Why it was so difficult to develop new methods to inspect work organization and psychosocial risks in Sweden

Abstract

In 2001–2003, the Swedish Work Environment Authority (SWEA) ran a project to develop better methods to inspect psychosocial risk factors at work. The objective was twofold: to develop methods to enable most inspectors to effectively inspect such health risks, and to set a standard for method development within SWEA. This article presents our evaluation of the project and a discussion of this as an example of regulatory implementation. The methods project largely failed. Major reasons were the lack of general provisions on psychosocial risks, isolation from other policies in SWEA that affect the inspection of such risks, and a lack of engagement and guidance by top management on how to prioritize and conduct this very challenging development project. Underlying this was possibly a preoccupation with other major internal reforms, a limited competence and an unwillingness to challenge the employers on psychosocial and organizational issues within SWEA’s top management. Yet, the project probably had some indirect positive effects by raising awareness within the authority of psychosocial risk factors, of the complexities of inspecting such risks and of the difficulty to develop effective methods to supervise them.

Full text

Why it was so difficult to develop new methods to inspect work organization
and psychosocial risks in Sweden
Anders Bruhn
a,*
, Kaj Frick
b
a
Örebro University, School of Humanities, Education and Social Sciences (HumES), 70182 Örebro, Sweden
b
University of Mälardalen, School of Sustainable Development of Society and Technology (HST), Sweden
article info
Article history:
Available online 21 August 2010
Keywords:
Labour inspection
Psychosocial risk factors
Inspection methods
Work environment policy
Occupational health and safety
management
abstract
In 2001–2003, the Swedish Work Environment Authority (SWEA) ran a project to develop better methods
to inspect psychosocial risk factors at work. The objective was twofold: to develop methods to enable
most inspectors to effectively inspect such health risks, and to set a standard for method development
within SWEA. This article presents our evaluation of the project and a discussion of this as an example
of regulatory implementation. The methods project largely failed. Major reasons were the lack of general
provisions on psychosocial risks, isolation from other policies in SWEA that affect the inspection of such
risks, and a lack of engagement and guidance by top management on how to prioritize and conduct this
very challenging development project. Underlying this was possibly a preoccupation with other major
internal reforms, a limited competence and an unwillingness to challenge the employers on psychosocial
and organizational issues within SWEA’s top management. Yet, the project probably had some indirect
positive effects by raising awareness within the authority of psychosocial risk factors, of the complexities
of inspecting such risks and of the difficulty to develop effective methods to supervise them.
Ó2010 Elsevier Ltd. All rights reserved.
1. Introduction: Sweden had to reduce psychosocial risks at
work
This article is about a special project to develop new methods to
inspect psychosocial risks at work in Sweden. We, the authors of
this article, were commissioned to evaluate this project (Frick et
al., 2006). The objective here is to describe and discuss the devel-
opment process, its results and our main evaluative conclusions
of the methods project. The central analytical question is how
and under what circumstances effective inspection methods on
psychosocial risks can be developed, as seen in the Swedish con-
text. We will also place this question in the wider setting of regu-
lation and supervision of the psychosocial work environment, in
order to better understand the difficulties of the project and there-
by contribute somewhat to a more general understanding of this
process of regulatory implementation.
The psychosocial aspects of the work environment have now
been regulated for a decade and a half, at least within the EU
(through the Framework Directive, 89/391/EEC). Ill-health due to
such risks is widespread and increasing. According to the Fourth
European Working Conditions Survey 2005, 22.3% of the working
population in the EU 27 countries see themselves confronted with
work-related stress. Thus, with backache (24.7%), muscular pain
(22.8%) and fatigue, stress is one of the four major work-related
health effects (European Foundation, 2007).
1
There are also many
attempts to reduce such risks and improve the psychosocial work
environment. Some major examples are the European social part-
ners’ agreement against stress at work (European Social Partners,
2008) and the Health and Safety Executive’s stress management
standards in the UK (HSE, 2004).
The situation in Sweden is no different. The incidence of work-
place accidents and ill-health caused by various kinds of technical
risks in the work environment steadily decreased during the dec-
ades before the millennium shift. In spite of this, however, sickness
absenteeism increased heavily. Between 1996 and 2002 – the
strong growth after the economic crisis in the early 1990s – the so-
cial insurance costs for sick-leave pay and early retirements dou-
bled (GNP from 2% to 4%). Even though no simple and clear
causes for this can be discerned, stress and other psychosocial risks
at work are believed to be important contributing factors (Ds,
2001:28; Frick, 2004). These risks were incorporated into the
broadened concept of the work environment, legislated already
in 1977 (the Work Environment Act). However, before the end of
the 1990s, not much had been done in Sweden to develop methods
on how to inspect such psychosocial risks at work.
0925-7535/$ - see front matter Ó2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ssci.2010.07.011
*Corresponding author. Tel.: +46 19303428.
E-mail addresses: anders.bruhn@oru.se (A. Bruhn), kaj.frick@mdh.se (K. Frick).
1
It should be noted that stress also strongly affects musculoskeletal health
(Lundberg and Melin, 2002).
Safety Science 49 (2011) 575–581
Contents lists available at ScienceDirect
Safety Science
journal homepage: www.elsevier.com/locate/ssci

In 2001, the Swedish Work Environment Authority (SWEA, be-
low including its predecessor until 2001, the National Board for
Occupational Health and Safety, ASS) was allocated more re-
sources, in part as a reaction to the alarming costs for ill-health.
To use the new resources efficiently, and as an answer to govern-
ment demands for better supervision of psychosocial risks (Ds,
2001:28), SWEA organized a development programme – ARNE,
2
between January 2001 and December 2003.
3
It consisted of four
parts: training of inspectors, recruitment of new ones with more
competence to supervise psychosocial risks; the development of
new inspection methods; and finally a review of available re-
search/information on these issues, thus the two squares in the mid-
dle of the model below. As mentioned above we were commissioned
to evaluate the third part, the methods development project (repre-
sented by the fourth square in the model).
The process of regulatory implementation – the part for which
SWEA is responsible – can be described as a supervision chain. The
point of departure is state objectives and how these are expressed
in the Work Environment Act. SWEA has a mandate to develop the
objectives into mandatory through provisions, to have competent
staff with adequate methods control that these regulations are
complied with, to require improvements where the inspectors find
non-compliance and to enforce the requirements, where so
needed. What gets changed at work depends much on how the
separate links in the supervision chain connect to and support each
other – on SWEAs ability to combine them into a well-integrated
supervision strategy. To evaluate the MDP and its results one has
to understand this context. Both the wider context – that is objec-
tives and goals at the political level (input) and the situation at
work (output) – and the narrow one, that is how the inspection
methods are to act as a part of SWEA’s supervision strategy. In this
article, we will at first deal with the wider context, the history,
background and wider conditions behind the ARNE-programme
and its method project (Section 2). Sections 2.3 and 2.4 also
touches upon the narrow context, the development of the regula-
tory part of the supervision strategy. Sections 3–5 concerns our
main focus, the development of the method project and its results.
Section 6contains our conclusions of the project and some com-
ments on how these exemplify regulatory implementation pro-
cesses at a general level.
2. The ARNE-programme
2.1. Work organization and psychosocial risks in Swedish work
environment supervision
Workplace psychosocial risk factors were, as mentioned, incor-
porated into the new and broadened Work Environment Act of
1978 (SFS 1977:1160. Its section 2:1 requires that:
Technology, work organisation and job content shall be
designed in such a way that the employee is not subjected to
physical or mental strains which can lead to ill-health or acci-
dents. Forms of remuneration and the distribution of working
hours shall also be taken into account in this connection. Clo-
sely controlled or restricted work shall be avoided or limited.
Efforts shall be made to ensure that work provides opportuni-
ties of variety, social contact and co-operation, as well as coher-
ence between different tasks.
Furthermore, efforts shall be made to ensure that working con-
ditions provide opportunities for personal and vocational devel-
opment, as well as for self-determination and professional
responsibility.
In 1991, a principal revision of the Act introduced a requirement
for employers to organize a Systematic Work Environment Man-
agement (AFS, 2001:1; 1993–2000 called ‘‘Internal Control”). This
regulatory revision had domestic origins but it is part of the inter-
national move towards process regulation and it transposes EU’s
Framework Directive (89/391/EEC) into Swedish law. A major mo-
tive was to enhance the hitherto poor assessment, prevention and
abatement of psychosocial risks (Frick, 2002). The management of
such risks was not helped by the lack of any provisions to specify
what 1978 Act required employers to do with the psychosocial
working conditions (see below).
However, SWEA did recruit some staff with psychosocial com-
petence. They were trained in groups, relations and individual psy-
chology but less in organizational issues and processes. They
developed a set of training materials for inspectors in 1984. During
1987–1989 a project tried to improve the inspectors’ supervision
of psychosocial risks, and after that each labour inspection district
assigned an inspector to promote this supervision. In 1990, a five-
week basic course for inspectors was organised. A handbook was
produced with examples of how to inspect psychosocial risks and
with examples of provision sections to cite when the inspectors re-
quired improvements of these conditions. However, SWEA’s
recruitment of more psychosocial competence was prevented by
the cutbacks in funding, due to Sweden’s economic crisis of
1991–1994. A few attempts by the authority’s experts to develop
more specific provisions on psychosocial risks were also cut short
by the Directors-General.
2.2. Focus on mandatory work environment management to inspect
psychosocial risks
The psychosocial handbook was withdrawn in 1993, when
SWEA started to implement the new regulatory strategy of system-
atic work environment management (as mentioned above). These
provisions required a proactive management of the quality of the
work environment, which created a legal base for a new method
to inspect psychosocial risks. Compared to technical risks, these
are harder for the inspectors to assess in detail and thus to specify
which improvements are required. ‘‘Simple” and delimited solu-
M
odel: The interaction between supervision chain and supervision strategy
State
objectives
Regulation Staff
Competence
Inspection
methods
Measures
required
Psychosocial
Work
Environment
Su
p
ervision chain
Input Output
WHAT:
HOW:
Supervision strategy of SWEA
2
The letters are an abbreviation of the Swedish words for work organization and
negative stress.
3
In practice though, some parts ended in late 2004.
576 A. Bruhn, K. Frick / Safety Science 49 (2011) 575–581

tions are rare in risks with organizational and relational origins,
and therefore for the labour inspectors to require. With the Sys-
tematic Work Environment Management provisions, the inspec-
tors could instead focus on the workplace procedures to manage
these risks. Employers are regularly ordered to assess their
employees’ subjective experiences of stress and other such risks.
They may also have to evaluate the effectiveness of the measures
they take to reduce the psychosocial risks. However, due to the
nature of the risks and the lack of more precise provisions, what
inspectors can require must in practice largely be based on negoti-
ations with local managers (and other stakeholders) to reach a
common perception of the psychosocial risks.
SWEA’s implementation of this regulated self-regulation – i.e.
mandatory management of the work environment – tries to sup-
port such a strategy of local development (Bruhn, 2006). It com-
bines inspectors’ traditional detection and assessment of risks
with a new supervision of the procedures by which employers en-
gage the workplace actors in a preventive management, one which
is effective to at least detect health and safety risks, and hopefully
also to and abate them. As a consequence of the new regulatory
strategy and of the increasing focus on psychosocial risks, SWEA’s
labour inspectors supervise these important aspects of how
employers organize and manage their businesses.
2.3. No comprehensive provisions on psychosocial risk factors
As mentioned, only the Act regulates psychosocial risks at work.
Although there is a principal consensus that psychosocial issues
are essential aspects of the work environment, since the work envi-
ronment reforms of the 1970s, it remains controversial to upgrade
the advisory ordinance from 1980 into mandatory psychosocial
provisions (Andersson et al., 1981; see Section 5.2).
There are three provisions on special delimited psychosocial
risks, on Work in Isolation (1982), on Violence and Threats in the
Working Environment (1993) and on Victimization at Work (or
bullying, 1993). As inspectors are able to base their improvement
demands on the specified regulatory requirements, these risks
are easier to supervise. However, this has also influenced what
was inspected. Before 2001, psychosocial inspections were often
reduced to these three issues; several inspectors even came to
equate them with all psychosocial risks (Frick et al., 2006).
2.4. The Swedish Work Environment Authority and the ARNE-
programme
However, at the end of the 1990s, there was a growing political
support and pressure on a more effective prevention also of psy-
chosocial health at work. Firstly, this came from the mentioned sig-
nals of costly psychosocial problems in the workplace (Ds,
2001:28). Secondly, there were some inspection incidents followed
by complicated judicial controversies. Last but not least, there was
a growing internal demand from groups of inspectors to develop
policy and procedures on these issues (Frostberg, 2005).
After a proposal from the authority, the government merged the
separate labour inspection districts and the National Board for
Occupational Safety and Health into the Swedish Work Environ-
ment Authority as of January 2001. The objective was to improve
the coordination and synchronization of the inspection activities,
and to ensure a uniform application of the provisions by the
inspection districts. SWEA then revised the internal control provi-
sions from 1993 into the present Systematic Work Environment
Management, and undertook to implement these provisions more
strongly.
To support a more effective prevention through inspection, the
government increased SWEA’s annual funding by some 7 million
Euros with directions to develop its competence on and inspection
of psychosocial risks:
The objective is to achieve a reduction of ill-health caused by
factors related to the organization and management of work.
The work environment factors referred to here are those that
can be influenced by concrete demands. They can concern neg-
ative stress, social factors, workload, repetitive physical work
etc. It can also concern on-the-job rehabilitation. (SWEA,
2001; SWEA’s appropriation letter from the government, our
translation)
This became SWEA’s mentioned ARNE-programme. ARNE com-
prised both ongoing and new development projects, with different
backgrounds, aims and results, including the methods develop-
ment project (Frick et al., 2006).
3. The methods development project and our evaluation design
3.1. Traditions of method development among labour inspection
districts
The methods development project and the rest of the ARNE-
programme, was started when the newly amalgamated SWEA tried
to integrate the formerly autonomous inspection districts and their
inspectors in the culture, co-operation and practice of a joint
authority. Districts and inspected had regarded the National Board
for Occupational Safety and Health as a source of expertise and
provisions on what employers must do, but less of methods on
how inspectors are to promote the implementation of these provi-
sions through their supervision. Some district methods were devel-
oped during the 1990s, to supervise the provisions on Internal
Control (the present Systematic Work Environment Management)
or psychosocial risks. The result was a (still limited) number of
methods, developed in parallel by different districts and often with
overlapping content. None of them attempted to integrate the two
issues, despite the expressed intention to do so in the reform of the
Work Environment Act of 1991 (Frick, 2002). Also, the districts’
own methods became part of a competition between them, which
strengthened their local cultures. For example, one method, ‘‘Focus
inspection”, developed in one district was forbidden in other dis-
tricts. This method was developed by inspectors with special com-
petence about psychosocial risks. It was based on a conviction that
it is sometimes worth while to allocate extensive time and re-
sources (here first and foremost on focus group interviews with
employees at different organizational levels) in order to develop
solid knowledge and gain long-term inspection results. The other
districts found this too much of a waste of resources.
This divergence in policy between districts was strengthened as
a consequence of an increasing number of such inspectors with
some competence in behavioural and social sciences in some dis-
tricts, while they remained very few in others in 2001. And where
inspectors with psychosocial competence had been recruited, they
saw the lack of suitable inspections methods and sometimes initi-
ated the development of such methods. But before ARNE, no
inspection method had been thus developed across the districts.
In 2001, SWEA assembled some local methods in a national
‘‘Handbook of inspection methods” in an attempt to counter this
fragmentation. The six methods included had all been approved
by SWEA’s development unit. Three were methods to inspect Sys-
tematic Work Environment Management, and one on psychosocial
risk factors – the highly contested Focus-method mentioned above.
However, ‘‘method”, was quite broadly defined. All should be based
upon the inspection process, as regulated by SWEA’s rules of
inspection (SWEA, 2008). However, a closer scrutiny in 2004 re-
vealed that several of the selected six methods were not really
A. Bruhn, K. Frick / Safety Science 49 (2011) 575–581 577

properly evaluated (Eriksson et al., 2004). For some that were, the
results were not entirely positive. Finally, despite being issued in a
national handbook, the use of these methods was both limited and
quite uneven between the districts.
Before this, SWEA’s Director-General had in 2002 issued a pol-
icy on how to develop inspection methods. Such projects had to
be approved by the Head office in Stockholm and their results
had to be evaluated and approved according to common criteria.
Although initiatives for new methods would still mainly originate
from the districts, these were to be under central control, to make
them more widely known, accepted and used.
3.2. The methods development project – objectives, programme base
and strategy
This history was a major part of the background of the methods
development project within the ARNE-programme. The project
contained seventeen quite varying special projects. Most tried to
improve the inspection of special issues of psychosocial risks.
ARNE’s general objective was ‘‘to achieve a distinct and coordi-
nated effort to strengthen the supervision in the field of work orga-
nization” (SWEA, 2002a; our translation). This objective was
explicitly connected to the implementation of Systematic Work
Environment Management in Swedish workplaces. This Manage-
ment was the ‘‘...most important basis for the improvement of
the psychosocial work environment”. Small firms and enterprises
were to get special attention within the methods development
project.
SWEA’s brief to the methods project (SWEA, 2002b), states that
monitoring the Systematic Work Environment Management is the
basis to inspect psychosocial risks, as it is a good entrance to these
conditions of work. However, the next sentence of the brief claims
that the inspection cannot always use these provisions when psy-
chosocial issues are inspected in small firms. There is therefore a
need for methods, general enough to be used in all kind of
workplaces.
SWEA gave the project very ambitious objectives. New methods
should enable inspectors to both detect and analyse sources of ill-
health, and identify the risks that may cause them in a proper legal
manner. Further, these inspection methods should stimulate and
support the overall organizational development at the workplaces.
The methods should also contribute to the identification of organi-
zational deficiencies behind risks that mainly are caused by ma-
chines, systems of production and chemical products. Despite the
recruitment of personnel with special psychosocial competence,
and despite training to increase the competence of existing person-
nel (i.e. the other ARNE projects), the methods to be developed
should provide all inspectors, even those without any particular
competence on psychosocial risks, with tools to effectively inspect
and issue legally sustainable requirements of different kinds of
risks in this quite broad and complex area. As there normally is a
very limited specialization between inspectors, technical oriented
inspectors thus also supervise psychosocial health risks at work.
3.3. The evaluation design as feedback to the Swedish Work
Environment Authority
Our evaluation of the methods development project was com-
missioned in the summer of 2003, when only half a year remained
of the programme. It was therefore designed as a hybrid between a
formative evaluation, a process evaluation and a more summative
one (Vedung, 1998). We used several sources of data: document
studies, so-called shadowing (following inspectors), qualitative
semi-structured interviews and focus groups (Kvale and Flick,
2007; Morgan, 1997; Pawson and Tilley, 1997). Data was gathered
between autumn 2003 and autumn 2005. Our evaluation does not
monitor the workplace effects of any particular inspection method;
the main focus is instead on the developmental process within the
methods project. A proper summative evaluation would have re-
quired a quite different design, with more time and resources to
measure the situation before and after the deployment of various
inspection methods. Nevertheless, our process evaluation would
have benefited from a start much earlier in the process of the
methods development project, in order contribute more interac-
tively to its development process.
The evaluation was commissioned in three parts. The first and
second studied the sub-projects on KASP and KOPS. The interim re-
ports on them evaluated the applicability and capacity of these two
methods (Bruhn et al., 2004; Lehto et al., 2004). The final report
evaluated the whole methods development project. On the one
hand, how the project set a joint and higher standard to develop
methods within SWEA, and on the other the utility and usefulness
of the methods developed (Frick et al., 2006). In the next section
we will deal with the first two parts of our evaluation – the KASP
and KOPS methods. In Section 5 we will go onto the main task in
this article, the developmental project as such.
4. Two inspection methods within the ARNE-programme
4.1. To communicate psychosocial risks through a prevention cycle tool
KASP is a general method to inspect psychosocial risks by apply-
ing the ‘‘cycle of control” in the Systematic Work Environment
Management. Like in all quality control, the steps of the cycle mon-
itor how workplace actors detect, assess, prevent or abate risks,
evaluate the results, and – if needed – take remedial action.
Although it focuses on psychosocial risks, KASP is to be applicable
by all inspectors in all inspection, and also to complement control
of specific elements, such as machinery or ventilation. It is to be
simple to use at the inspectors’ workplace meetings, primarily with
managers and safety reps. It is supported by a manual. This in-
cludes a broad list of possible psychosocial risk factors: large work-
load, high pace of work, complex tasks, simple repetitive tasks,
working alone, shift-work and irregular hours, risks of violence,
bullying, conflicts at work, unclear work roles or expectations, con-
tinuous change, working with people, social contacts and the phys-
ical environment and accident risks.
KASP was developed in collaboration between some inspectors
and researchers in communication psychology. It simplified an ear-
lier method to inspect, Systematic Work Environment Manage-
ment (Strangert, 2000). The quality control circle can be effective
both to monitor risks and to transmit knowledge about risk pre-
vention if it is used by an inspector with time to explain KASP as
steps in a preventive management. However, the method does
not contain any special tools to inspect psychosocial risks. The list
above requires special training in psychosocial risks to be applied
in inspection practice, as was underscored by the researchers in-
volved. Nor does KASP cover the entire workplace inspection, as
it is a tool for the primarily meeting only. It is therefore seen and
used more of a means of communication by the inspectors, than
a method to monitor psychosocial risks. After our evaluation, the
final development of KASP was halted, possibly because its objec-
tive and effectiveness was unclear from the beginning.
4.2. Communicating psychosocial risks in small firms
KOPS is presented as a method and an educative tool for inspec-
tors to work with and motivate employers of small firms (with 5–
20 employees) to more actively develop their work environment. A
KOPS-inspection starts by sending an introductory letter and a
questionnaire to employers and the safety reps at small workplac-
578 A. Bruhn, K. Frick / Safety Science 49 (2011) 575–581

es within the selected industry, with open questions on the organi-
zation and psychosocial climate. On the basis of the answers re-
ceived, the inspectors select firms to inspect (including those
that do not answer). At the workplace the inspector first meets
the manager and a safety rep or another employee representative
to inform them of the purpose and form of the inspection. Then
he/she chair a dialogue with five to eight employees with a focus
on stress and other difficult psychosocial situations and how these
are handled. At the end, the inspector tries to get a consensus on
which questions can be raised with the employer, and which
may possible be too sensitive. In the final meeting, the inspector,
discusses these questions with the employer–manager, and possi-
bly the safety rep, and explains and motivates the requirement in
the coming inspection notice to the firm.
In KOPS, the inspector assesses psychosocial risks on a three-le-
vel scale. The first is evident risks and deficiencies that result in
concrete requirements in the inspection notice. The second also
stands for a high risk level, but with more complex sources and
where no concrete requirement can be formulated to reduce the
risks. The third level of KOPS indicates only the presence of general
types of risks, but not at a directly hazardous level. In level one or
two firms, KOPS includes a follow-up inspection.
KOPS was the only method fully developed within ARNE’s
method development project. The need for effective methods to in-
spect small firms was well known, as these firms often lack knowl-
edge on work environment risks, and are aversive to the systematic
management and documentation mandated in the provisions on
Systematic Work Environment Management (Antonsson et al.,
2002). ARNE’s methods development project therefore appointed
a project leader to develop a method for small firms in co-opera-
tion with inspectors in one district. This inspector network was la-
ter extended to nearly all districts. As KOPS broke the isolated
district way of developing methods, it became quite well anchored
in SWEA’s organization.
Without this being mentioned in the KOPS-manual, the inspec-
tor uses a simple form of the quality control circle in the interview
one with employees. To apply this to psychosocial risks requires an
inspector with special competence on these issues. The unclear
connection between the provisions on Systematic Work Environ-
ment Management – which the inspector thus applies – and KOPS
was criticized in our evaluation. After some adjustments a shorter
course in how to use KOPS was developed within SWEA. Today the
method is used (under a different name) but only by inspectors
with special psychosocial competence.
5. Why the methods development project failed
5.1. Few workable psychosocial inspection methods were developed
The methods development project resulted in few workable
methods to inspect psychosocial risks, and KASP was never fin-
ished. Besides KOPS, only one more project initiated in the meth-
ods development project has resulted in an inspection method.
This is the originally Danish OBS (observation) method. Many of
the other projects were not really about inspections methods,
and a few that were, were never finished. After the methods devel-
opment project, SWEA only had two new inspection methods, but
only for use by specialized inspectors. To these can be added the
controversial existing Focus method, two others that were started
by districts before the methods development project and are now
used only or mainly within these districts, and a special method to
inspect stress at constructions sites that is being used within
SWEA. In all, the project did thus not achieve its objective to devel-
op general methods to inspect psychosocial risks that could be
used all of SWEA’s labour inspectors.
The methods development project may still have raised the
awareness within SWEA of psychosocial issues and of the difficulty
of inspecting complex issues. Yet, it failed also as a learning project.
It did not develop knowledge of what is required of psychosocial
risks inspections, not the least because there was little dialogue be-
tween the separate method sub-projects and their inspectors and
specialists. For example, the two methods described above repre-
sented two extremes that could have learnt from each other. KASP
was ‘‘oversteered”. The manual instructed the inspectors what to
do in too much detail, with not enough professional discretion
and practical adaptation to the unique situations. The inspectors
therefore ‘forgot’ the manual and only used the KASP-circle, which
they appreciated as a guidance and communication tool. KOPS on
the other hand was ‘‘understeered”. The manual was brief and gen-
eral with quite limited guidance on how to organise and manage
the method’s workplace dialogue. In practice, KOPS inspections re-
quired inspectors with personal competence in group relations and
other psychosocial issues, some of which should have been built
into the method.
5.2. Broad and unclear tasks to be solved in isolation by the methods
project
Why, then, did the methods development project not reach its
objectives? The short and simple answer is that SWEA treated it
as if it was a simple and delimited issue of guidelines or similar
tools. This might require work – for which some funding was allo-
cated – but the details could be left to a few experts to develop
among themselves. Given this, they could produce clear solutions
that could be applied all across the organization. This principal
and practical isolation made it hard for the methods development
project to achieve its given objectives.
To start, the methods project was isolated from SWEA’s strategy
and its renewed project on psychosocial provisions. Supervision is
a link in the chain of regulation and its implementation (see the
model above). It connects state policy and regulation on the one
hand with the local work environment on the other. The inspection
thereby transforms regulation into practice, into demands and ac-
tions of change (Hutter, 1997). In this process, the purpose of an
inspection method is, to help build a bridge between what is re-
quired in the general provisions and the competence of inspectors
to assess unique local work environments. Such bridges-methods
are especially needed, when the regulation is complex and there
are no evident criteria against which local compliance can be mon-
itored. This is the case with regulations for both psychosocial risks
and Systematic Work Environment Management, but less so for
more clear-cut technical provisions.
Yet, no method can bridge a too wide gap, i.e. from a vague and
broad regulation (such as the psychosocial section of the Act) to
the inspectors’ workplace assessments. Inspections, and methods
to support them, require a clear enough regulatory base, with guid-
ance on what is required of employers. Inspection methods thus
depend on the regulations, but on the other hand, they may also
influence what regulations stand for in practice, as the methods’
guidance may affect how inspectors interpret and apply the rules.
Despite this, ARNE’s methods development project had little con-
tact with SWEA’s renewed attempt to develop more specific psy-
chosocial provisions. Under instruction from the government, the
authority set this up from 2001. The joint effort of SWEA’s legal
and psychosocial experts resulted in draft provisions, which in
2003 were circulated for comment by the social partners and oth-
ers concerned. However, these were heavily criticized, especially
by the employer organizations that found them too far-reaching
and too interfering with their management prerogative to organize
production and therefore impossible to comply with. SWEA’s
Director-General – who had much been ordered to start the project
A. Bruhn, K. Frick / Safety Science 49 (2011) 575–581 579

by the government – then withdrew the proposal and cut the pro-
ject short (Ds, 2001:28; SWEA, 2001; Linde et al., 2005). The psy-
chosocial experts felt that they did not get the chance to fully
explore the feasibility of psychosocial provisions that were more
precise than the general requirements of the Act and still legally
fully applicable.
However, the Director-General feared that this was less relevant
as new provisions would be hard to implement against such a
strong opposition. Traditionally Swedish employers have a good
relation with SWEA as a supervising authority. They largely accept
the requirements in the inspectors’ improvements notices, with no
need for a much more cumbersome legal enforcement (Frick,
2009). SWEA feared that the enforcing controversial psychosocial
provisions could damage these positive but sensitive consensus
relations. They could also result in many court cases. And as the
provisions could anyhow not become as precise as technical ones,
these cases might result in negative precedents. The authority
therefore preferred to continue to rely on the employers’ duty to
assess and prevent or abate risks in the Systematic Work Environ-
ment Management, as its main strategy to improve the psychoso-
cial work environment. However, ARNE’s methods development
project was involved in neither the aborted provisions project,
nor in the top management’s decisions on the choice of strategy
to reduce psychosocial risks. The methods project thus got no guid-
ance on the more specific risks to reduce by the inspectors’ super-
vision and how this was to be done.
As mentioned, SWEA’s choice of Systematic Work Environment
Management as its main psychosocial strategy was confirmed in
the brief to ARNE’s methods project. Despite this, there was no
interaction between the methods development project and how
SWEA attempted to increase its supervision of the provisions on
this management after they were modernized from ‘internal con-
trol’ in 2001. As also mentioned, the methods project did not itself
connect the inherited KASP methods project or the new provisions
to their other inspection methods. Instead, the methods develop-
ment project rather cultivated the idea that Systematic Work Envi-
ronment Management was not very applicable, especially in small
firms, i.e. those forming the vast majority of employers. As a result,
neither SWEA nor the methods project gave the inspectors any
guidance on how to combine the two complex issues of what to
change of psychosocial risks and how to change this through the
strategy of Systematic Work Environment Management.
How different inspection methods can be applied – and thus
should be developed – also depends on the professional compe-
tence and leeway of inspectors, i.e. their competence and authority
to use their judgement in the process of the workplace inspection.
Again, this is much less clear-cut in inspecting psychosocial risks
than technical ones. There were – and still are – some important
policy issues on how to balance various aspects in the practice of
inspection that are unresolved and not much discussed within
SWEA, which also contributed to the too vague prerequisites for
the methods to be developed within the ARNE-programme. Firstly,
inspectors always have to balance between quantity and quality in
their supervision. Is SWEA to supervise as many workplaces as pos-
sible, to inspect the worst ones, to prioritize long-term improve-
ments in somewhat fewer workplaces, or something else? The
mentioned disagreement on the Focus-method to inspect psycho-
social risks – is it worth the extra time to use the method or not
– highlights this issue of quantity versus quality in the supervision.
Secondly, labour inspectors also have to combine control/sanctions
and support/advice. SWEA’s official focus is on the former, policing
role of inspectors, but this is hard to maintain in the supervision of
Systematic Work Environment Management and of psychosocial
risks. With the changes towards such more complex regulatory
requirements, inspectors increasingly have to support and guide
workplace actors in improving their work organization (Bruhn,
2006). This was also advocated in an internal SWEA report, at the
start of the ARNE-programme (Olsson, 2001). Finally, the practice
of inspection also has to strike a balance between on the one hand
the detailed regulations and SWEA’s aspiration of uniform inspec-
tions, and on the other the need for inspectors to use their discre-
tion in relation to the unique workplace circumstances. The latter
requires the development of the inspectors’ professional compe-
tence, but ARNE’s methods were to be usable for all labour
inspectors.
5.3. A lack of a base-line, of interaction within SWEA and of
managerial guidance
The method development project was also isolated from
SWEA’s practice of supervising the psychosocial work environ-
ment. No inventory was made of what the new the inspection
methods were to change, i.e. of the work organization and psycho-
social risks in working life. Nor did the project charter the inspec-
tors’ problems to supervise such risks and what methods they
needed to support them in this. Finally, no specific psychosocial
risks were identified as objectives for separate methods (except
for stress in construction, which not really was part of the project).
Besides the lack of a goals for the inspection methods – as de-
scribed above – the methods project therefore also lacked a
description of the existing situation and problems as a base-line
from which to proceed.
Finally, ARNE’s project to develop methods was left much to it-
self. The person in charge was placed within the department of
work environment experts in the main office, but he mainly
worked with the districts’ inspectors. Most of the diverse ideas
and projects were also inherited from the districts, several of which
were not intended as inspections methods. Neither did SWEA’s
high-level steering group help to alleviate the projects’ practical
problems and principal isolation. The group started late and was
more occupied with their separate interests as stakeholders than
with the objectives and strategies of ARNE and its inspection meth-
ods. These top managers may have been too busy with the devel-
opment of the newly merged authority, and the recruitment,
training and integration of many new inspectors. All the top man-
agers involved in ARNE were also engineers. They were trained and
competent in SWEA’s traditional techno-legal mission and culture
but had no professional competence in organizational or psychoso-
cial issues. They may have underestimated the complex problems
of inspecting psychosocial risk factors – i.e. inspecting the opera-
tion and results of how workplaces are organized and managed –
and how methods for this therefore unavoidably interact with
SWEA’s broader policies and practices. The recruitment of new
‘‘psychosocial” inspectors may indicate such a limited perspective.
These had training and–or experience in behavioural and group is-
sues, but not in organization and management, i.e. not in the over-
all organization and quality management of work that employers
are required to apply in their systematic work environment
management.
6. Conclusion: the methods development project in a broader
setting
Returning to our model and the central analytical question for-
mulated in the introduction – how and under what circumstances
effective inspection methods on psychosocial risks can be devel-
oped – we may state that the principals of the methods project
failed to analyze and relate to the project’s prerequisites. How
was this link in the chain to connect to the other ones? Any such
development project that does not take this, and the complexity
of its wider social context, into account is likely to fail. The diffi-
580 A. Bruhn, K. Frick / Safety Science 49 (2011) 575–581

culty to develop effective inspection methods on these kinds of
risks have to be tackled on at least two levels. At the first, there
is the problem as such: How to effectively inspect psychosocial
risks at the workplace? What is the nature and essence of such
risks? And, what kinds of methods are most effective to help
inspector to detect risks and to stimulate improvements that are
in accordance with the objectives set at the political level? In this
article we have only touched indirectly upon this level of the
inspection practice (see further Bruhn, 2006). At the second level
stands the question of how such effective methods can be devel-
oped in this very complex organizational and socio-political con-
text. The main objective of the article is to discuss this in relation
to a concrete project in the Swedish environment.
So what can be learned by this, possibly also other jurisdictions
with different work environment politics and infra-structures? No
simple answer can be given but some comments are possible:
(a) Labour inspection is built on a techno-legal tradition. If ins-
pectorates are to supervise the principally new organiza-
tional and psychosocial risks, their top managers need to
understand the issues and their complexities, and not to
see and handle them as ‘‘normal” technical regulations.
(b) Psychosocial regulations, and methods to supervise them,
require a challenging development co-operation between
psychosocial and legal experts. However, SWEA never fully
tested how far it is possible to specify legal requirements
of the psychosocial work environment. The co-operation
between legal and psychosocial experts in the provisions
project was instead cut short before they could reach any
such conclusion.
(c) Regulation and supervision of the psychosocial work envi-
ronment intervenes in how employers organize and manage
their businesses. It is a more sensitive issue than to regulate
technical conditions (which can be contentious enough). The
authority’s specification and application of psychosocial
requirements is therefore very much situated in the national
legal, political and industrial-relations structure. The Swed-
ish authority did not want to risk a head-on conflict with
the employers by pushing through more specific psychoso-
cial provisions, irrespective if these would have been feasible
or not. Instead, it continued to use the strategy of mandatory
quality management of the work environment that both
social partners supported.
(d) Whatever way it is regulated, an inspection of the psychoso-
cial work environment is therefore inevitably part of the
authority’s supervision strategy. It can never be an issue of
simple measuring compliance, but rather relates to the
broader issues of inspection strategies, such as policing ver-
sus advising, monitoring and guidance of inspectors versus
giving them leeway for professional interpretations of local
situations, and quantity versus quality in the allocation of
scarce inspection resources.
At a general level the failure of this Swedish method develop-
ment project illustrates that regulatory implementation is, as
many authors already have pointed out (see e.g. Hutter, 1997), a
very complex task. To specify and give proper guidance in how
to interpret and apply the state’s general objectives is difficult en-
ough in this very complex socio-political context of different stake-
holders. The next step, to implement, to make practice out of these
interpretations via concrete actions and methods is no less diffi-
cult. To have any chance to do so, the authority concerned must
thus face and try to resolve the underlying, complex issues that ob-
structed the development of workable inspection methods within
the Swedish ARNE-programme.
References
AFS, 2001:1. Systematiskt Arbetsmiljöarbete. <www.av.se/lawandjustice/
provisions>.
Andersson, P., Axelsson, Y., Braathen, T., Wahlström, A., 1981. Från ax till limpa –
eller framväxten av ASS
´publikation ‘Psykiska och sociala aspekter på
arbetsmiljön’. Institutionen för tillämpad psykologi, Uppsala University.
Antonsson, A.B., Birgersdotter, L., Bornberger-Dankvardt, S., 2002. Small Enterprises
in Sweden. Health and Safety and the Significance of Intermediaries in
Preventive Health and Safety, Arbete och hälsa 2002:1. Arbetslivsinstitutet,
Stockholm.
Bruhn, A., 2006. The inspector’s dilemma under regulated self-regulation. Policy and
Practice in Health and Safety 04 (2), 3–23.
Bruhn, A., Frick, K., Lehto, A., 2004. KASP – metoden: kommunikation om
arbetsorganisatoriska, sociala och psykologiska risker: utvärderingsrapport.
Arbetslivsinstitutet, Stockholm.
Ds, 2001:28. Långsiktig verksamhetsutveckling ur ett arbetsmiljöperspektiv.
Näringsdepartementet, Stockholm.
Eriksson, O., Adler, K., Holmlund, L., Sjöström Ross, Å., 2004. Vad vet vi om
effekterna av olika inspektionsmetoder? En rapport från projektet Effekter av
inspektion. Slutrapport oktober, Arbetsmiljöverket, Solna.
European Foundation for the Improvement of Living and Working Conditions, 2007.
Fourth European Working Conditions Survey. European Foundation for the
Improvement of Living and Working Conditions, Dublin. <http://
www.eurofound.europa.eu/pubdocs/2006/98/en/2/ef0698en.pdf>.
European Social Partners, 2008. Implementation of the European Autonomous
Framework Agreement on Work-related Stress. ETUC/CES, Business Europe,
CEEP, UEAPME, Brussels, 18 June 2008.
Frick, K., 2002. Sweden: occupational health and safety management strategies
from 1970–2001. In: Walters, D. (Ed.), Regulating Health and Safety
Management in the European Union. P.I.E. Peter Lang, Brussels.
Frick, K., 2004. Varför är arbetshälsan inte god? (Why isn’t the work environment
healthy?). In: Johansson, B., Frick, K., Johansson, J. (Eds.), Framtidens
arbetsmiljö- och tillsynsarbete. Studentlitteratur, Lund.
Frick, K., 2009. Health and safety representation in small firms: a Swedish success
under threat. In: Walters, D., Nichols, T. (Eds.), Workplace Health and Safety –
International Perspective on Worker Representation. Basingstoke, Palgrave
Macmillan.
Frick, K., Bruhn, A., Letho, A., 2006. Metodutveckling: en delrapport i utvärderingen
av utvecklingsprogrammet ARbetsorganisation och NEgativ stress (ARNE).
Rapport 2006:5. Arbetsmiljöverket, Stockholm.
Frostberg, C., 2005. Interview with Carin Frostberg, in charge of Systematic Work
Environment Management at ASS, 051130.
HSE, 2004. <http://www.hse.gov.uk/stress/standards>.
Hutter, B., 1997. Compliance: Regulation and Environment. Oxford University Press,
Oxford.
Kvale, S., Flick, U., 2007. Doing Interviews, first ed. Sage Publications, Thousand
Oaks, Calif.
Lehto, A., Frick, K., Bruhn, A., 2004. KOPS – metoden: kommunikation om
psykosocial arbetsmiljö i småföretag: en formativ utvärdering av AVs
pågående utveckling av KOPS-metoden. Arbetslivsinstitutet, Stockholm.
Linde, P., Lundberg, A.-L., Löf, B., 2005. AFS:en om sociala och psykologiska
arbetsmiljöförhållanden–vart tog den vägen? Uppsats i arbetsmiljörätt 10
poäng. Örebro Universitet, Örebro.
Lundberg, U., Melin, B., 2002. Stress in the development of musculoskeletal pain.
Pain Research and Clinical Management 12, 165–179.
Morgan, D., 1997. Focus Groups as Qualitative Research. Sage, London.
Olsson, H., 2001. Utveckling och förstärkning av Arbetsmiljöverket.
Arbetsmiljöverket Intern Rapport 2001-06-15. Solna.
Pawson, R., Tilley, N., 1997. Realistic Evaluation. Sage, London.
Strangert, B., 2000. Studier av yrkesinspektionens metodik vid systeminspektion av
internkontroll. Rapport 1/2000. Enheten för Arbets- och Organisation-
spsykologi. Umeå Universitet.
SWEA, 2001. Regleringsbrev för 2001, verksamhetsgren nr 1 Tillsyn 2000-12-21.
SWEA, 2002a. ARNE programbeskrivning utkast 2001-08-20, antagen 2002-04-22.
Arbetsmiljöverket, Solna.
SWEA, 2002b. Beställning av ARNEs metodutvecklingsprojekt, utkast 2001-12-14,
definitiv 2002-01-22. Arbetsmiljöverket, Solna.
SWEA, 2008. <www.av.se/rules_for_inspection_1_2008.pdf>.
Vedung, E., 1998. Utvärdering i politik och förvaltning. Studentlitteratur, Lund.
A. Bruhn, K. Frick / Safety Science 49 (2011) 575–581 581