Bending the Rules II

BENDING THE RULES II

Why do people break rules orfail to follow procedures?

andWhat can you do about it?

The Vtolatlon Manual

P T W HudsonW L G Verschuur

Leaden UnwersltV

R LawtonD Parker

J T Reason

Manchester Unwerslp

David

Rectangle

Vlolauon Handbook Vewon 1 2

Contents

0 Management Summary 3

Part I 5

1 0 Introduction 5

1 1 Sclentlfic 5

1 2 Approaches 6

1 3 Remedies 6

1 4 The Theory The Behavloural Cause Model 8

1 5 Remedlatlon 9

1 6 Dlagnoas-Remedy Matrices 1 0

1 7 Conclusion 1 3

Part II 1 4

2 0 Detatled Introduction 1 4

2 0 1 The Nature of the Problem 1 4

2 0 2 The Am of the Manual 1 5

2 0 3 Procedures 1 6

2 0 4 Natural Born Violators Wolves m Sheeps’ Clothmg 1 7

2 1 Why IS Vlolatlon a Problem7 - Human Error2 1 1 Human Error and Danger2 1 2 The Types of Vlolatlon2 1 3 Why IS Vlolatlon so Dangerous?

2 2 What Should You Do3 2 4

2 2 1 Proactive and Reactwe approaches 2 4

2 2 2 What can be done to make thmgs better? 2 4

2 2 3 The Behavloural Cause Model 2 4

2 2 4 Testing the Model 2 7

2 2 5 An Altematwe Model - Detection and Pumshment 2 7

2 3 The Manual - Structure

Part Ill 3 0

3 0 The Proactive Approach - The Diagnosis-Remedy Handbook 3 0

3 1 Diagnosis Section 3 1

3 1 1 Vlolatmg Behavlour 3 1

3 1 2 lntention 3 2

3 1 3 Expectation 3 4

3 1 4 Plannmg 3 6

3 1 5 Attitudes 3 8

1 92 02 12 2

2 8

Vlolauon Handbook Version 1 2

316Not-m 40

3 1 7 Feelmg of Control 43

3 1 8 Oppoltumty 44

3 1 9 Poweriklness 46

3 1 10 Powerlessness 48

3 2 The Remedies3 2 1 Dlvlslon I Remedies3 2 2 1 Supervlslon3 2 3 2 Analysis of Exlstmg Vlolatlons3 2 4 3 Analysis of Pre-condltlons for Vlolatlon3 2 5 4 Structured DIscussIons3 2 6 5 Procedures3 2 7 6 Du-ect Management

3 2 2 Dlvlslon II Remedies 57

3 2 2 1 7 Selection 57

3 2 2 2 8 Trammg 58

3 2 2 3 9 DetectIon 59

3 2 2 4 10 Reporting 60

3 2 2 5 11 Incentives 61

3 3 Remedies and Problem Areas3 3 1 Vlolatmg Behavlour3 3 2 Intentton3 3 3 Expectation3 3 4 Plannmg3 3 5 Attitudes336Noms3 3 7 Feehng of Control

3 4 Summary of Dlagnosls-Remedy Approach

Part IV4 0 Introduction to the Reactive Approach4 1 Identlfymg the Problem4 2 Unmtentlonal Violations4 3 Routme Vlolatlons4 4 Situational Vlolatlons4 5 Optimismg Violations4 6 ExceptIonal Vlolatlons4 7 Summary of the Reactive Approach

5 0 References 85

6 0 Appendix Reactive Test Items 86

495050 1

5 1 15253 r-5556

6262636466686970

7 1

727272737577798183

I

F-1

Vlolatlon Handbook Version 1 2

Bending the Rules IIWhy 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 assumptionthat people ~111 follow the guldelmes and procedures as laid down m the Safety ManagementSystem When this assumption 1s broken the whole basis of the Safety Management System1s 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 theJob without slavishly followmg the procedures,3 Seemg the Opportunities that present themselves for short cuts or to dothmgs ‘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 avery 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 threatof 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 notbreakmg rules, but end up domg so anyway because of the other factors One factor thathelps IS havmg strong personal norms, values, that may be translated as behevmg m goodengmeenng practice

Thus manual provides two ways to combat vlolatlon, proactzve and reactzve

l The reactzve approach mvolves finding out what rules are bemg broken andfixmg the reasons why that IS happening

l TheproactIve approach concentrates upon finding out whether the scene ISset for vlolatlons, even If they are not happemng nght now, and ensunng thatproblems don’t anse

Because most people do not find themselves breaking the rules that frequently, It IS oftenbetter to concentrate upon the proactive approach The field studies have found that themajority of Shell workers offshore m the North Sea can be described as Natural BornVzolators or ‘ WoZves ’ However, at the same time, most people don’t report that muchvlolatlon, and the majority have broken few rules m the previous SIX months, so the Wolvesare Wolves tn Sheeps’ Clothing Wolves are opportunists, lookmg for advantage and takmg

3


it, usually for what they see as the benefit of the company rather than Just personallyManagers 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 orbending IS so dangerous that the process has to be carefUlly managed Most vlolatlons are, mfact, preventable Thus can be done by either changing the rules or avoiding the problemWhat remains must be managed by recogmsmg the types of people who are performmg thework - using the theory - and by understanding why violations are so dangerous

The recommended remedies for violation problems are

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

Other possible remedies are also covered, but these are not recommended because they maybe less effective or have slgmficant negative side-effects Any one remedy will only beeffective for some causes A Diagnosis-remedy matrix IS provlded showmg whchapproaches are effective and what side-effects may be expected

This manual starts m Part I with a simple descnptlon of the problem and the solutionsproposed The diagnosis-remedy matnx forms part of this bnef overview The remainder ofthe manual 1s optional, wntten for people who want to go finther Readers intending toapply the diagnosis and remedy techniques are recommended to read those sections Part IIprovides a more detailed mtroductlon to the theory for understanding vlolatlon and theproposed remedies Part III documents the proactlve approach There IS a section detailingeach contnbutmg factor to help diagnose whether there 1s a problem with e g , expectationor powerfulness, and a section with associated remedies These form the background to thediagnosis-remedy matnx Because of the complexity these sectlons are internally cross-referenced Part IV documents the reactive approach To enable readers to restrictthemselves to single sections, there 1s some degree of redundancy m the text to allow eachsection to be stand-alone

The knowledge m thus manual 1s based upon field studies carried out by Leaden andManchester Umveratles for SIEP, m co-operation with NAM and Expro In thus study atheory was developed to understand deliberate failures to follow procedures The theory hasbeen tested on offshore operators, supervisors and contractors m the North Sea and foundto 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% ofvlolatmg behavlour The results apply to the North Sea operations and, probably, m slrmlarcultures elsewhere m the world While the theory 1s general, the relative importance ofspecific factors will depend upon the local culture

Vlolatlon Handbook.

Part I

Bending the Rules II

Version 1 2

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 1s an 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 IS that, wlthm Shell, these problems have been solved, at least so far 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 doing so anyway because of the other factors One factor that helps 1s havmg strong personal no?mA, values, that may be translated as behevmg m good engineering practice

1.1 The Scientific Background

This knowledge 1s based 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

l The proactwe approach concentrates upon finding out whether the scene 1s set for vlolatlons, even if they are not happenmg nght now, and ensuring that problems don’t anse

Because most people are not breakmg the rules that frequently, It 1s often 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 don’t 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 IS that advantages are there to be taken, but rule- breakmg or bending 1s so 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 l Analjsis of existing violations - find the background behmd vlolatlon behaviour l Ana@sis of violation potential - use the theory to see if people will find vlolatmg easy l Strucfured discussions - discover what IS gomg on and get people to agree on actions l Procedures - improve the quahty of procedures so vlolatlon 1s not seen as necessary l 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 won’t be tempted vlolate to start with l Training - tram people more on the real nsks and the possible consequences l Detection - sharpen the lookout for vlolatlons l Reporting - encourage people to report when they or others bend the rules l Incentives - pay for compliance or purush people who break the rules

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

6


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 1s essential m competence-based workforces. but possesslon of extra knowledge 1s one 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 Verston 1 2

1.4 The Theory: The Behavioural Cause Model

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 one’s own actlons

l The loss of a feehng of control can mean that previously well thought out plans are short- circuited 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” oweriessness $ < ‘@q

Fgure I The Behawoural Came Model


1.5 Remediation

Vlolatlons aren’t 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 IS to fall to understand a complex phenomenon The pay-off for success when rule-breaking, taken together wth the perceived chance of success, 1s much greater than the possible bad consequences, especially when people strongly believe that It won’t happen to them

A number of different approaches to stoppmg or managing vlolatlon are avallable The solution to tackling the disease of vlolatlon 1s to seek out the appropriate remedies and apply them rigorously The remedies can be dlstmgulshed mto Fzrst and Second Dwzszon remedies 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 IS the 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 1s important, powerfulness, 1s Just the one those managers and supervisors probably possess most strongly - whch IS why, when the opportumty arises, it 1s Just those people who exercise lmtlatlve, by bendmg the rules

9


1.6 The Diagnosis-Remedy Matnces

I Dlvlslon

l Supervision - watch more closely over people, plan their work, provide an example l Analysis of existing violations - find the background behmd vlolatlon behaviour, and fix it l Analysis of violation potential - use the BCM to see If people ~111 find vlolatmg easy,&

fix the problems found l Structured discussions - discover what IS gomg on and get people to agree on actions that

get carried out l Procedures - improve the quahty of procedures so vlolatlon IS not seen as necessary l Direct management - a techmque to vary the procedures as circumstances demand

Dlvlslon II

l Selection - pick people who don’t vlolate to start wtth l Training - tram people more on the real nsks and the posstble consequences l Detection - sharpen the lookout for vlolatlons l Reporting - encourage people to report when they or others bend the rules l 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 1s that 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 I 2

Division I Remedies Powerfulness 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

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

Opportunltles and improvements once Identified can be procedurahsed Opportunltles need to be sought for and taken up m the planning process

Supervision

Analysis of existing violations

Analysis of violation potential

Structured discussions

Procedure improvement

Direct management

Expectation 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

Effective If carried through fully Requires action after the talk Crucial to the whole approach Can solve the problems m one

Needs people to do It

Crucial and effective Allows powerfulness

Crucial

Planniug Supervrsors’ major task, rather than watching over people A good supervisor manages Are vlolatlons a response to poor planning? Is this common practice? Dlstmgulsh which part of the planning process needs to be fixed

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

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

Sets social group norms m public

Allows mltlatlve

Contmenls 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


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, don’t 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 be an effective solution, most of the violators m the study were tqmg to get 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


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 dehberate The 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 and fatalities Many 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 1s called 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 assured safety 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 1s broken the whole basis of the SMS IS momentarily put at nsk This manual IS Intended to help manage that nsk

14


2.0.2 The aim of the manual

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 approaches are taken aproactwe approach and a reactwe approach

l The proactive approach IS based upon 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

l The reactive approach IS based upon the existence of vlolatlons and 1s set m a five step approach Thus five step approach to lmprovmg safety by encouraging procedural comphance 1s based 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 extended by the work of Free (1994) Zt IS a particularly usehI classlficatlon m terms of remedlatlon, because It 1s causally 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 serves to introduce the idea that vlolatlon IS a corn wth two faces, the other of whch 1s uutlatlve 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 1s followed There 1s a dragnostzc and a remedzal section Each of the components from the The Behavioural Cause Model 1s presented 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 1s followed 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 represents both 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 ease of use All the proposed remedies form best practice, so whle over- eager 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 people’sperceptlons 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 reasons why 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 reasons why 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 seen as 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

Vlolaoon Handbook 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 balances built m Many procedure problems anse because there 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 needed to perform the procedure as a whole Optional steps may have acquired an obhgatory status that IS mappropnate for the probablllty of the outcome that 1s supposed 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 causes and therefore feelmg satisfied that the problem has been solved, a study offshore (Hudson & Verschuur, 1995) suggested that even m the absence of 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, appearances can be deceptive Another way of expressing ths 1s to 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 Sheep’s Clothing (up-front sheep) were 22 5% of the workforce responding Thus first group represents the guardians of standards

l The largest group, with 33 8% are Wolves m Sheep’s 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 weren’t 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 ensure that 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 won’t 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 011 and Gas industry m the North Sea only highlights the need for timely and effective management The proactive approach 1s Intended 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 because there 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 1s a 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 someone who 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 1s usually 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 because the violator intends harm In such cases we would talk of cnmmal actlvlty

19


2.1.1 Human Error and Danger

There 1s a 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, because they 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 r

Safety Management System As slips are caused by factors such as haste and dlvlded 1 attention, good work plannmg 1s also a highly effective remedy F‘

l Lapses are more dangerous because they may be rmssed, it 1s harder to spot that you haven’t 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 person’s thmkmg that they have done the Job completely, so an extra round of checks 1s no 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

l Mistakes are even more dangerous than lapses because those 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

l Vlolatlons are most dangerous of all They often represent a quite dehberate mtentlon not 1 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 I could be automatically expected to survive all sorts of vlolatlons

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 1s lmposslble 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


the rules Thus may be particularly relevant to new employees or when completmg tasks that require adherence to 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 1s important 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 employee’s 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 because the equipment specified m the procedure IS not available

Optimising violations. Thus category 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 non- fimctlonal aspects of work e g a desire to impress or to relieve boredom These vlolatlons are also assocrated wrth 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 necessary for managers to asks themselves a number of questions # Do employees know and understand the procedures7 + Do we need all of these procedures7 ? Are there xtuatlons when rf IS rmposslble to apply procedures7 + Does the job 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 1s worth remembenng when reading ths guide that deviance from procedures 1s not always bad - sometlmes It saves hves, and so it follows that comphance with procedures 1s not 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 absence of 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 and Jumped

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

l Vlolatlons can themselves be errors when the mdlvldual does not know or understand the rule Thus lack of understanding IS dangerous m Itself, because whle not appreclatmg the risk, people often fall to protect themselves

l Vlolatlons can take people mto new or unpracticed sltuatlons, m whtch the person IS more hkely to make an error

Vlolatlons mvolve knowz~.@y domg somethmg or, conversely, knowzngiy fazlzng 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 1s no 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 1s Just what 1s required 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


i+iizigy The Edge

Fpre 3 7’he operatzng envelope Specrjc events such as A perturbatron In the case of A, wlthzn the Area of Normally

and B each zntroduce some 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 sys tenr rn to normally safe operation In thus representatzon an error may be regarded as rncreaszng the radrus of ou tcomes around the event porn t

Speeding m the desert 1s not 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 speed hrrut, 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 Questronnalre exammed 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


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 happened The 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 analyses are pnmanly reactive, but strong because somethmg 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 hasn’t gone wrong yet won’t necessarily go wrong m the fbture

For vlolatlons there are also two equivalent approaches The reactive approach mvolves lookmg to see what 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 reasons for vlolatlon are so many, It 1s necessary to first find out what and where the problem rmght be, and then to select the remedial approaches that are most likely to be effective There are many perfectly effective steps that can be taken What should be avoided, however, IS takmg unnecessary action 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 1s not really a problem - people have good mtentlons The steps needed to solve intention problems are, therefore, only necessary to 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 1s almost 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

: b

F-

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 Shell’s UK and Netherlands North Sea operations

24


r Igure 4 The BehavIoural Cause Model The sltuatron IS added to the model and the lndrvldual’s 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 mdlvtdual’s behavmg m a particular way, but they are generally separated for the purposes of study For a practical theory, however, It IS necessary to 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 person’s IntentIons and Expectations plus the opportunltles that are recogmsed at the time

Actual behaviour, m this sense defined 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 appearance of unexpected opportumtles

25


Motlvatlon IS seen as a general driver, set at least partially by external goals, and mfluenctng basic inputs such as Attitudes and Norms Motivation almost certainly also influences the ablhty to recogruse oppotturutles when they anse Motlvatlon IS a two-edged sword 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 Element Athtude

Habits

Sod Noms a) Comphance

Soual Noms b) Conformity

Soual Norms c) Personal Norm

Feelmg of Control a) Powerfulness

Feehng of Control b) Powerlessness

Intention

Expectahon

External Goals

Consequences

Mohvatlon Behaviour

Defuution The mcbvldual’s posihve or negahve evaluation of performmg the parhcular behaviour of interest 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 lntemahsed moral norms & anhapated regret reflechng expected emohonal response If tlungs go wrong 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&vldual’s 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 The mtemal drive to act m a gWen way Vlolatlon behaviour

26


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 mdlvldual’s scores on these factors allows the prediction of whether they are likely to vlolate or not The prediction 1s 2/3 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 changes m 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 assumes that 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 Supervzse CC 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% Thus small 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 message 1s clear management of violation requires understandmg of the slgmficant drlvmg factors rather than bemg based nutlal 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 associated tolerances, and the values allow prlontlsatlon on the basis of data rather than impression and prejudice

27


2.3 The Manual - Structure

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 1s expected to be the usual approach

Each section m the dlagnosls part of Part III consists of the followmg elements

l What isX? - e g Intention, Planmng, Powe&lness l How do you define it ? - The questions that you can ask to detect problems l The Components - Pomtmg to more detailed underlymg causes l 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 approaches to 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 IS to 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 Ease of use Easy l Counter-productrwty No side-effects l Effectweness Low effectiveness l 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


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 1s given 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 doesn’t work

The possible approaches to remedlatlon are

Division I

8 Supenwion - watch more closely over people, plan their work, provide an example l Analysis of misting violations - fmd the background behmd vtolatlon behaviour l Analysis of violation potential - use the theory to see If people ~11 find vlolatmg easy l Structured discussions - discover what 1s going on and get people to agree on actlons l Procedures - Improve the qualrty of procedures so vlolatlon 1s not seen as necessary l Direct management - a techruque to vary the procedures as circumstances demand

Division II

l Selection - pick people who don’t violate to start with l Training - tram people more on the real nsks and the possible consequences l Detection - sharpen the lookout for vlolatlons l Reporting - encourage people to report when they or others bend the rules l Incentives - pay for compliance or pumsh people who break the rules

29


PART III

3.0 The Proactive Approach - The Diagnosis-Remedy Handbook

Thus part of the Vlolatlon Manual 1s mtended to allow the reader to dlscover whether thereare grounds for vlolatmg behavlour wlthm the workforce Each section can be readIndependently, although reading one may lead to another For each of the important factorsuncovered and used m the Behavloural Cause Model there IS a dlagnosls section, of one ortwo pages If the specific factor 1s recogmsed as a problem, or a potential problem, thereader can turn to the appropnate remedy sectlon

Do not expect that everythmg WIII be a problem If you read a medlcal dlctlonary and feelyou have every disease you read about you are either already dead or are suffenng fromhypochondria The same applies here If you find a problem with every sectron then youeither have a very senous problem Indeed (wlzlch IS most unlikely) or you are reading toomuch For instance, Intention 1s unlikely to be a serious problem wtile expectation IS often amajor problem To calibrate yourself go out and ask people the questions m the sectlons (thescales are given next to the questlons)

l-‘

The basic scales used for answermg questlons or agreemg mth statements are

Likely Very unlikely - unlikely - neutral - likely - very hkely

Desirable Verydesirable

- desirable - n e u t r a l - undesirable - veryundesirable

AgreementTotallyagree

- agree - neutral - disagree - totallydisagree

Frequency Never - hardly ever - sometimes -often - almost always I4

Approval Stronglyapprove

- approve - not approve/ - disapprove - strongly r‘not disapprove disapprove

3 0

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 dopeople bend or break the rules or procedures’ Behavlour 1s seen as the performance of aplan of action If the plan 1s a good one we expect the behavlour to be ‘good’ as well Errorsarise when local factors, such as haste, overload or dlstractlon allow other actions to beperformed (slips) or essential actions to be slupped (lapses) Poor plans are rmstakes, whchresult In intended but faulty behavlour

Two factors can be seen to shortclrcult the plannmg process and ‘hijack’ behavlourl The sudden dIscovery of highly attractive opportumtlesl The reahsatlon (nustaken or not) that a specific course of action has to be followed, there

IS a feeling of having no control over one’s own behavlour

How do you define violating behaviour?

The followmg statements served to define peoples’ behavlour

How often, m the last SIX months, have you found yourself not followmg the rules7[Frequency]

Posltwe answers to either of theseSometimes 1 deviate from the rulesI sometimes bend the rules

(Agreement][Agreement]

What are the components3

The Behavloural Cause Model defines the effective components leadmg to vlolatmgbehavlour The field studies showed that four components are the most importantl Expectationl Powefilnessl Planningl Opportumty

Because the actual mcldence of vlolatlon may be quite low, certainly what IS reported, It ISImportant to remember the the maJorlty of the workforce will probably be Wolves, so It 1sstill worth followmg back to these four components before stopping Remember, If youthmk that there isn’t a problem, you will be all the more surprised when somethmg doeshappen1Do you have a problem? If you think so after reading this section go to the remedy_section.

31


3.1.2 Intention

What is Intention?

Intenttons are what you have before you actually do somethmg Intentrons define what youwant to do If you behave m a certam way we assume that there 1s some mtentron behmd thebehavtour In general we do what we intend to do, bamng error, rmshap or ctrcumstances Ifwe behave wzthout an obvrous and explamable mtentton, we feel that that 1s a sign ofsomethmg wrong More generally we do thmgs we know we should not do because we havethe 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 theirintentions

How do you define intention?

Intention can be gauged by the followmg Items

l I intend to follow the proceduresl 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 allnegatrvely The first 1s obvrous and failure here is crrttcal (or even crnnmal) The second ismore 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 thecomponent factors discussed below nnght be the root cause of the problem The field studiesm the North Sea suggested that Intentron, on Its own, IS probably not a problem If rt ts, thrsIS where you have to start If It IS not, then exhortation and more trammg are unhkely tomake any difference and would be a waste of resources better employed m solvmg theproblems that still cause people wrth good mtentrons to vrolate

32


What are the components of Intention?

The theory behmd the Behavloural Cause Model assumes three major components toIntention

1 Attitudes Attitudes to work determme what 1s felt to be more or less ImportantThere IS a certain hnk to Personal Norm m that what IS felt to be less important IS morehkely to be dlsregarded If attttudes to work are poor thts can appear as (1) Incorrectassignments of what IS and IS not Important and (3) Incorrect beliefs about what canhappen

2 Norms Personal and GrouD Norms Group norms (my mates/colleagues) mayconflict with desirable behavlour Personal norms may not be strong enough to overcometemptation

3 Feeling of Control Feelings of control over the work and the sltuatlon contrast mth lossof control Is there a can-do culture or are there sltuatlons where people feel they have nochoice?

These all contribute to the formulation of Intentions to act or, m ths case, to violate andbend rules

Feeling of Control I/

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

3 3


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 doJust that Expectation turns out to be the biggest single factor leading to rule-breakmg EvenIf a person’s intentions are good, and there 1s no great pressure, expectations dnvebehavlour 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 timeunless somethmg has clearly changed m the mean time People will usually explam why theybehaved the way they did by referrmg to the circumstances, such as time pressure, lack ofalternatives, or poor procedures These attrrbutzons are not always acceptable to someoneoutslde, 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 I may not always follow the rules In future (Agreement]l In the near fLture I do not expect that I ~111 find myself always workmg accordmg to the

rules [Agreement]l I am sure that thmgs will still go wrong m the future [Agreement]l 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 How likely do you thmk it IS that a sltuatlon WIII arlse In the f&ure m which you willchoose not to follow a procedure7 [Llkehhood]

l All thmgs considered, how likely 1s it that you will actually always follow the procedures?(Llkehhood]

L.

r‘.

suggest that there 1s an expectation problem

Vlolabon Handbook

What are the components of Expectation?

Version 1 2

Indirect AttitudesIndirect Attitudes

Personal NormPersonal Norm

Powerlessness

l Past Behavlour Past behavlour, generated m sltuatlons which led to vlolatlon mthe first place, creates a VICIOUS circle Unless this cycle IS broken the best of mtentlonswill never get off the ground

l Attitudes Attitudes to work determme what 1s felt to be more or less ImportantThere 1s a certam hnk to Personal Norm m that what 1s felt to be less Important IS morehkely to be disregarded If attitudes to work are poor thus can appear as (1) incorrectassignments of what 1s and IS not Important and (2) Incorrect beliefs about what canhappen

l Powerlessness Powerlessness appears when sltuatlons arise m which there appears to beno choice TIE 1s charactenstlc of sltuatlonal vlolatlons Are there such sltuatlons whichare regularly confronted? Indications of this are frequent complamts from different peopleabout the same work, often when different contractors are required to do thelob Thiscan be hldden when routine vlolatlon has become the standard way of solvmg theproblem

l Comphance Whether or not one wishes to Ident@, comphance implies that one 1swllhng to go along Weanng ID passes, even when one IS well known, or filling tn reportsmay seem unnecessary, but can prove vital Is there an atmosphere of non-compliance oreven defiance7 Note that ths IS dlstmgulshed from conforrmtv

l Personal Norm Personal norms form a primary defence against poor expectations Arethere dlstmct boundanes over whxh people will not go? If this IS not the case, then this IS

a serious problem


35


3.1.4 Planning

What is Planning?

Plannmg 1s a well known phenomenon, everyone does It The problem IS that plannmg ISmost notlceable when it goes wrong Planmng mvolves settmg goals and definmg theprocesses needed to attam those goals, asslgnmg resources to perform the tasks, settmgmeasures of success and, finally, placmg the whole m a time-based framework Muchplannmg assumes that the details will be filled m by those present at the time, ths 1s wherethtngs go wrong

Plannmg failures can produce vlolatlons m a number of ways, whch 1s why we dlstmgutshthree different sorts of plannmg process

r’

1

r-

*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 etcaAction 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 tothat failure If they decide to deviate from known and acceptable procedures to fix theproblem, then ths becomes a vlolatton, especially If It goes wrong If thmgs go well it maybe 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 asnecessary and therefore asslgned elsewhere, the time alloted for theJob may not be longenough, the appearance of a problem may create a totally new sltuatlon A vlolatlon occurswhen a person deades, on the spot, to do It hs own way With all the time m the world onecan possibly go back and start agam, but somettmes a situation has started and can not beleft unfimshed The toothpaste won’t go back m the tube If there IS one thmg people feelthey are tired for It 1s to exercise their mmatlve and specialist knowledge and experience Amajor solution to vlolatlon IS the exercise of planrung, off the Job, and the recogmtlon ofplannmg 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 aJobThe Job IS always thought through m detailWe are always aware of all the problem which rmght anseJobs are planned together with supervisorsWork IS planned m an ad hoc manner bob by Job)

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

36


When work IS carried outThe tools and equipment are always available (Apeement]

Job plannq

When work IS carried outThmgs rarely go wrong [Agreement]We are left to fimsh the work we have started (Agreement]Plans are often changed before work IS stat-ted [Agreement]Work 1s often rescheduled at short notice [Ag eement]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 [Agreement]

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


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 theyperform What they thmk of as important, what they feel 1s not necessary etc If you feel thatsomethmg IS particularly Important, then you are more likely to try harder If you behevethat fallmg to attend can have serious consequences, and also beheve these consequences arequite hkelj to happen, then you are more likely to attend If, on the other hand, somethmg ISseen as tnvlal, unlikely to be dangerous or go wrong, then less attention ~11 be pald WhenIt 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 yearsTwenty years worth of experience 1s a strong personal counter to someone wth a hgher Jobgroup simply saying that your experience IS wrong Consrderatlon of negative consequenceshas to be welghed m with the expected probablhtles to arrive at personal Judgements of nskPoor attitudes m the workforce may be confUsed with different behefs m the set of possibleoutcomes

Vahes What IS seen and more or less Important or desirable Values are attached tooutcomes, actlons and behavlours

Attitudes can be either positive or negative The “don’t grve a damn” attitude contrasts withattitudes whch express “ttis IS really important” When someone decides to slup part of aprocedure, their attitudes form a major factor m decldmg what gets slupped If we knowwhat they consider important, and what they believe about outcomes, we can predict prettyaccurately whch steps m a procedure they will always adhere to, and which parts they ~11shP

TP

Knowledge of the potential negative consequences provides one of the inputs to theformatIon of attitudes, but these can be filtered by many factors, such as the feelmg ofpowefilness, which reduce a person’s assessment of the (known) probablhtles hsk IS seenas probability x consequence, once someone has decided that the bad consequences are notwhat ~111 happen to them personally, their attrtudes to their own and others’ behavlour maybe altered

Attitudes can be altered by outside mfluence If figures whom someone respects appears totake some thmgs seriously, the effects of norms may affect the construction of attitudes

38

Vlolahon Handbook Version 1 2 .

How do you define attitudes?

Agreement with the followmg Items

People cut corners In ON orgamsatlon becauseThey can get the Job done fasterThey can get the Job done more easilyThey can meet the performance targetsThe Job becomes more mterestmgiless bonng

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

Rating as likely or very likely the foliowmg

If I were to cut corners thenIt would save timeIt would make theJob easierIt would allow me to meet performance targetsIt would make the Job more interesting

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

and rating as desirable

Savmg time isMakmg theJob easier ISmeeting performance targets ismakmg the Job more interesting 1s

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

What are the components?

l Beliefs m the importance to the orgamsatlon of getting work done quickly so that thenext Job can get started The belief that rapid work, tigh production rates, IS work of highvalue and that speed itself 1s important

l Beliefs rn the rmportance to the mdlvldual of having an easy life, of not doing more than1s necessary (note the attitudinal component of what 1s to be defined as necessary)

l Beliefs m the importance to the mdlvldual and the orgarusatlon of meeting presetperformance targets T~.E may be seen to condone occasional corner cutting as thepenalties of not meeting targets (hgh probablhty unless comers are cut) are felt tooutweigh the nsks of something going wrong (low probabdlty given the mdlvldual’s owncompetence - see powerfulness)

l Behevmg that vanatlon in workmg methods enables the worker to keep doing the Job andreducing boredom TIE factor 1s probably the least important of the four components


39


3.1.6 Norms

what are Norms?

There are two types of norml Personal Norms define the lines past whch an mdlvldual will not go, come what mayl 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 beseen as followmg a different set of norms from the other group members This can lead tobehavlour which someone would regret if there were not a broader posltlve payoff m termsof acceptance by the group Bendmg or breakmg a rule, wth little chance of being caught orof thmgs going wrong, can be traded off against being allowed mto the group Group normsaren’t necessarily bad If the values are high people are as likely to raise their standards asthey 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 wouldhke to be” An Important factor 1s the feehng of ‘antlapated regret’ d somethmg should gowrong

How do you define norms?1 Personal Norms

Agreement with the followmg statements mdlcates hrgh Personal Norms

l In my opnuon a professional always follows the procedures [Agreement]l If damage should result from my failure to follow the rules, I would feel bad about It

afterwards [Agreement]l I’d find It hard to Just@ bending the rules to my colleagues [Agreement]l Personally I strongly disapprove when rules are bent [Agreement]l 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

[Agreement]l It 1s n-responsible to your colleagues not to follow -he rules [Agreement]l I disapprove when someone IS not followmg the rules [Agreement]l One should always follow the rules [Agreemenf]

40


2 Group Norms

Generally the group norms at different levels are measured by approval or disapproval of comercutting using approval of the followmg statement

When people cut comers m our orgarzlsatlon they expectl Their colleagues/peers to [APProveIl Their supervlsors to rAPProvel Their managers to rAPProve

Compliance by mdlvlduals IS mdlcated by agreement with the followmg statements

l If I were to cut comers thenl My supervrsor would disapprove [Agreement]l The company would disapprove [Agreement]l My work mates would disapprove [AFeement]l Others (techmclans) would disapprove [Agreement]Exactly whch groups’ norms are being complied wrth depends upon which group 1s feltwould disapprove

Agreement mth the followmg general statements mdlcates mdlvldual compliance to groupnorms

l I follow the rules, whether I agree with them or not [Agreement]l I only bend rules if others allow me to [Agreement]l You had better follow the procedures, It’s m your own interest [Agreement]l When somethmg goes wrong others wrll cover themselves by using the rules to blame me

[Agreement]l You better stick to the rules because others will cover themselves tf you don’t [Agreement]l If you don’t follow the rules you should expect to be putushed [Agreement]l People who disobey the rules are Just bemg awkward and makmg hfe hard for themselves

[Agreement]l I would not question the rules If they were wntten by experts [Agreement]l I am m favour of very strict enforcement of all laws no matter what the consequences

[Agreement]l An experienced worker 1s Justified m not domg the Job exactly to the book [Agreement]l A really slulled worker IS one who knows which rules to bend so as to get theJob done more

efficiently [Agreement]l People who rnmdlessly follow the rules are Just covermg themselves [Agreement]l I would not obey rules enforced by a person who does not deserve their posltlon of power

[Ageement]l I don’t obey rules unless I agree with them (Agreement]l If a manager asked me to do something whch I did not agree with, I would not do It

[Agreement]

41


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

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


l Anticipated Regretl Unthmkmg comphancel Pragmatic rule-bendingl Perceived legltlmacy see powefilness


42


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 themand how you do It When one or more of these IS constrained, you may feel that you havelost control over your own actions We have defined feehngs of control m terms of whetheryou feel that you have power over your actions, or that events have power over you, so thatyou have no choice

Powerlessness IS a very senous problem because it can short-circuit the best of mtentlons Ifyou feel that you really have no choice, It takes a very strong counter-force to overcome thednve to do what you know you shouldn’t

Powerfulness IS ofienJustlfied, but allows people to overlook what they know to beImportant problems on the grounds that “that won’t happen to me”

Both powefilness and powerlessness are mtenslfied by strong motlvatlon While someonewho IS little motivated to achieve an end will not be too affected by ablhty or constraint, ahighly motivated person will be very frustrated when they experience powerlessness, andwill be even more likely to see their own ablhtles as a Justlficatlon for doing what they areabout to do

How do you define it?

See PowerlessnessSee Powerfblness


PowerfulnessPowerlessness

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

43


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 Opportumtlesexist at a personal and at a Job level Opportumty appears to be a major factor mdetermmmg whether people will violate or not Are there many opporturutles or few? Dopeople 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 dosomethmg else, or are they about how careiil people were when lgnonng the opportumtlesthat arose’?

This posltlve aspect of opportumty should be dlstmgulshed Corn the luck of opportumty that1s better covered by powerlessness under loss of control Here the best and proper way 1sblocked - a lack of opportumty

People expect to find themselves cutting comers becauseThe procedures take too long to followThere are too many proceduresThe procedures are too comphcated

andThe procedures change too oftenThe procedures need to be updatedMany sltuatlons aren’t covered by procedcures

How do you define opportunity?

Sl tuational Opbortumty

The exrstence of sgnlficant Srtuatronal Opportunrty IS rndrcated by agreement wrth thefollowing statementsl 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


Personal 0pportunrQ

l SometImes there are sltuatlons m which I could get the Job done a lot faster [AgreementJl I’m able to take shortcuts to speed up the work [Agreement]l If I take shortcuts It would make hfe easier [Agreement]l In some sltuatlons I can do theJob better than the one In the procedure [Agreement]l 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 knowof them and, even, use themHowever, motivation plays an important role Unmotivated people are likely only to seeopportunmes for their own ease, the highly motivated, on the other hand, may see room forImprovements and grab them before It 1s too late


45


3.1.9 Powerfulness

What is Powerfulness?

Powerfulness 1s the feelmg that you can control events and outcomes You know what youare domg, you have been tramed for the task and are tughly experienced This may notalways be true, m whch case the feeling of control 1s illusory Illusory control 1s whatamateurs have, or young drivers when they have Just passed their drlvmg tests In the oil andgas industry powerf3ness reflects real ablhty, although feelmg powetil may suddenlybecome a problem when the situation 1s not exactly as usual PowerWness reflects provencompetence, but no one mdlvldual’s experience 1s ever gomg to be so wide that they willhave personally tested themselves agamst everything that could be requu-ed This means thatthe downside of competence 1s over-confidence m one’s own abilities - the feeling ofpoweri%lness

How do you define Powerfulness?

Agreement wrth the followvg statements

It 1s mamly up to me whether or not I cut comers

l It IS OK for somebody with my experience to cut comers [Agreement]l When you are slulled you know how to deviate from the rules safely [Agreement]l My experience tells me when bending the rules ~111 not be harti to me [Agreement]l You can bend the rules if you know what you are doing [Agreement]l I don’t follow rules if I know how to manage without them [Agreement]l My experience enables me to devlate from the procedures safely [Agreement]l If you are slulled you know when it IS OK to break or bend the rules [Agreement]l My professlonal knowledge tells me whether certain rules are appropriate or not

[Agreement]l People wth a lot of experience are expected to cover shortcommgs m procedures

[Agreement]l I don’t need rules to tell me what to do /Agreement]l I don’t follow rules that are time-consummg fAgreementJl I wouldn’t run mto problems even if I didn’t follow the rules /Ag cement]l Sometimes experience means that you know a better way of domg the job than the one

defined m the rules /AFeementJl I know when there IS a better way to do theJob /Agreement]l There are situations m which I use my own Judgement, whether that tnvolves followmg a

procedure or not fAgreementJl I know from experience when I should follow the procedures or not /Agreement]

4 6


What are the Components7l Trainmgl Experiencel Attitudesl Feehng of control


47


3.1.10 Powerlessness

What is Powerlessness?

Powerlessness IS the feehng you have when you no longer have control over events andpeople Thus feeling, m combmatlon wth a high degree of motlvatlon, can be extremelyfi-ustratmg A feeling of loss of control, powerlessness, can result m rule-breakmg despztethe best mtentlons and reasonable expectations Powerlessness 1s recogmseable when youwatch someone else doing aJob you know you can do as well (or better - e g powerfulness)and you know they aren’t domg It well, but you can’t do anything to do it better

How do you define Powerlessness?

Agreement with some or all of the followmg I

l Some procedures are so time-consummg they can’t be followed anyhow [Agreement]l Sltuatrons sometlmes anse which force you to break the rules [Agreement]l 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 Trammgl Attitudesl Motlvatlon


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 1sintended to help you select the appropnate remedies In this section a general overview 1sgiven of the possible approaches to remedlatlon Each one IS evaluated wth Its advantagesand disadvantages and a sample ratmg 1s given of each approach As such evaluations arecrude rt IS only a three-point scale The mtentlon IS at least to force people to thmk about therated charactenstlc and not to lmagme that every approach ~111 be equally effective, easy touse or will take effect nnrnedlately

Remedies are rated m terms of

Good Badl Ease of use Easy Medium D&ltl Counter-productwq No side-effects Some Side-effects Major side-effectsl Effectiveness Low effectiveness Effective figh effectivenessl Time scale to success Immediate impact Months More than a year

The mdlcatlons on these scales are given m bold typeface

There are a wide variety of possible ways to remedy vlolatlon or potential vlolatmgbehavlour Some of these are really effective, some only appear to be so Some may beeffective but have such bad wde-effects that they are best not pursued or only chosen andimplemented with condlferable care The dlstmctlon between recommended, effectiveremedies and the less effective 1s given m terms of Dlvlslon I and Dlvlslon II remedies Thesecond dlvlslon remedies are only Introduced and dtscussed here m order to show that thefi~ll range of posslblhtles has been considered and to prevent people from proposrng whatlooks attractive but doesn’t work

The possible approaches to remedlatlon are

Division Il Supervision - watch more closely over people, plan then- work, provide an examplel Analysis of existing violations - find the background behind vlolatlon behavlourl Analysis of violation potential - use the theory to see If people will find vlolatmg easyl Structured discussions - discover what IS going on and get people to agree on actionsl Procedures - improve the quality of procedures so olatlon IS not seen as necessaryl Direct management - a techmque to vary the procedures as cn-cumstances demand

Division IIl Selection - pick people who don’t vlolate to start withl Tratning - tram people more on the real risks and the possible consequencesl Detection - sharpen the lookout for vlolatronsl Reporfing - encourage people to report when they or others bend the rulesl Incentives - pay for compliance or pumsh people who break the rules

49


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 getout of hand Supervisors can actively search for problems that lead to vlolatlon, suchas poor procedures and inadequate planrung, they can iden@ who are sheep and whoare wolves and allocate them appropnately They are the people who should performor check the plannmg process and have the experience to recogruse opportuxutresearlier Supervisors should not. however, merely act as policemen. watclung otherswork m the hope of detecting vlolatlons

Advantages

SupervIsors are close to the action If there IS any rule-breakmg going on, they shouldbe aware of It They are also m a position to do somethmg about the work plannmgand 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 “Don’t you do It this way unttl you are as experienced asme” The problems with work plannmg may be being caused by supemsors (thesolutron 1s to tram them to plan better, see trammg) Also too close personalsupervision may cause resentment, m competent workers, at being watched and feltnot to be capable of doing the work on therr own

l Ease of use Easyl Counter-productlmty No side-effectsl Effectiveness Low effectivenessl Time scale to success Immediate impact

Mediumside-effectseffectivemonths

DifficultMajor side-effectsI-hgh effectivenessMore than a year

50


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 nothidden 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, sheep’sclothmg If there are no vlolattons to analyse, the lmpresslon can easily be gamed thatthere IS nothing to worry about

l Ease of use Easyl Counter-productlvlty No side-effectsl Effectiveness Low effectivenessl Time scale to success Immediate impact


DlficultMajor side-effectsHigh effectivenessMore than a year

51


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 ofoperators, supervisors and managers with a view to uncovering the grounds forvlolatlon Look for whether opportumttes are to be found, and people praised andrewarded for takmg them (successfilly) Examme whether the quality of the plannmgprocess IS hable to let people down Just at the moment that they feel they need tocontmue m order to get theJob done Analyse the current expectations and find outwhy they arlse Use the dlagnosls sections of this handbook to ascertam m whch areasthere are likely to be problems Use the remedy sectlons to select the appropnateremedy 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 hasrequired an mvestlgatton of why a vlolatlon took place

Drsadvantages

Requires slttmg down and bemg proactlve when there appears to be no reasonJustlfymg proactlve work IS always hard as “nothmg IS wrong” Requires people to usethe mformatlon here systematlcally, when they may take one look at the names of thefactors and unmed~ately draw their own conclusions without having the drsclphne toactually follow through

l Ease of use Easyl Counter-productrvrty No side-effectsl EfSectrveness Low effectivenessl Tune scale IO success lmmedlate Impact


DlfiicultMajor side-effectsHigh effectivenessMore than a year

52


3.2.1.4 4 Structured Discussions

One major method that can be used 1s a structured dlscusslon session with theworkforce In small umts or at the managenaUsupervlsor levels, all mvolved can andshould be present In larger groups a selection has to be made, but with more than onesession more people can be directly and personally mvolved

Unless there 1s a specific structuring rule a dlscusslon session IS hkely to turn mto abitch-session, with the air cleared but no progress Thus IsJust the kmd of lack ofprogress that creates negative expectations The structured approach IS related toTnpod DELTA generation and dlagnosls sessions The task IS to discover the reasonsfor failure, to see If there are specific underlying causes that can be ldentlfied fortreatment (cure the disease, not Just the symptoms), then to find remedial actions andcarry them out

Method

z Select a dlscusslon group=> Start with a general dlscusslon of vlolatlonq Use the techmque of askmg each mdlvldual to describe a vlolatlon they personallyperform (don’t 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 morehonest)=> Agree that vlolatlons are present and need to be fixed3 Go round the group m search of routme vlolatlons Collect at least one from eachmember Group leader goes first to break Ice If necessary3 DISCUSS the reasons for these routme vlolatlons Use the Diagnosis Sections to help3 Go round the group m search of sltuatlonal vlolatlons Collect at least one from .each member If at all possible3 Inventonse the sltuatlonss 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 wthremedies 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 mtounproductive argument, dommated by people who want to get their personal hobbyhorse fixed and run by managers who want to blame everyone but refuse to accepttheir own role m the sltuatlon

53


l Ease of use Easyl Counter-productwty No side-effectsl Effectiveness Low effectwenessl Time scale to success Imrnedlate Impact

Mediumside-effectseffectwemonths

DifficultMajor side-effectsHigh effectwenessMore than a year

54

Vmlauon Handbook Version 1 2

3.2.1.5 5. Procedures

Vlolatlon IS all about deliberately not followmg procedures The solution IS to removeunnecessary procedures, improve those that are found to Induce vlolatlon because theyare too long, dlffkult or n-relevant Routme vlolatlons are almost mvanably caused byIll-thought out and incorrect procedures, so sensible behavlour becomes vlolatlon and,as a result, vlolatlon becomes a habit Sltuatlonal vlolatlons are caused by havmgprocedures that have not been well enough thought through for all the sltuatlons mwhich they should apply Improvmg those procedures can only be beneficial

Advantages

Getting the procedures nght removes all the routme vlolatlons and should remove allthe sItuatIona ones as well With a considerably reduced frequency of vlolatlon theculture will shfi of Its own accord At this point any vlolatlon should be much moreobvious and, If felt to be necessary, the direct management approach (See 11) with aprocedure for vanatlon due to circumstances will be easier to operate Getting theprocedures nght ~11 almost certainly have considerable benefits m terms of operatingefficiency

Disadvantages

Re-exammmg and rewntmg procedures IS a lengthy and potentially expensive processThe main reasons why procedures have been changed m the past IS the result of asingle incident (Shuttmg the stable door after the horse has bolted), but the effect IS tooverlay a procedure with controls for the highly Improbable and to lose all sense ofwhat 1s really vital Because of the mdlvldual mcrdents it can be hard to argue againstremoving what seems a sensible precaution, the overall view IS hard to mamtam whenm the middle of reviewing an old but suspect procedure

l Ease of use Easyl Counter -productrwty No side-effectsl Effectrveness Low effectivenessl ike scale to success Immediate Impact


DtfficultMajor side-effectsHigh effectivenessMore 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 yourworkforce may be Wolves, born violators, then one solution to ensuring safety andeffectiveness IS actrve and direct management of vlolatlon behavlour I e Allowvlolatlon by creating a procedure that makes vanatlon pemssable under strictlydefined precondltlons

2 Ensure all commumcatlon devices are avallable and operatmg (telephones, hand-held radios, line of sight)= Get people to recogtuse when they are vlolatmg=> Recogmse a vlolatlon-provokmg situation3 Apply Rule of Three to identify extra sources of errorz Inform all tn vlcunty that rules are bemg broken, put up flags, flashmg hghts etc= Note the reasons why vlolatlon appeared necessarys Routme - IdentlfL and consider changing the rulea SItuatIonal - identify and remedy sltuatlons=> Optnmsmg - consider for Improvement of procedures*FIX necessity for next time and before next tzme3 Hangmg offence IS NOT followmg ths procedure when vlolatlon occurs

Advantages

Takmg this approach allows work to continue, opporturuttes to be taken andimprovements to be Implemented

Dzsadvantagesl

This IS such a drastic approach It can really only apply once most of the reasons forvlolatlon have already been fixed Because it generates a fi.111 ‘paruc’ mode people maybe tempted to sktp It (“Just ths once”), which IS why failure to use ths procedure hasto be a full hanging offence

t

l Ease of use Easyl Counter-productlwty No side-effectsl Effectiveness Low effectivenessl Time scale to success Immediate impact


DifficultMaJor side-effectsHigh effectwenessMore than a year

5 6

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 notviolate to start with Selection would mvolve creating a comphant workforce that didnot 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 tasksthat real& require sheep rn sheep’s clothmg To the extent that these mdlvlduals canbe 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 beautomated (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 becovered by people taking direct mltlatlve As plannmg failures form a much morepotent cause of vlolatlon than simple failures of intention, selecting a totally obedientworkforce may prove costly and even dangerous In the US nuclear mdustry it wasfound that the maJorlty of procedure causes tn incidents were caused by peoplefollowmg the procedures, not by deliberate farlures to obey1 So, tf you want to go thsway. 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, ratherthan those Just good enough for competent workers to do the nght thmg

l Ease of use Easyl Counter-productlvlty No side-effectsl Eflectrveness Low effectivenessl Tzme scale to success Immediate Impact

Medmmside-effectseffectivemonths

DifficultMajor side-effectsfigh effectivenessMore than a year

5 7


2.3.3.2 8. Training

Trammg can mvolve safety-specific trainmg or more general techmcal or professionaltrammg Trammg may be class-room or on-the-Job and ~111 spread across a person’scareer The expectation IS that people can be persuaded not to violate The reahtylooks more complex Trammg can mform people about potential consequences ofactlons, help them to thmk complex sltuatlons through The down side IS that trammgincreases peoples’ feeling of competence and therefore their confidence m theu- ownactions (powerfulness) When someone deviates from a procedure, clamung that theywrote the procedure m the first place, you have an example of both why they werehired and what damage they might do

AdvantagesF

Trammg operates to mfluence attitudes by provldmg people with mformatlon aboutconsequences and sendrng messages about acceptable norms The lessons learned canbe 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 oneof the contributing factors to vtolatlon Competence 1s seen as an mvltatlon toexcerclse ttutlatrve and that 1s what people are bred for

l Ease of use Easy Medium Difficultl Counter-productlvlty No ade-effects side-effects Major side-effectsl Effectiveness Low effectiveness effective I+gh effectivenessl Time scale to success Immediate impact months More than a year

Vlolahon Handbook Version 1 2

2.3.3.3 9. Detection

Detectmg vlolatlons can be done after the event, usually m an accident report, or bydirect supervision, as on walk-rounds

AdvantagesA vlolatlon that has been detected can be analysed (see remedy 2) Really dangerousmdlvlduals, those with poor Intentions, can be detected and dealt with Other peoplecan help to uncover the grounds for their vlolatlon The mam reason for usingdetection appears to be as a preventative measure, I e supervisors are there to be ableto detect violation (See remedy 1)

DrsadvantagesDoes 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 bevery ineffective and probably much too late For self-detection see remedy 10 Whenusing supervlsors as detectors It IS all too often not clear what will happen when theydo find them This problem IS often exacerbated by the knowledge that it IS sometimesthe supervisor who IS the pnme violator (see 1)

l Ease of use Easyl Counter-productrwty No side-effectsl Effectrveness Low effectivenessl Time scale to success Immediate impact


DifficultMajor side-effectstigh effectivenessMore than a year

59


2.3.3.4 10. Reporting

Reporting IS a form of self-detection that 1s like a near-nuss system Such systems aremdlcatlve but haphazard Reportmg on others can be dangerous m most culturesWhile reporting should be encouraged, this will only occur if there IS no blame cultureIn the aviation industry there are confidential reporting systems, such as CHIRP, buteven here the legal consequences mean that not all such mformatlon can be trulyconfidential

Advantages

Makes people aware of what they are supposed to report Similar to Unsafe ActAuditing

Disadvantages

Requires a very open company culture, which IS rare and difficult to put mto placefrom outside among some contractors Unsystematic as It requires mdlvlduals to havea sufficiently deep understandmg to recogmse what It 1s they have to report Could benusused to get other results, such as cntlclsmg supervisors and managers for forcingviolations

t-’

i

l Ease of use Easyl Counter-productwrty No side-effectsl Effectiveness Low effectivenessl Tzme scale to success Inunedlate Impact


DifkultMajor side-effectsfigh effectrvenessMore than a year

60

Vlolabon Handbook Version 1 2

2.3.3.5 11. Incentives

Paymg people to behave properly, or at least finmg them if they break the rules, 1s aclassIcal techmque Smularly within orgamsatlons careers can be influenced by theextent to whch people do or do not comply As a techmque rt seems simple andeffective The problem with mcentlves IS that they are applied by people who thmk thatthey ~111 work for them, but they don’t break the rules or the system doesn’t apply,wtule they are untended for other people (e g , Prison IS felt to be an effective threatfor the law-abldmg, who assume it ~111 be equally effective as a deterrent for thecnmmal The size of the prison population seems to disprove the Idea) The use ofmcentlves m a simple form Implies that mtentlon IS the determmmg factor, asmcentlves will work on the mtentlons As the studies have shown that wthm Shellcompanies offshore mtentlon IS already a solved problem, and the causes for vlolatlonhe elsewhere, mcentlves are much less likely to be effective than naive outsiders wouldimagine

A more subtle approach mvolves maupulatmg incentives for compliance, as suggestedm the ongmal Bendmg the Rules report (Reason et al, 1994) Here again the problemIS that compliance appears to be of little effect m determmmg whether someone ~11vlolate, as the Important factors (Plannmg, Opportumty, Power-filness andExpectation) 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 notvery sure process

Advantages

Appears to be what 1s needed Acceptable to outside bodies as an mdtcatlon of takmgthmgs seriously

Disadvantages

Not very effective and potentially counter-productive, leadmg to mls-reporting morder to obtain the benefits Thus will make thmgs worse rather than better as whathttle mformatlon there IS may be suppressed to obtam the short term rewards

l Ease of use Easyl Counter-productmty No ade-effectsl Effectiveness Low effectrvenessl Tune scale to success Immediate Impact


DifficultMajor side-effectstigh effectivenessMore than a year


3.3 Remedies and Problem Areas

All of the remedies described above can be applied Some are less effective than othersTrammg and SelectIon have probably little margmal benefit, what advantage there was hasalready been obtained Incentives are superficially effective. but suspect SupervIsIon,DetectIon and Reporting should be employed, but with recogmtlon of the disadvantagesThe three analysis techmques, Direct Analysis, ProactIve Analysis and StructuredDlscusslons should be apphed to see exactly what problems there are Finally lmplementlngImprovements m Procedures should have a maJor effect Once these have been started andshown to have effects, the Direct Management approach can be star-ted

T

The next sections define briefly the appropnate remedies for problems Identified The mostusehI remedies are tighllghed m bold

t-

3.3.1 Violating Behaviour

Tell them not to This IS not very effective on its own Try gomg back toward the rootcauses The message management has to send 1s that wtule mttlatlve may be desired,vlolatlon 1s not A vlolatlon management approach mvolves

1

23

Recogmsmg that someone (oneself) 1s about to break or bend a rule This becomes easierwhen vlolatlon 1s no longer routine and the rules are well known and not impossibleCallmg a time-out to signal a situationDlstmgulsh between sltuatlons offenng no choice, an obvious improvement or a newproblem

45

67

Either create enough time and resources locally so as not to violate or,Flag a variance and mform everyone what IS gomg on The most effective defence agamstan ongoing vlolatlon IS the knowledge that it IS happenmg, who, where and when Radios,special signals etc all sewe thus purpose See Direct ManagementReduce the chances of errors ocumng elsewhere that have any chance of mteractmgRecogmse 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 andcertainly defined as below the ALARP threshold In the SMS

a SelectIonl Trammgl Supervisionl Detectionl Reportingl Analysis of existing violationsl Analysis of violation potentiall Structured dlscussionsl Proceduresl Incentivesl Direct management

62


3.3.2 Intention

The problem with intentions lies m one or more of the three components1 Attitudes2 Norms3 Feeling of control

1 Attitudes - These are defined by the the thmgs people find important and what they feelare 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 theirmates/colleagues would have them behave m a certain way7 Does a person have a personalhnut beyond which they are not prepared to go7

3 Feelmg of Control - Poor mtentlons anse when people feel that they lack control overtheir actlons Why 1s ths so7 Where does this feelmg anse7

The primary solutions are selectlon, trammg and perceived rewards1 Select people with value systems that are right to start with - don’t expect to take peopleand then try to make them better overnight Unless there 1s no choice choose well-mtentloned people This selects on Norms and Attitudes2 Trammq tells people about what IS important and predonunantly Influences attitudes bymformmg people about the consequences when thmgs go wrong3 Rewards shape the way people want to behave Those who wish to get on , to start withIf all else falls Fn-e them

l Selectionl Trainingl Supervisionl DetectIonl Reportingl Analysis of exlstmg violationsl Analysis of vlolatlon potentiall Structured dIscussIonsl Proceduresl Incentivesl Direct management

63

Vlolauon Handbook, Version I Z

3.3.3 Expectation

Expectatron has been found to be the smgle largest factor determmmg whether peopleviolate or not We expect to find ourselves vtolatmg because the srtuatron m the past seemedto requn-e It and now because nothmg has changed The old srtuatron still exrsts and thesame pressures will arise again Unless somethmg IS changed and seen to have been changedfol the better, expectatrons wtll remam as they are and vrolatrons wtll remam likely Amanager or supervisor faced with an expectatron problem can expect an uphrll strugglebecause they are mstrumental m creatmg the condrtrons for vtolattonl Tlus IS whatcommrtment m Enhanced Safety Management IS all about

Expectatron 1s made up of a number of different components Each of these requires adifferent solutton Two approaches can be taken

1 Analyse the srtuatron on your own Draw up a one year plan and Implement tt The bestway to do thrs IS to have the plan as one of your personal tasks and set measureable targetsfor success Make sure that the targets are feasible and that the changes are vrstble to thoseliable to have the poor expectatrons, otherwise they won’t believe anythmg has been doneand 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 knowwhat they are, the problem IS we fall to understand Just how tmportant they are), some willtake time because fixmg expectation requires both fixmg the root cause and then havingthose affected change then behavtour as a result of dtscovenng that real change has takenplace For instance, haste IS a major problem The solutron for haste lies m preparationplanrunq and manpower allocatton Until a full cycle of work has been gone through peoplemay not believe that real change has taken place With a 12 month productron- maintenancecycle 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 dragnosrsof the underlymg and nnmedrate problems Start with having each mdrvrdual put forwardan example of a bent or broken rule, a procedure that they, personally, have knowmglyfailed to follow fully Once the group has a full hst,start to search systematrcally for thereasons why the rules were bemg broken Use the questrons m the defimtron ofexpectation and then work backwards through the different components as a way ofstructurmg the dragnosrs drscussron

+ Go through the exercise m Part IV to tdenttfjr routme rule-breaking and srtuatlons whrchseem to require vrolatron for success Revrsrt the procedures

+ Discover the Personal Norms

64

Violation Handbook

+ Check on IntentIons to make sure there Isn’t an Intention problem-

* SelectIonl Trammgl SupervisIonl DetectIonl Reportmgl Analysis of existing violationsl Analysis of violation potentiall Structured discussionsl Proceduresl Incentwesl Direct management

Version 1 2

6 5


3.3.4 Planning

Fnung poor plannmg requires first ldent@ng the problem m more detail Use thebreakdown

Preparatron planning - The lmtlal defirutlon of the work to be doneJob Planmng - The local details of how the Job IS to be done, by whom etcActlon 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@ whether1 The work actually needed to be done2 The original planning was camed out with hands-on knowledge3 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’

1

r-

Was plannmg ad hoc, opporturustlc, top-down or hnear’ What tt.lls means 1s Is the plannmgprocess Itself planned? Ad hoc and hnear plannmg processes are to be avolded Top-downplannmg needs to be tempered by reality

+ Ad hoc planrungPlannmg takes place on a moment-to-moment basis The work plan 1s very rough andready and there 1s conslderable rehance on experience

+ Oppofturustlc planmngPlans are easily set aside when opportumtles are discovered Not necessarily bad, butneeds to be controlled

+ Top-down plannmgPlanning starts with the most general activity and IS broken down m increasing detailProblems arlse when the level of detail IS either too little or, sometlmes, too much Planshave to be flexible enough to cope with unexpected situations and top-down planmng 1sngld

+ Linear planningPlannmg starts at the beginning and goes on till the end This approach 1s liable to get intotemble problems If It IS found it must be replaced

l Selectionl Trammgl Supervisionl Detectionl Reportingl Analysis of existing violations

66

Uolauon Handbook Version 1 2

l Analysis of vlolatlon potentiall Structured dlscusslonsl Proceduresl Incentwesl Direct management

67

Version 1 2Uolauon Handbook

3.3.5 Attitudes

Flxmg poor attitudes requires bnngmg people to understand what IS really important and tohave to correct emotional response to situations

1 What is really important Ttus 1s trammg, especially much safety-related trammg, whchconcentrates upon the consequences First people have to know what those consequencesare

.

2 How should you feel. Thus IS much harder to fix Our emotional responses are theproduct of our genes and our early environment What you can do IS to try and apply thedesired emotional response to specific sltuattons byframrng situations to appear Forinstance, situations can be described m positive or negative terms A simple medical analogyIS found In an expenmental task where people must choose treatments If they are framed interms of 65% survive then they make quite different choices than when the task IS framedwith 35% die Logically there 1s no difference, m practice people are really affected by theuse of the word survive vs die

SelectIonTrainingSupervisionDetectionReportingAnalysis of existing violationsAnalysis of violation potentialStructured discussionsProceduresIncentivesDirect management

6 8

Version 1 2Uolauon Handbook

3.3.6 Noms

There are two types of Norm, Group and PersonalGroup norms require careful selectlon If the workforce operates with norms that are dlstrnctfrom those required for safe and comphant behavlour, one should consider breakmg groupsup and selecting figures For Instance people are more likely to change their dnvmgbehavlour for a well-known racmg dnver (Damon Hill or Michael Schumaker) than for ananonymous pohcemanPersonal norms are one of the most Important defences against vlolatlon These will havebeen set early on Younger people may still be In a more labile stage, whch means that theypick up and retam the values by reference to status figures Umversltles and professionalorgarusatlons represent slull norms Engmeers for instance, should have lnmts past whichthey will not knowmgly go. defined by “Good Engmeenng Practice”

Do not expect supervisors to set group norms They can, however, set personal ones

SelectionTrainingSupervisionDetectionReportmgAnalysis of exlstmg vlolatlonsAnalysis of violation potentialStructured discussionsProceduresIncentivesDirect management

69


3.3.7 Feeling of Control

Powerlessness

Powerlessness means that people feel they have no control over then- actlons, the externalenvironment IS doing the dnvmg The fixes ~11 either mvolve altenng the setting so thatalternatives are obvious, or changing the person by letting them take control when thesituation arises 1

u-’l Create situations in which the feeling doesn’t have to mse by plannmg m advance At the

workface level techruques like JSA (Job Safety Analysis) can help by lookmg at what has1

to be done Just before doing it rl FIX the situations so that altematlves are always avalable Thus IS a task for supervisors

and planners Change the procedures appropnatelyl 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, thusmay be a time to involve the Direct Management approach

l Report situations where lack of choice becomes a problem If supervisors aren’t told,they may never know any other way

l Have a procedure for stopping or tellmg others, 1 e the Direct Management techmqueneeds 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 theyknow they are, the feeling of powefilness and control IS itself potentially dangerous

l Learn to recogmse one’s shortcommgs and recogmse the feeling that one IS takmgcontrol when it may be unwarrented

l Learn to use the Direct Management techmquel 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 cutsl Tram people by putting them In situations where they create their own problems and learn

to understand that powetilness has a downside

SelectionTrainingSupervisionDetectionReportingAnalysis of existing violationsAnalysis of violation potentialStructured discussionsProcedures

1

h‘

70

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 Do employees know and understand the procedures? l Do we need all of these procedures? l Are there sltuatlons when it IS Impossible to apply procedures7 l Does the job itself encourage vlolatlons7 l Is It possible to have a procedure for every sltuatlonv l 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 1s not 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 IS presented to allow the orgarusatlon to assess the 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


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 1s obviously 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 1s not the rules per se that are at fault, but the distnbutlon and trammg with respect to the rules Developing and pubhshmg good rules 1s essential, 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 non- comphance, 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 1s encouraged. 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 Vernon 1 2

4.3 Routine violations

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 1s caster 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, IS to 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 the prevalence of routine violations

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 Infringements of some rules 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 or no risk 16 I can get the Job done quicker by lgnonng 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 bhnd eye to some rule violations 20 Some rules are only of value to protect management’s 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


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 1s easy 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 do’s and do not’s, 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

2 I Schedules seldom allow enough time to do the lob accordmg to the rules 22 I sometimes can’t get the eqmpment necessary to work to the rules 23 Sometimes condmons at the workplace stop me working to the rules 24 There are mcentives to ignore some rules 25 There are circumstances where managers will support rules being broken 26 StafF shortages sometimes result m rules being broken to get the job done 27 Some rules are impossible or extremely dn?icult to apply 28 Supervisors recognise that m some snuations deviations from rules are unavoidable 29 It IS necessary to bend some rules to achieve a target 30 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 IS because 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 IS worth 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 1s to 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~hty m a system where actlon’state rules predonunate, means that the focus 1s on compliance rather safe behaviour

If the rules do not work wlthm a particular sltuatlon. then people ~111 violate m order to do 4

the job There may an unw&ngness to find alternatIve ways of domg the job safely, probably r

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

78


4.5 Optimising violations

Optm-nsmg vroiattons 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 1s concerned 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 1s drffrcuit It 1s worth 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

3 1 Sometrmes the operatmg hmrts prescrrbed m rules are too restnctrve 32 I sometrmes don’t know why I have to follow rules 33 I am sometrmes tempted to do work that IS not my responsrbrlrty 34 There are financial rewards to be gamed from breakmg the rules 35 I am not given regular break per-rods when I do repetitive and bonnglobs 36 Workmg to the rules removes slulis 37 Devtatmg from some rules demonstrates knowledge of the job 38 There are no personal benefits from strictly foiiowmg rules and procedures 39 Fmdmg aitematrve ways of doing the job makes tt more mterestmg 40 You can’t 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 IS not 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 1s to mvolve the staff m wrttmg and evaluating procedures Tlus 1s partrcuiariy 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 1s a hrgh level of

79


comtment 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 1s to 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 1s taken Corn Reason et al (1994)

“A pair of engineers are mspectmg an 011 pipeline 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 1s attempting to solve a difficult and unusual problem for whch there 1s no 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 1s m 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 1s more 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 I ofien come across situations with which I am unfanuhar 42 I have not been tramed m the rules to be used IP unusual cn-cumstances 43 I ofien encounter situations where no prescnbec rules are available 44 There are some rules where your natural reaction would be to break them 45 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 Vernon 1 2

4.7 Summary of Reactive Approach

The five main steps m lmprovmg safety through procedural comphance are as follows Each step 1s hnked 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

1s to 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 BuZZetrn of the Brltzsh PsychoZogzcal Socrey, 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~t Leaden 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- 18 I

85

Uolauon Handbook Vernon 1 2

6.0 Appendix Reactive Test Items

1 = strongly disagree 2 = disagree 3 = neither agree nor disagree 4 = agree 5 = 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

I

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

8 I sometimes come across a rule I did not know about 1

9 I sometimes fall to fiAly understand which rules to apply 1

10 I have rules for tasks I will never have to do 1

11 The rules do not always describe the best way of working 1

12 There are some rules that make the job less safe/efficient I

13 Infimgements of some rules occur all the time 1

14 I have found better ways of doing my job than those 1 given m the rules

15 Short cuts are acceptable when they mvolve little I

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

86

Vlolauon Handbooh Version 1 2

or no risk

16 I can get the job done quicker by lgnormg some rules I 2 3 4 5

17 Some rules are only for mexpenenced workers 1 2 3 4 5

18 Some rules do not need to be followed to get the job done safely

I 2 3 4 5

19 SupervIsors turn a blind eye to some rule violations 1 2 3 4 5

20 Some rules are only of value to protect management’s back 1 2 3 4 5

21 Schedules seldom allow enough time to do theJob I 2 3 4 5 accordmg to the rules

22 I sometImes can’t get the equipment necessav to 1 2 3 4 5 work to the rules

23 SometImes conditions at the workplace stop me I 2 3 4 5 workmg to the rules

24 There are mcentlves to Ignore some rules 1 2 3 4 5

25 There are circumstances where managers wtll support I 2 3 4 5 rules bemg broken

26 Staff shortages sometImes result m rules being broken 1 2 3 4 5 to get the Job done

27 Some rules are lmposslble or extremely difficult to apply 1 2 3 4 5

28 Supemsors recogruse that m some situations 1 2 3 4 5 devlatlons from rules are unavoidable

29 It 1s necessav to bend some rules to achieve a target 1 2 3 4 5

30 At certain times rules have to be bent to keep the Job gomg 1 2 3 4 5

3 1 SometImes the operating hmlts prescribed m rules 1 2 3 4 5 are too restnctlve

32 I sometImes don’t know why I have to follow rules 1 2 3 4 5

33 I am sometImes tempted to do work that IS not my 1 2 3 4 5

87

I

Vlolahon Handbook

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 can’t 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

1

1

1

1

I

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

3

3

Version 1 2

4

4

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

5

5

5

5

88

Bending the Rules II

Documents

Transcript of Bending the Rules II