Reincentivizing Work: A grounded theory of work and sick leave

Hans O. Thulesius, Ph.D. & Birgitta E. Grahn, Ph.D.

Abstract

Work capacity has a weak correlation to disease concepts, which
are insufficient to explain sick leave behavior. With data mainly
from Sweden, a welfare state with high sickness absence rates,
our aim was to develop an explanatory theory of work and sick
leave.We used classic grounded theory for analyzing data from
130 individual interviews of people working or on sick leave,
physicians, social security officers, and literature. More than
60,000 words and hundreds of typed and handwritten memos
were the basis for the writing up of the theory. In this paper we
present a theory of “reincentivizing work”. To understand
incentives we define work disability as hurt work drivers or work
traps. Work drivers are specified as work capacities + work
motivators, monetary and non-monetary. Incentives are
recognized when hurt work drivers are assessed and traps
identified. Reincentivizing is done by repairing hurt work drivers
and releasing from traps. In our theory of reincentivizing work,
hurt work drivers and traps are recognized and then repaired and
released. The theory may add to social psychological research on
work and sickness absence, and possibly inform future changes in
sick leave policies.

Background

The sickness absence rate in Sweden is one of the highest in
the world (Ljungqvist & Sargent, 1998). Sweden has generous
sick leave policies and strong job protection legislation. There is
no upper time limit for sick leave, and a low risk of loosing
employment due to sickness absence. Monetary compensation
from social security limits the loss of buying power to 0-20% after
tax for people with low to average incomes on sick leave (Esser,
2005). A too soft and disincentivizing social security system was a
central political issue leading to a shift in Swedish government
2006.

In welfare states such as Sweden monetary work motivators
are weak as compared to laissez-faire economies such as USA
(Rae, 2005) while non-monetary motivators for working such as
plight and pride are stronger in welfare states (Johansson &
Palme, 2004). Although Swedish sick leave compensation is
generous a sick leave trajectory often involves shame and
distrust. Against this background common disease concepts are
inadequate to explain sick leave behavior since work capacity
alone shows little correlation to disease severity (Melamed,
Groswasser& Stern, 1992; Riegel, 199; Englund, 2000). Therefore
our aim of this study was to generate an explanatory theory of
work and sick leave.

Methods

Data collection started in 2003. We did 22 formal and 40
informal interviews with people working and on sick leave,
informal interviews with 30 Swedish health care professionals
(physicians and nurses), and formal interviews and focus group
interviews with 6 employees of the Swedish social insurance
agency (Försäkringskassan, FK). We did secondary analysis of
taped and transcribed formal interviews with 20 participants in a
Swedish rehabilitation study (Grahn, Stigmar & Ekdahl, 1999;
Grahn, Borgquist & Ekdahl, 2004) and 12 American employees of
a public transportation company (Potts, 2005). We examined data
from expert group meetings, conferences, and literature data as
well as quantitative data on sick leave in a cohort of 196 people.
Participation by the first author in international grounded theory
workshops 2003-2006 was a source of both interview data and
memos.

We did classic grounded theory (GT) analysis according to
Glaser (Glaser, 1978, 1998, 2001, 2003, 2005) aiming at
generating conceptual theories that are abstract of time, place
and people. Classic GT differs from most studies using qualitative
data that often are presented as GT (Sekimoto, Imanaka, Kitano,
Ishizaki & Takahashi, 2006) by presenting explanatory concepts
rather than descriptions.

All of the data mentioned above was compared in the
analysis according to the grounded theory “all is data” dictum
(Glaser, 1998, p 145). Field notes from interviews not taped were
coded and compared in the same way as transcripts from taped
interviews. Concepts and categories emerged through a cyclic
process of collecting, coding, and comparing incidents in the data
by which concepts and categories relating to the incidents
originated. These concepts were then compared with each other
and new incidents as more data was collected and compared.
Theoretical memos, in the shapes of text, diagrams, and figures,
were written, typed, or drawn in the comparative process. More
than 60.000 words and hundreds of pages of typed and
handwritten memos sit in the memo bank from which this paper
was sorted and written up. ”Memos are the theorizing write-up of
ideas about substantive codes and their theoretically coded
relationships as they emerge during coding, collecting and
analyzing data, and during memoing” (p 177). Memos yielded
creative ideas of where to sample more data (theoretical
sampling), which was coded in a more selective way after a core
variable was discovered. More memos were produced followed by
sampling of more data and so on. Memoing is “the core stage of
grounded theory methodology” and should be done at any time
and place in order to capture creative ideas (p 177).

Memos were sorted and later written up in the last steps of
the grounded theory methodology. We now thoroughly compared
relationships between categories and concepts using different
theoretical codes (Glaser, 1978, 2001, 2005), and eventually a
dense substantial theory emerged. The writing of two working
papers enhanced the sorting of the memos: in 2005 for a research
report in Swedish (Thulesius, 2005) and in 2006 for a grounded
theory seminar. The intensity of the analytic process varied but
increased over time, and the theory was modified until the last
writings of this article.

Many quantitative clinical research methods consider
persons or patients as units of analysis, whereas in GT the unit of
analysis is the incident (Glaser & Strauss, 1967). The number of
incidents being coded and compared often amounts to several
hundred in a GT study since every participant often reports many
incidents. When comparing many incidents in a certain area, the
emerging concepts and the relationship between them are in
reality probability statements and therefore GT should not be
considered a qualitative method but a general method that can
use any kind of data (Glaser, 2003, p 1). However, although they
deal with probabilities most GT studies are considered as
qualitative since statistical methods are not used, and figures not
presented.

The inductive nature of GT with hypotheses being generated,
not tested as in traditional quantitative research is important for
the method, and has its roots in quantitative inductive research
(Glaser, 1998, pp 22-31; Lazarsfeld & Thielens, 1958). Thus, the
results of grounded theory are not reports of facts but probability
statements about the relationship between concepts, or an
integrated set of conceptual hypotheses developed from empirical
data. Validity in its traditional sense is consequently not an issue
in GT research, which instead should be judged by fit, relevance,
workability, and modifiability (Glaser, 1998, p 18). Fit has to do
with how close concepts fit with the incidents they are
representing, and this is related to how thorough the constant
comparison of incidents to concepts was done. A relevant study
deals with the real concern of participants and grabs attention.
The theory works when it explains how the problem is being
solved with much variation. A modifiable theory can be altered
when new relevant data is compared to existing data. A GT is
never right or wrong, it just has more or less fit, relevance,
workability and modifiability, and readers of this article are
asked to try its quality according to these principles.

This study was approved by the regional ethics committee at
Lund University, and formal interviews were made with informed
consent from participants.

Results

The theory of “reincentivizing work” first requires the
understanding of concepts generated for this study that explain
different aspects of hurt incentives. Then we show how these hurt
incentives are recognized, and finally different reincentivizing
activities are presented in Table 1.

Table 1 Reincentivizing work, theory outline.

• Understanding incentives by specifying drivers and traps

• Recognizing hurt incentives:

– driver assessments

– mode traps

• Reincentivizing work:

– capacity repair

– capacity and non-monetary motivator repair

– monetary motivator repair

– trap release

Understanding Incentives by Specifying Drivers and
Traps

Understanding what incentives consist of is important for
reincentivizing. The notions of traps and drivers can explain
many issues regarding disability and sick listing. In this study
drivers are specified as a combination of motivators and
capacities to pursue a certain mode of occupation such as work.
Traps are situations where drivers are locked in a certain mode.

A driver is what makes you go on in a mode and a
trap is what prevents you from getting out of it”
from theoretical memo

We define a driver as a combination of motivators and
capacities. Work capacities are education, health, training,
physical and psychological conditions, and social skills etc. Nonmonetary
work motivators are fellowship, identity, meaning,
desire, plight, pride, and “flow” (Csikszentmihalyi, 1995) but also
shame-avoidance (Sachs & Krantz, 1991) etc. Monetary work
motivators (or non-work motivators) are wage and sick leave
compensation but also unemployment benefits, fringe benefits or
expenses such as meal costs, clothes, traveling, and time for
repairs of homes, cars etc.

A change in motivators or capacities can hurt mode drivers.
Illness may hurt a work mode driver if the work mode capacity
goes down. Eventually a hurt work driver may cause sick leave.
Hurt capacities and motivators eventually trap a person in a
certain mode, (see below). In addition, time dependant inertia can
trap mode drivers. This means that the longer a person has been
in a certain mode the more difficult it is to change that mode, and
thus the person gets trapped. Thus, if a person has been on sick
leave for a long time it is difficult to go back to work since the
inertia that comes from being in the sick leave mode for a certain
time prevents the person from going back to work.

…after two to three months of sickness absence
the patient often gets stuck in a sickness role that
is very difficult to get out of” (physician expert,
FK social insurance)

… after three months of sick leave it is difficult
for people to return to work (physician, male
middle aged)

Recognizing Hurt Incentives

Many driver assessments are done in sick leave situations.
This either results in reincentivizing or disincentivizing a work
return.

Mode driver calculation. Primarily, every person aims for her
optimal “being mode” by an automatic mode driver calculation
(Mdc), modified after Ekström (2005). A Mdc has three main
outcomes: preserving a mode, limiting losses within a mode, and
eventually reevaluating a mode. The Mdc weighs up mode
motivators and capacities in a cost-benefit calculus. Let’s say an
ill person is uncertain about being able to work since he/she feels
depressed or suffers pain while working. So the work driver is
hurt. i) Then mode preserving is first done: Enduring anxiety and
pain by sticking to fundamental beliefs, strategies, and
explanatory models. Keeping up habits, goals and daily life and
continue working. ii) Or the person goes on to limiting losses:
Trying to stay in the mode as long as possible. Trying to master
the situation by seeking knowledge, investing in life style
changes, new health care contacts, or cutting down work,
changing work tasks, taking short sick leaves or holidays. iii) Or
the person eventually re-evaluates the situation: Changing the
mode by going on long sick leave, or changing job. If illness is
severe enough the preserving and limiting stages are bypassed
into immediate reevaluation.

Sick leave in itself can be seen as a mode with its own
drivers. The Mdc basically determines whether an ill person
works or stays at home. Ill health is then only one factor in the
calculus. An ill person with an otherwise high work capacity
combined with strong work motivators (monetary and nonmonetary)
has a strong work driver and a low risk of sick leave.
Another ill person with an otherwise low work capacity and weak
work motivators has a high risk of sick leave. (See Table 1)

Being modes affect motivators. While working the ill person
primarily wants to preserve the work mode, but may eventually
reevaluate the situation and go on sick leave. Having been on sick
leave for some time the sick leave mode gets stronger through
inertia. The Mdc now preserves another status quo and thus
either reincentivizes work or chooses sick leave (see trapped mode
drivers).

Should I go on working despite my symptoms or
stay at home?

Should I return to work now as my symptoms are
reduced or should I stay at home until they are
completely gone?

If I stay home from work what is the cost in terms
of money and/or humiliation from my
employer/fellow workers and/or the social
insurance and what are the gains in terms of
reduced suffering? (From a theoretical memo)

Other participants in the sick leave situation also make
assessments and mode driver calculations:

Employer assessment. Employers may use a calculus similar
to the Mdc when an employee turns ill. Preserving the existing
situation is first done. This is followed by limiting losses, i.e.
having the person cut down on tasks. Finally, the employer is
reevaluating the situation, often by replacing the person by
another employee, eventually permanently. This replacement
reevaluating strategy disincentives work return. But, if the
employer regularly contacts absent employees and cooperates
with the FK (social insurance) this can reincentivize work return.
In rehabilitation planning the employer input is crucial for a
work return.

Physician assessment. Physicians also calculate hurt work
drivers in a physician assessment, which can either reincentivize
or disincentivize work return. When writing sick leave notes
physicians are either reincentivizing work return by being
restrictive about sick leave:

You don’t need sick leave for this condition; you
can actually go on working! (middle-aged male
physician),

or disincentivizing it by doing what the patient wants (Carlsen &
Norheim, 2005):

How long sick leave do you want [me to write in
the sick leave note] (middle aged male physician)

Social insurance (FK) officials ’assessments. When assessing
requests for sick leave FK either reincentivizes work return by
handling cases restrictively – “Tiredness is not a reason for sick
leave
” (FK executive), or disincentivizes work return by speeding
up “client” turn-over and promptly providing sick leave benefits
the trick is to feed the PUMA (permanent and automatic benefit
payment without control of sick leave status, abbreviated PUMA
in Swedish
)” (FK official). So whether an ill person goes on sick
leave depends on the Mdc, and assessments of employer,
physicians and FK officials. But there is also a higher societal or
macro level that determines sick leave behavior:

Macro level assessment. On the macro or society level the
social insurance has three ways to go in the sick leave situation.
Either preserving the present sick leave policies regarding
legislation and compensation levels; or limiting sick leave by
moderately restricting policies or by influencing attitudes
towards sick leave; or reappraising the situation by radical
changes of policies.

Mode traps further explain why a person is on sick leave. A
person can get trapped in a certain mode through different
(dis)incentivizers such as inertia, changing motivators or
capacities. There are different drivers for different modes and
these can trap the individual from reincentivizing work, i.e. from
going back to work or go on working. Below are a few examples of
traps associated with work and sickness absence.

Body trap. A person suffering pain or ill-health can be said to
have a body-trapped work driver. This is the traditional reason
for sick leave. The work motivators may be there, but they are
locked in the hurt body. Basically, “body trap” means that your
body prevents you from working. Work motivators could be high
but body capacity is low.

It is like your body energy is trapped, you can
barely handle everyday tasks and work is
unthinkable” (middle-aged woman)

When body says no, work incentives are low
(middle-aged man).

Poverty trap. Monetary disincentivizers in the Swedish labor
market have been recognized by the government report “Out of
the poverty trap” (Swedish Gov Report, 2001). In Sweden, it is
difficult for persons on long-term sick leave with a low income to
increase their income by returning to work due to marginal
effects of the social security system. These marginal effects are
disincentivizing work by reincentivizing non-work modes.

Fox trap. A person on sick leave having a limited work
capacity belongs neither in the work mode nor in the sick leave
mode. Instead of being on full time sick leave or to go on working
full time despite illness a person in Sweden partial time sick
leave, which is quite common in Sweden. However, being without
employment the person is in the fox trap – “you are too healthy to
be on sick leave
” says the FK official, while the employment
service agent says “you are too ill to be working”.

System trappers. Some people in welfare states take
advantage of the compensations in the welfare system. One might
say that they are “working the system”, and we could also call
them “system trappers” since their behavior could be compared to
that of hunters and gatherers. In our data “system trappers” are
more frequent in remote areas where people traditionally make
their living from hunting, fishing, and forestry, and where
regular jobs are limited. In scarcely populated parts of Canada it
is considered acceptable to work as little as possible and yet get
social security. In Sweden scarcely populated areas have the
highest number of people on sick leave and unemployment
benefits. Attitudes towards such benefits in these areas are less
linked to shame than in other areas with a stronger labor market.
So, work motivators seem to vary culturally, geographically and
demographically.

Honey trap. Too much stimulation by strong motivators, both
monetary and non-monetary will eventually trap a person in a
high pace work mode difficult to get out of. This might provoke
illness and a limited future work capacity. People working with
creative tasks thus risk getting stuck in the honey trap. Family
life and leisure becomes annoying breaks in work, which becomes
the primary meaning of life. The honey trap involves a
reincentivizing positive feedback mechanism. The incentive
makes you work more, which gives more incentive, and finally
you cannot stop working at a pace that is too high for your
capacities.

At X the honey trap is a fact. People get here from
all over the world. They love their work – solving
problems etc, and if they don’t watch out they get
stuck in the (honey) trap… (Middle aged employee
with creative job at multinational company)

Reincentivizing Work by Capacity and motivator
Repair

Reincentivizing work is done through repairing hurt work
drivers, i.e. hurt capacities and motivators for work. When
drivers are hurt they need repair, and by repairing them the
traps get released.

Capacity repair

Reincentivizing work by improving the health and well-being
of a person on sick leave is fundamental. We call one aspect of it
body repair. Body repairs for impaired body capacities are
medication, physiotherapy, surgery, rehabilitation programs
(Grahn et al, 1999, 2004) and alternative therapies. Successful
treatments eventually reincentivize the work return. Irreparable
illness often leads to disability compensation such as a sick leave
pension.

Capacity and non-monetary motivator repair

Self repair. Socializing with friends and relatives, keeping
pets, physical exercise, and hobbies may enhance non-monetary
motivators and restore work capacity at the same time. This is
achieved by an improved well-being which ameliorates work
return. However, long duration of self-repair activities may
weaken the work driver since time away from work
disincentivizes work return.

Work-place repair. Making the work place a better
environment for the employee can reincentivize work. Emotional
strains caused by bad management risks eroding work identity, a
powerful work motivator. It is therefore important that
supervisors try to create a positive emotional atmosphere
(Nordqvist, Holmqvist & Alexanderson, 2003). Structured backto-
work programs where absent employees are contacted and
fellow workers informed of possible changes in task assignments
when the absentee returns are also beneficial. It seems as the
more employers are engaged in rehab programs the more work
can be reincentivized (Nordqvist, Holmqvist & Alexanderson,
2003).

Rehumanizing. Strengthening non-monetary work
motivators and thus increasing work capacity can prevent a
person from going on sick leave. This can be achieved by joining
support networks in the workplace that may initiate a
rehumanizing process (Holton, 2006) promoting authenticity,
safety and healing. By giving network members challenge,
experimentation, and creativity, this can provide the worker with
new energy and learning.

Monetary motivator repair

Controlling sick leave insurance. There are three main ways
to reincentivize work by controlling the sick leave insurance.
First, making it more difficult to obtain by controlling its
eligibility. Second, controlling non-monetary motivators, and
third making it less financially beneficial to be on sick leave.

Controlling insurance eligibility. Reincentivizing would be
enhanced by a stricter control of the sick leave insurance
eligibility, which has been characterized as being too “soft”. A
stricter control means that FK and employer assessments have to
be tougher. Hence, the trust in the Mdc and physician assessment
is often reduced. A 2006 government report suggests the use of
Medical Disability Advisor (MDA) guidelines from the USA for
limiting the length of sick leave periods (Swedish Gov Report,
2006).

Controlling non-monetary motivators. Shame, fear and plight
could disincentivize sick leave. In national multimedia ad
campaigns FK linked sick leave to shameful behavior and subtle
fraud. Hence, by inflicting shame, and appealing to societal plight
people would became less prone to go on sick leave.

…it (the ad campaign) puts a sick leave controller
in the head of the person on sick leave. (Regional
FK CEO)

Controlling monetary compensation. Hurt monetary
motivators disincentivize a return to work for those who have
been on sick leave long enough to trust the monthly payments
from the FK. By cutting down monetary compensation levels of
sick leave (and of unemployment benefits) it might be possible to
reincentivize work.

Sick leave would probably go down if
compensation levels were lowered… (Former
national FK CEO)

Strengthening monetary work motivators. Making work
monetarily advantageous in relation to non-work could be done
on the macro level by using tax policies. This is used in the UK
where working families get a special tax deduction as compared
to families on welfare. In Sweden the new 2006 government
launched a tax deduction eligible only for workers, not for people
on sick leave or retirement pension.

Trap Release

Traps are essentially released by the above repair strategies.
Either body and/or work place repair can release from a body
trap
. Improving impaired health situations and work place
conditions can help workers with health problems to return to
work. Controlling sick leave insurance and strengthening
monetary work motivators might get people out of the poverty
trap
. By all three strategies a fox trap can be released. Education
or job training programs could release from the fox trap by
increasing work capacity.

System trappers can be controlled by sick leave insurance
repair. A stricter control of eligibility and reduced monetary
compensation of different types of social insurance will prevent
people from abusing the welfare system.

A Honey trap can be prevented by work place repair. Some
employers are aware of “honey-traps” and prevent their
employees from getting consumed by over-motivating jobs. They
sense signals of over-stimulation and require that employees take
time off. So a release from the honey-trap can be done through an
initiative from the employer or another person in order to prevent
a future damage to the work driver.

Discussion

In this study of work and sick leave we present a theory
explaining why it may be difficult to return to work after sick
leave, and what can be done to reincentivize the return.
“Reincentivizing work” indicates that work motivators, both
monetary and non-monetary, and not only health related factors
are important in the process of a work return (Fryers, 2006).
Reincentivizing is a theory that fits with the wide range of data
from which it was generated. It also works to explain many work
and sick leave related issues. The theory applies to the Swedish
situation with one of the highest sickness absence rates in the
world, but we think reincentivizing is relevant for other settings
as well. Reincentivizing starts with understanding incentives by
specifying the driver and trap concepts that are central to
comprehend the theory: then follows recognizing hurt drivers and
traps. Third, reincentivizing work is done by repairing hurt
capacities and motivators, and releasing from traps.

To develop the reincentivizing theory we did classic
grounded theory (GT) analysis according to Glaser (1978, 1998,
2001, 2003, 2005). We interviewed people working or on sick
leave as well as physicians and social insurance officials, and also
analyzed literature. Our procedure was comparable to a previous
study in a different substantive area Thulesius, Hakansson &
Petersson, 2003). GT is the most quoted single method for
analyzing qualitative data according to a Google Scholar search.
Yet classic GT studies are rare. They represented 10% of 200
consecutive studies referring to the method in a PubMed search
in 2005-2006 done by the first author. Most studies were
descriptive and lacked a core variable theory, which is required in
classic GT.

The concept driver is fundamental to this study and
commonly used in contemporary Swedish language: “what is your
driver?”, “what is the driver in your life…”. In GT this is called an
in-vivo code, i.e. it comes from the interview data. Trap is another
in-vivo code from the area of sick leave used by unions, employer
organizations, and government agencies. The body trap concept is
also an in-vivo code. To be in a honey trap resembles the
colloquial expression “workaholic”. Poverty trap is a concept
borrowed from a Swedish government report (Swedish Gov
Report, 2001). A similar concept is called “low pay traps” that are
disincentives for people to stay in the workforce (Quintini &
Swaim, 2003). Fox trap is a concept found in a white-collar
workers union report. The mode driver calculation (Mdc) is a
concept generated by inspiration from two grounded theories –
“Cutting back after a heart attack” (Mullen, 1978), “Keeping my
ways of being” (Ekstrom, Esseveld & Hovelius, 2005) and Jeremy
Bentham’s “hedonic calculus” (Bentham, 1996). Mullen suggested
that people having suffered a heart attack “cut back” in their
lives after a complex calculus. Ekström proposed that women in
midlife apply a personal calculus to keep up their way of being
when faced with insecurity caused by midlife changes. Jeremy
Bentham in 1798 claimed that every person was aiming for
ultimate happiness by applying a “hedonic calculus” in life:
“promoting whatever factors led to the increase of pleasure and
suppressing those which produced pain”.

Our theory of reincentivizing work fits in the literature on
work, sickness absence and unemployment in several diverse
fields such as sociology, economics and medicine. It attempts to
integrate previous research findings together with new empirical
data in an explanation of what motivates complex fundamental
human behavior such as work. Theoretical explanatory models
for sick leave behavior are scarce. A process model explaining
absenteeism with data from the USA has been presented (Steers
& Rhodes, 1984). It includes different variables such as workrelated
attitudes, personal factors, market factors and cultural
and organizational norms in an organizing framework for
understanding absence research. Our theory of reincentivizing
seems to fit into that framework, yet with a more parsimonious
explanation. Historically work incentives seem to be about
balancing between working for a greater good such as society or
God, and working for profit (apart from working for supporting
life processes). In our study the non-monetary and monetary
motivators for working represent this balance. In typical welfare
states such as Sweden plight motivators are stronger than in
laissez-faire economies such as USA (Esser, 2005). This is
reflected in high Swedish compensation levels and a weak control
system for sick leave. The Swedish expression “writing your own
sick leave note” typically indicates the ease by which sick leave
may be attained in this country. Societal stability motives for
having generous sick leave policies — possible reduced costs of
health care, basic social welfare, policing, and drug control —
could legitimize the present high compensation levels. But
between 1997 and 2003 both unemployment and sickness absence
increased in Sweden to levels allegedly threatening the working
morale of the population and eventually the foundation of the
welfare state (Rae, 2005) Hence, a crucial issue in the 2006
parliament election campaign was to reduce the high number of
people outside of the work force. This led to the first shift in
government for 12 years with the new government suggesting
lowered compensation for sick leave and unemployment. This was
a political risk taking since 14% of the Swedish population
depends on sick leave insurance or disability pension for their
daily living (Rae, 2005). A November 2006 government report
suggested stricter sick leave assessments using the length of
mean sickness absence periods in the USA as a standard
(Swedish Gov Report, 2006).

It may be argued that the value of our study is limited since
it is not traditionally deductive. Neither is it a full description of
the sick leave phenomenon. It is rather inductive since GT is
primarily an inductive method. Reincentivizing work is according
to GT a theory with a certain degree of probability and ability to
provide an explanatory account of the area under study. It is not
a presentation of proven facts but a suggested conceptual
explanation of what is going on in the area of work and sick leave.
We admit that we may have missed data in our comparative
analysis of sick leave and work. We did for instance not study
self-employed people. Yet there is Swedish data showing lower
odds of sick leave in self-employed despite more subjective illness
as compared to matched controls (Holmberg, Thelin, Stiernstrom
& Svardsudd, 2004). However, we trust that our theory is
modifiable when “missing” data is entered into the analysis.
Thus, by adding new data more concepts will eventually be
generated that will add to the theory, not contradict it. We
therefore encourage readers to pursue research in this field, and
refine and improve the suggested theory of reincentivizing work.

Conclusions

We have developed a theory suggesting that complex drivers
determine people’s behavior. These drivers can work either to
incentivize or to reincentivize different modes. To deal with sick
leave according to the theory of reincentivizing work first
requires an understanding of the concepts of drivers and traps.
Then hurt drivers and traps are recognized and eventually
repaired and released. The theory of reincentivizing work could
give ideas for future research, and possibly inform changes in sick
leave policies.

Table 2 Work Driver & Risk of Sick Leave
A 2×2 table presenting work driver and risk of sick leave as a
function of degree of work capacity and work motivators
[please see PDF version for table]

Authors’ Contributions

HT and BG together conceived, designed, and collected data
for the study. HT did the grounded theory analysis and drafted
the manuscript in collaboration with BG. Both authors read and
approved the final manuscript.

Acknowledgements

We thank: Alf Södergren and Peter Burman,
Försäkringskassan Kronoberg, for supplying data and fundings;
patients, colleagues, officials at Försäkringskassan; participants
at international GT workshops; Dr Barney G. Glaser for valuable
support, and input for naming the core variable; PhD-candidate
Bibi Potts, for providing data from the USA for secondary
analysis; Professor Olav Thulesius for help with the manuscript.

Authors

Hans Thulesius, Ph.D.Department of Clinical Sciences Malmö, Division of Family
Medicine, Lund University, Sweden
hans.thulesius@ltkronoberg.se

Birgitta E. Grahn, Ph.D.
Department of Health Sciences, Division of Physiotherapy, Lund
University, Sweden.
3 Welfare Research and Development Centre of Southern
Smaland, Box 1223, SE-351 12 Växjö, Sweden
birgitta.grahn@fouviss.se

References

Bentham J: An introduction to the Principles of Morals and
Legislation. London: Oxford University Press. 1789.
Reprint 1996.

Carlsen B, Norheim OF: “Saying no is no easy matter” A
qualitative study of competing concerns in rationing
decisions in general practice. BMC Health Services
Research 2005, 5:70

Csikszentmihalyi M: Creativity: Flow and the psychology of
discovery and invention. New York: Harper Perennial;
1995.

Ekström H, Esseveld J, Hovelius B: Keeping My Ways of Being:
Middle-aged women dealing with the passage through
menopause. The Grounded Theory Review 2005, 5:21-53.

Englund L: Sick-listing – Attitudes and doctor’s practice. With
special emphasis on sick-listing practice in primary
health care. PhD-thesis. Uppsala University, Department
of Public Health and Caring Sciences; 2000.

Esser I: Why Work? Comparative Studies on Welfare Regimes
and Individuals’ Work Orientations. PhD-thesis.
Stockholm University, Department of Sociology; 2005.

Fryers T: Work, identity and health. Clin Pract Epidemol Ment
Health 2006, 2:12.

Glaser BG: Theoretical Sensitivity: Advances in the methodology
of grounded theory. Mill Valley: Sociology Press; 1978.

Glaser BG: Doing Grounded Theory. Issues and discussions. Mill
Valley: Sociology Press; 1998.

Glaser BG: The Grounded Theory Perspective I:
Conceptualization Contrasted with Description. Mill
Valley: Sociology Press; 2001.

Glaser BG: The Grounded Theory Perspective II: Description’s
Remodeling of Grounded Theory. Mill Valley: Sociology
Press; 2003.

Glaser BG: The Grounded Theory Perspective III: Theoretical
Coding. Mill Valley: Sociology Press; 2005.

Glaser BG, Strauss AL: Discovery of Grounded Theory: Strategies
for qualitative research. Chicago: Aldine; 1967.

Goine H: Sickness absence–Aspects of measurement, impact of
the labour market and effects of intervention. PhD-thesis.
Mid Sweden University, Health Sciences Department;
2006.

Grahn B, Stigmar K, Ekdahl C: Motivation for change in patients
with prolonged musculoskeletal disorders: a qualitative
two-year follow-up study. Physiother Res Int 1999, 4:170-
89.

Grahn BE, Borgquist LA, Ekdahl C: Rehabilitation benefits
highly motivated patients: a six-year prospective costeffectiveness
study. Int J Technol Assess Health Care
2004, 20:214-21.

Holmberg S, Thelin A, Stiernström EL, Svärdsudd K:
Psychosocial factors and low back pain, consultations, and
sick leave among farmers and rural referents: a
population-based study. J Occup Environ Med 2004,
46:993-8.

Holton, J: Rehumanising Knowledge Work through Fluctuating
Support Networks: A Grounded Theory. PhD-thesis.
University of Northampton, 2006.

Johansson P, Palme M: “Moral hazard and sickness insurance:
Empirical evidence from a sickness insurance reform in
Sweden”. Institute for Labour Market Policy Evaluation
(IFAU). Working Paper 2004, 10
http://ideas.repec.org/p/hhs/ifauwp/2004_010.html

Lazarsfeld P, Thielens W: The academic mind. Glencoe: Free
Press; 1958.

Ljungqvist L, Sargent TJ: The European Unemployment
Dilemma. J Polit Economy 1998, 106:514-550.

Melamed S, Groswasser Z, Stern MJ: Acceptance of disability,
work involvement and subjective rehabilitation status of
traumatic brain-injured (TBI) patients. Brain Inj 1992,
6:233-43.

Mullen PD: Cutting back after a heart attack: an overview.
Health Educ Monogr 1978, 6:295-311.

Nordqvist C, Holmqvist C, Alexanderson K: Views of laypersons
on the role employers play in return to work when sicklisted.
J Occup Rehabil 2003,13:11-20.

Potts B: Interviews with 12 RTA-employees 2002-2003. Personal
communication, GT workshop Mill Valley 2005.

Quintini G, Swaim P: Employing the non-employed. OECD
Observer 2003, 239.
http://www.oecdobserver.org/news/fullstory.php/aid/1079/
Employing_the_non-employed.html

Rae D: How to reduce sickness absences in Sweden: lessons from
international experience. OECD Economics department
working papers, 2005, 29.
http://www.olis.oecd.org/olis/2006doc.nsf/43bb6130e5e86e
5fc12569fa005d004c/8b02a0bbb12ae243c12571fe002c1d17
/$FILE/JT03214486.DOC

Riegel BJ: Contributors to cardiac invalidism after acute
myocardial infarction. Coron Artery Dis 1993, 4:315-20.

Sachs L, Krantz I: Anthropology of medicine and society: A new
perspective for a multidisciplinary audience. Stockholm:
Department of International Health Care and Research;
1991.

Sekimoto M, Imanaka Y, Kitano N, Ishizaki T, Takahashi O: Why
are physicians not persuaded by scientific evidence? A
grounded theory interview study. BMC Health Serv Res
2006, 6:92

Steers R, Rhodes S: Knowledge and Speculation about
Absenteeism. In: Goodman P, Atkin R, editors.
Absenteeism. San Francisco: Jossey-Bass;1984: 229-275.

Swedish Government Reports: Out of the poverty trap [Swedish].
Stockholm: SOU 2001: 24 (in Swedish).
http://www.regeringen.se/download/82f99a95.pdf?major=1
&minor=2776&cn=attachmentPublDuplicator_0_attachm
ent

Swedish Government Reports: More insurance and more work
[Swedish]. Stockholm: SOU 2006: 86 (in Swedish).
http://www.regeringen.se/download/fdb4ae15.pdf?major=1
&minor=72124&cn=attachmentPublDuplicator_0_attach
ment

Thulesius H: Evaluation of Diagnostic Centre in Kronoberg
County 2001-2003: A more reliable sick-listing?
[Swedish]. Växjö: FoU Kronoberg; 2005.

Thulesius H, Håkansson A, Petersson K: Balancing: a basic
process in end-of-life cancer care. Qual Health Res 2003,
13:1353-7.

Facebooktwittergoogle_plusredditpinterestlinkedinmail