuatIonal factors such as staff shortages, supervlsov pressure, equipment non-avallablhty, poor condltlons and tigh work load, aSect the hkehhood of vlolatlons by mcreasmg the pressure to violate m order to acheve targets and keep to time By takmg seriously problems such as these, management are showing a commitment to safety and the use of rules to acheve safe operations What ofien happens m these circumstances IS that managers and supervIsors Ignore such vlolatlons unless somethmg goes wrong m whch case the rule IS held up m defence Thus puts the burden of responslblhty on the employee who can be left feeling disillusioned about rules generally 4
r

78

Vlolauon Handbook

Version 1 2

4.5 Optimising violations Optm-nsmgvroiattons refiect the fact that people have many goals, one of which may be getting the lob done m a professronai manner Non-functronal goals that compete with tins goal mrght be getting a buzz, havmg a laugh, avordmg boredom or demonstratmg slu1ls to fellow workers When the organrsatron Itself provides mcentrves for rule vrolatrons (1 e bonus for achrevement of targets) such practices may be further encouraged There may also be srtuatrons m which the rules are overly restrrctrve and m fact increase the boredom and monotony associated wrth the Job, so that employees feel that they are not valued and are unable to act on then own mrttatrve Step 4 1sconcerned wrth reducmg vrolatrons that occur m an attempt to make the lob more interesting Achrevmg the balance between varrabriny and flexrbrhty m behaviour and predrctabrhty and reguiarny 1sdrffrcuit It 1sworth consrdermg, however, that over procedurahsatron can strfle the abriny of employees to thmk on their feet m novel srtuatrons Relymg heavrly on procedures to control the behavtour of highly sktlied employees may only serve to reduce the efficiency gains made by employing and trammg such staff Agreement wrth the statements below may indicate that employees are vroiatmg for reasons other than the achievement of production targets 31 32 33 34 35 36 37 38 39 40 Sometrmes the operatmg hmrts prescrrbed m rules are too restnctrve I sometrmes dont know why I have to follow rules I am sometrmes tempted to do work that IS not my responsrbrlrty There are financial rewards to be gamed from breakmg the rules I am not given regular break per-rodswhen I do repetitive and bonnglobs Workmg to the rules removes slulis Devtatmg from some rules demonstrates knowledge of the job There are no personal benefits from strictly foiiowmg rules and procedures Fmdmg aitematrve ways of doing the job makes tt more mterestmg You cant move round here unless you do rt by the book

There are a number of approaches to the reduction of optmnsmg vroiatrons Job redesign, rule redesign and enforcement For pamculariy bormg and monotonouslobs, where vtoiattons are artsmg fi-om a need for variety and reitef from the tedtom tt may be necessary to consrderjob redesign Thrs ISnot the place to consrder these methods m detatl Suffice rt to say that there are three mam kinds oflob design improvement to consider, lob enlargement, lob enrrchment and flexrble workmg groups These methods are well descrrbed m the literature (e g Barley, i983) The second approach to opttmrsmg vroiatrons 1sto mvolve the staff m wrttmg and evaluating procedures Tlus 1spartrcuiariy relevant where staff are hrghiy tramed Empowermg employees and aliowmg them to use then knowledge of the lob to help develop local rules which are approved at a hrgher level IS one way to ensure that there 1sa hrgh level of

79

Violation Handbook comtment

Version 1 2

to rule compliance Group members themselves may then provide sanctions to ensure compliance Finally, putting m place mcentives for rule compliance and safety more generally, rather than simply purushmg rule uolations may promote adherence to the eustmg procedures Anecdotal evidence from British Rail tram drivers suggests that at one time there was a great deal of pnde associated wth knowmg and followmg all of the rules One way to encourage tlus 1sto promote a sense of owner&p of the rules Local supervisors can also influence employees by recogmsmg staff who comply with rules under d&cult circumstances or even sanctlonmg safe alternatives to the rules It IS worth bearmg m mmd that the more people are managed through external controls (extrinsic motlvatlon), the less likely they are either to enjoy workmg or to take personal responslblhty for their behaviour Ttus does not mean that all procedures should be dropped, but suggests that they should only be the nummum necessav to provide the structure and securtty for safe and efficient workmg

SO

Vlolatlon Handbook

Vernon 1 2

4.6 Exceptional violations ExceptIonal violations occur m novel, difficult or dangerous situations m whch there may not be any procedural guidance These violations may occur m rare but tramed for situations, or where an unlikely combmation of mdlvidualiy fanuhar circumstances arise The followmg example of an exceptional vlolatlon 1staken Corn Reason et al (1994) A pair of engineers are mspectmg an 011pipeline One of the jumps down mto the inspection pit to sort out a problemand IS overcome by deadly hydrogen sulphide times I% companion, although tramed to cope with such a situation (I e to radio for help and stay out of the pit), obeys a prmutive impulse and Jumps down to help lus partner- whereupon he too ts overcome ExceptIonal violations very often occur when an employee 1sattempting to solve a difficult and unusual problem for whch there 1sno procedure and m the course of solvmg the problem violates a safety rule hother example comes from shunting A shunter signals to the locomotive driver to propel his vehicles towards a set of statIonaT vehcles that ~111 then become attached via an automatic coupler The shunter suddenly notices that the coupler 1sm the wrong position and will be damaged by impact with the vehcles HeJumps down Corn the platform and attempts to reposItion the coupler, but he does not get out m time and IS crushed by the oncoming vehicles In both of the situations described above, incorrect decisions were made m response to strong emotions In the first, the suffering of a colleague m the second the antlclpated damage to the vehcle and possible disclphnary actIon that might ensue It 1smore dBicult to measure the occurrence of these types of violations because they tend to be rare Agreement with the followmg statements should give some idea of the hkehhood of such violations arising Ttus kind of violation IS possibly more frequent m Jobs that require a great deal of novel problem solvmg (e g mamtenance activities) Step 5 mvolves accepting that it IS not possible to introduce a procedure for every eventuality What staff need to be aware of IS how close they are to the edge of the safety space 41 42 43 44 45 I ofien come across situations with which I am unfanuhar I have not been tramed m the rules to be used IP unusual cn-cumstances I ofien encounter situations where no prescnbec rules are available There are some rules where your natural reaction would be to break them I cotiont many novel problems in my Job

It IS difficult to imagine how to tackle the problem of exceptional violations and the imediate reaction may be to attempt to write rules to cover each novel situation as it ames However such a strategy IS not likely to have the desxed effect as it only serves to make the employee less able to solve problems and deal with novel situations Furthermore,

81

VIolaDon Handbook

Version 1 2

It 1s wrtually impossible to envisage a procedures manual that contains suffkient rules to

cover all possible hazards or hazardous situations ln fact employees require the shJls necessary to deal with such problems whle at the same time being able to envisage the possible dangers associated with their actions It IS trammg, rather than procedures that ~111 help to reduce the occurrence of these rare, yet dangerous violations Staff need to know how close they are to the edge of the safety space A relevant motto here IS thmk before you act

82

Violauon Handbooh 4.7 Summary of Reactive Approach

Vernon 1 2

The five main steps m lmprovmg safety through procedural comphance are as follows Each step 1shnked to one or more of the questions posed at the end of the mtroductlon 1 Ensure that eustmg rules are correct, comprehensible, available and understood Do employees know and understand procedures? 2 Get nd of unnecessary rules Do we really need all of these procedures? 3 Support rule compliance at the local level Are there situations where it is impossible to apply procedures? 4 Make rule followmg rewarding Does the job itself encourage violations? 5 Encourage employees to use then nutiative without taking risks Is it possible to have a procedure for every situation? Havmg camed out Steps l-5 the factors promotmg rule vlolatlons should have been largely eradicated What the orgarusation must now do IS encourage comphance wth the rules Thus can be acheved by consistent and fair enforcement of the rules by local supervIsors Situations should no longer arise m wkch the supervisor feels it IS necessary to deviate ii-om the rules m favour of productlvlty The management should contmuaily momtor the abihty of employees to follow procedures without productivity costs A good set of procedures should facilitate rather than hinder productimty Finally we come to the last question m the list - Are there alternatrves to procedures? The answer very simply is yes Reason et al (I 994) outlined the four prmcipal orgausational controls, extended here, of whch procedures are only one form Administratrve controls. There are two forms of admmlstrative controls process controls and output controls The former are attempts to standardise the work process using prescriptive rules and procedures, often backed up by the momtonng of behaviour and by sanctions for non-compliance The latter, output controls, are necessary when tasks are complex and unpredictable ln contrast to rules and procedures, they ental a feedback control process based upon comparing output measures with orgamsatlonal objectives Social controls. These emerge from mteracttons among members of work groups to produce ways of acting and thmkmg that the group considers legmmate Such norms need not always conform with the goals of the organisation The alrn of the orgamsation

83

Vmlatmn Handbook

Version I 2

1sto ensure that group norms support organisational goals The group discussion technique described m Reason et al (1994) IS one means by whch the power of group Influence can be used to achieve posmve safety rewards Personal controls. These are controls that people exert over thexr own behaviour through mtnnsic motivation and behavIoural economcs (I e maxnmse benefits, numrmse costs) Personal controls are strongly influenced by situational factors and social controls and, to a lesser extent, by adnumstrative controls Therefore, alternatIves to procedures are selectIon of employees who are motivated to work safely and the trammg of employees so that they are motivated to a&eve orgarusational goals without the precise mstructlons contamed m procedures Finally the orgamsatlon can influence personal control by developing rules whch represent best and most eficlent practice Technical controls. These include automatIon and various engineered feature desIgned both to hnut and to replace direct human partlclpatlon m the production process Techmcal controls may also be put in place to defend the system agamst an error or a vlolatlon on the part of the employee Fail safe systems accept the falhbihty of human behaviour and controls such as these are extremely Important where unpredictable behaviour can have devastating effects on the safety of the system

84

Vlolauon Handbooh

Vernon 1 2

5.0 References Bailey, J (1983) Job design and woj k organzsatlon Prentice Hall InternatIonal Free IX J (1994) The Role of Procedural ll?olatlons in Ral/wal Accidents Unpubhshed PhD thesq Umverslty of Manchester Hale. A R (1993) Sgfeo rules p~ocedujal>eedom OI actIon constramt? Paper to Bad Homburg Safety Conference Parker, D , Reason, J T . Manstead, A S R , & Stradhng, S G (1995) Dnvmg errors, dmmg vlolatlons and acldent mvolvement Ergonomics, 38, 1036-l 048 Punchard, E (1989) PIpeI A/pha A Sunwor s Story London W H Allen Reason, J T (1987) The Chernobyl errors BuZZetrnof the Brltzsh PsychoZogzcalSocrey, 106, 321-33 1 Reason, J T (1990) Human Error New York Cambrtdge Umverslty Press Reason. J T , Parker, D , & Free, R (1994) Bendmg the Rules The Varletles, Orqys and Management of Safety fioiatlons Leaden Rqks Umvers~te~tLeaden Report for SIPM Twmmg, W , & Mers, D (199 1) How to do thIngA wzth rules London Weldenfeld and Ncolson Verschuur, W L G , Hudson, P T W & Parker, D (1996) Vlolatlons of Rules and PI ocedures Results of item analysis and tests of the BehavIoural Cause Model Field Study M4.M and Shell Expro Aberdeen Report Leaden Umverslty for SIEP Zelthn, L R (1994) Failure to follow safety mstructlons faulty commumcat~on or t-sky declslons? Human Factors, 36, 172- 18I

85

Uolauon Handbook

Vernon 1 2

6.0 Appendix Reactive Test Items

1= 2= 3= 4= 5=

strongly disagree disagree neither agree nor disagree agree strongly agree

Please circle the number on the scale that best reflects your level of agreement with each statement. 1 The rules are not written m simple language 2 Some rules are vey d&icult to understand 3 Some rules are so complex that 1 lose track 4 Rules commonly refer to other rules 5 Some rules are factually mcorrect 6 No system emsts to check people understand procedures before they are used 7 I sometimes have dBicuity getting hold of written rules and procedures 8 I sometimes come across a rule I did not know about 9 I sometimes fall to fiAly understand which rules to apply 10 I have rules for tasks I will never have to do 11 The rules do not always describe the best way of working 12 There are some rules that make the job less safe/efficient 13 Infimgements of somerules occur all the time 14 I have found better ways of doing my job than those given m the rules 15 Short cuts are acceptable when they mvolve little 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5

1 1 1 1 I 1 1

2 2 2 2 2 2 2

3 3 3 3 3 3 3

4 4 4 4 4 4 4

86

Vlolauon Handbooh

Version 1 2

or no risk 16 I can get the job done quicker by lgnormg some rules 17 Some rules are only for mexpenenced workers 18 Some rules do not need to be followed to get the job done safely 19 SupervIsors turn a blind eye to some rule violations I 1 I 2 2 2 3 3 3 4 4 4 5 5 5

2 2 2

3 3 3

4 4 4

5 5 5

20 Some rules are only of value to protect managements back 1 21 Schedules seldom allow enough time to do theJob accordmg to the rules 22 I sometImes cant get the equipment necessav to work to the rules 23 SometImes conditions at the workplace stop me workmg to the rules 24 There are mcentlves to Ignore some rules 25 There are circumstances where managers wtll support rules bemg broken 26 Staff shortages sometImes result m rules being broken to get the Job done 27 Some rules are lmposslble or extremely difficult to apply 28 Supemsors recogruse that m some situations devlatlons from rules are unavoidable 29 It 1snecessav to bend some rules to achieve a target I

1 I

2 2

3 3

4 4

5 5

1 1

2 2

3 3

4 4

5 5

2 2 2

3 3 3

4 4 4

5 5 5

30 At certain times rules have to be bent to keep the Job gomg 1 3 1 SometImes the operating hmlts prescribed m rules are too restnctlve 32 I sometImes dont know why I have to follow rules 33 I am sometImes tempted to do work that IS not my 1

1 1

2 2

3 3

4 4

5 5

87

Vlolahon Handbook

Version 1 2

responslblhty 34 There are financial rewards to be gamed from breakmg the rules 35 I am not given regular break periods when I do repetmve and bonng jobs 36 Workmg to the rules removes .&As 37 Deviating Corn some rules demonstrates knowledge of the job 38 There are no personal benefits Corn strictly followmg rules and procedures 39 Fmdmg alternatIve ways of doing the job makes it more mterestmg 40 You cant move round here unless you do It by the book 41 1 often come across sltuatlons with which I am unfamiliar 42 I have not been trained m the rules to be used m unusual circumstances 43 I ofien encounter sltuatlons where no prescribed rules are avaIlable 44 There are some rules where your natural reaction would be to break them 45 I confront many novel problems in my job 1 2 3 4 5

1 1

2 2

3 3

4 4

5 5

1 1 1

2 2 2

3 3 3

4 4 4

5 5 5

88