Content uploaded by Bent Flyvbjerg
Author content
All content in this area was uploaded by Bent Flyvbjerg on Jan 08, 2016
Content may be subject to copyright.
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
117
FIVE MISUNDERSTANDINGS ABOUT CASE-STUDY
RESEARCH
Bent Flyvbjerg
flyvbjerg@plan.aau.dk
Five Misunderstandings About Case-Study Research
This article examines five common misunderstandings about case-study research: (1)
Theoretical knowledge is more useful than practical knowledge; (2) since one cannot
generalize from a single case, the single case study cannot contribute to scientific develop-
ment; (3) the case study is most useful for generating hypotheses, while other methods
are more suitable for hypotheses-testing and theory-building; (4) the case study contains
a bias toward verification; and (5) it is often difficult to summarize specific case studies.
The article explains and discusses these misunderstandings one by one, and concludes
with the Kuhnian insight that a scientific discipline without a large number of thoroughly
executed case studies is a discipline without systematic production of exemplars, and that
a discipline without exemplars is an ineffective one. Social science will be strengthened
with the production of better case studies.
Keywords:
Case study research, narrative, learning, case selection, hypothesis testing, validity,
scientific generalizability, vias, falsification, critical cases
Introduction
When I first became interested in in-depth case-study research, I was trying to
understand how power and rationality shape each other and form the urban
environments in which we live (Flyvbjerg 1998a). It was clear to me that in
order to understand a complex issue like this, in-depth case-study research was
necessary. It was equally clear, however, that my teachers and colleagues kept
dissuading me from employing this particular research methodology.
‘You cannot generalize from a single case,’ some would say, ‘and social sci-
ence is about generalizing.’ Others would argue that the case study may be well
suited for pilot studies but not for full-fledged research schemes. Others again
would comment that the case study is subjective, giving too much scope for the
researcher’s own interpretations. Thus the validity of case studies would be
wanting, they argued.
Sosiologisk tidsskrift: VOL 12, 117–142 © Universitetsforlaget 2004
Bent Flyvbjerg
118
SOSIOLOGISK TIDSSKRIFT
At first, I did not know how to respond to such claims, which clearly formed
the conventional wisdom about case-study research. I decided therefore to find
out where the claims come from and whether they are correct. This article con-
tains what I discovered.
The Conventional Wisdom About Case-study research
Looking up ‘case study’ in the
Dictionary of Sociology
as a beginning, I found the
following in full citation:
Case Study. The detailed examination of a single example of a class of phenomena, a case
study cannot provide reliable information about the broader class, but it may be useful in
the preliminary stages of an investigation since it provides hypotheses, which may be
tested systematically with a larger number of cases (Abercrombie et al. 1984, 34).
1
Even if this and other dictionaries later became a bit more subtle, this descrip-
tion is indicative of what was, and often still is, the conventional wisdom about
case-study research. It is not directly wrong, but so oversimplified as to be
grossly misleading. It is correct that the case study is a ‘detailed examination of
a single example,’ but as we will see below it is not true that a case study ‘can-
not provide reliable information about the broader class.’ It is also correct that a
case study
can
be used ‘in the preliminary stages of an investigation’ to generate
hypotheses, but it is misleading to see the case study as a pilot method to be
used only in preparing the real study’s larger surveys, systematic hypotheses
testing, and theory building.
According to the conventional view, a case and a case study cannot be of
value in and of themselves; they need to be linked to hypotheses, following the
well-known hypothetico-deductive model of explanation. Mattei Dogan and
Dominique Pelassy (1990, 121) put it like this: ‘one can validly explain a partic-
ular case only on the basis of general hypotheses. All the rest is uncontrollable,
and so of no use’ (see also Diamond 1996, 6). Similarly, the early Donald
Campbell did not mince words when he relegated single-case studies to the
methodological trash heap:
[S]uch studies have such a total absence of control as to be of almost no scientific value
… Any appearance of absolute knowledge, or intrinsic knowledge about singular iso-
lated objects, is found to be illusory upon analysis … It seems well-nigh unethical at
the present time to allow, as theses or dissertations in education, case studies of this
nature (i.e., involving a single group observed at one time only) (Campbell and
Stanley 1966, 6–7).
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
119
If you read such criticism of a certain methodology enough times, or if you
hear your thesis advisers repeat it, you begin to believe it may be true. This is
what happened to me, and it made me uncertain about case study methodol-
ogy. As I continued my research, however, I found out that Campbell had later
made a 180-degree turn in his views of the case study and had become one of
the strongest proponents of this method. I eventually found, with the help of
Campbell's later works and other works like them, that the problems with the
conventional wisdom about case-study research can be summarized in five
misunderstandings or oversimplifications about the nature of such research:
• Misunderstanding no. 1. Theoretical (context-independent) knowledge is
more valuable than practical (context-dependent) knowledge, and case
studies can only provide the latter.
• Misunderstanding no. 2. One cannot generalize on the basis of an individual
case; therefore, the case study cannot contribute to scientific development.
• Misunderstanding no. 3. The case study is most useful for generating
hypotheses; that is, in the first stage of a total research process, while other
methods are more suitable for hypotheses testing and theory building.
• Misunderstanding no. 4. The case study contains a bias toward verification,
that is, a tendency to confirm the researcher’s preconceived notions.
• Misunderstanding no. 5. It is often difficult to summarize and develop gen-
eral propositions and theories on the basis of specific case studies.
These five misunderstandings indicate that it is theory, reliability, and validity
which are at issue; in other words, the very status of the case study as a scien-
tific method. In what follows, I will focus on these five misunderstandings and
correct them one by one. First, however, I will outline the role of cases in
human learning.
The Role of Cases in Human Learning
In order to understand why the conventional view of case-study research is
problematic, we need to grasp the role of cases and theory in human learning.
Here two points can be made. First, the case study produces the type of con-
text-dependent knowledge which research on learning shows to be necessary
to allow people to develop from rule-based beginners to virtuoso experts. Sec-
Bent Flyvbjerg
120
SOSIOLOGISK TIDSSKRIFT
ond, in the study of human affairs, there appears to exist only context-depend-
ent knowledge, which thus presently rules out the possibility of epistemic theo-
retical construction. The full argument behind these two points can be found in
Flyvbjerg (2001, chapters 2–4). For reasons of space, I can only give an outline
of the argument here. At the outset, however, we can assert that if the two
points are correct, it will have radical consequences for the conventional view
of the case study in research and teaching. This view would then be problem-
atic.
Phenomenological studies of human learning indicate that for adults there
exists a qualitative leap in their learning process from the rule-governed use of
analytical rationality in beginners to the fluid performance of tacit skills in
what Pierre Bourdieu (1977) calls virtuosos and Hubert and Stuart Dreyfus
(1986) true human experts. Here we may note that most people are experts in
a number of everyday social, technical, and intellectual skills like giving a gift,
riding a bicycle, or interpreting images on a television screen, while only few
reach the level of true expertise for more specialized skills like playing chess,
composing a symphony, or flying a fighter jet.
Common to all experts, however, is that they operate on the basis of inti-
mate knowledge of several thousand concrete cases in their areas of expertise.
Context-dependent knowledge and experience are at the very heart of expert
activity. Such knowledge and expertise also lie at the center of the case study as
a research and teaching method; or to put it more generally, still: as a method
of learning. Phenomenological studies of the learning process therefore empha-
size the importance of this and similar methods: it is only because of experience
with cases that one can at all move from being a beginner to being an expert. If
people were exclusively trained in context-independent knowledge and rules,
that is, the kind of knowledge which forms the basis of textbooks and comput-
ers they would remain at the beginner’s level in the learning process. This is the
limitation of analytical rationality: it is inadequate for the best results in the
exercise of a profession, as student, researcher, or practitioner.
In a teaching situation, well chosen case studies can help the student
achieve competence, while context-independent facts and rules will bring the
student just to the beginner’s level. Only few institutions of higher learning
have taken the consequence of this. Harvard University is one of them. Here
both teaching and research in the professional schools are modeled to a wide
extent on the understanding that case knowledge is central to human learning
(Christensen and Hansen eds. 1987; Cragg 1940).
At one stage in my research, I was invited to Harvard to learn about case
methodology 'in action.' During my stay, it became clear to me that if I was
going to aspire at becoming an expert in my field of expertise, and if I wanted
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
121
to be an effective help to my students in their learning processes, I would need
to master case methodology in research and teaching. My stay at Harvard also
became a major step forward in shedding my uncertainties about the conven-
tional wisdom about cases and case studies. At Harvard I found the literature
and people who effectively argued, 'Forget the conventional wisdom, go ahead
and do a case study.' I figured if it is good enough for Harvard, it is good
enough for me, and I suggest others might reason like this, including whole
institutions of learning. There is much to gain, for instance, by transforming the
lecture format still dominant in most universities to one of case learning (Chris-
tensen and Hansen eds. 1987).
It is not that rule-based knowledge should be discounted: it is important in
every area and especially to novices. But to make rule-based knowledge the
highest goal of learning is regressive. There is a need for both approaches. The
highest levels in the learning process, that is, virtuosity and true expertise, are
reached only via a person’s own experiences as practitioner of the relevant
skills. Therefore, beyond using the case method and other experiential methods
for teaching, the best that teachers can do for students in professional programs
is to help them achieve real practical experience; for example, via placement
arrangements, internships, summer jobs, and the like.
For researchers, the closeness of the case study to real-life situations and its
multiple wealth of details are important in two respects. First, it is important for
the development of a nuanced view of reality, including the view that human
behavior cannot be meaningfully understood as simply the rule-governed acts
found at the lowest levels of the learning process, and in much theory. Second,
cases are important for researchers’ own learning processes in developing the
skills needed to do good research. If researchers wish to develop their own
skills to a high level, then concrete, context-dependent experience is just as
central for them as to professionals learning any other specific skills. Concrete
experiences can be achieved via continued proximity to the studied reality and
via feedback from those under study. Great distance to the object of study and
lack of feedback easily lead to a stultified learning process, which in research
can lead to ritual academic blind alleys, where the effect and usefulness of
research becomes unclear and untested. As a research method, the case study
can be an effective remedy against this tendency.
The second main point in connection with the learning process is that there
does not and probably cannot exist predictive theory in social science. Social
science has not succeeded in producing general, context-independent theory
and has thus in the final instance nothing else to offer than concrete, context-
dependent knowledge. And the case study is especially well suited to produce
this knowledge. In his later work, Donald Campbell (1975, 179) arrives at a
Bent Flyvbjerg
122
SOSIOLOGISK TIDSSKRIFT
similar conclusion, explaining how his work has undergone ‘an extreme oscil-
lation away from my earlier dogmatic disparagement of case studies,’ which
was described above. In a logic that in many ways resembles that of the phe-
nomenology of human learning, Campbell now explains:
After all, man is, in his ordinary way, a very competent knower, and qualitative com-
mon-sense knowing is not replaced by quantitative knowing … This is not to say that
such common sense naturalistic observation is objective, dependable, or unbiased.
But it is all that we have. It is the only route to knowledge – noisy, fallible, and
biased though it be (1975, 179, 191).
Campbell is not the only example of a researcher who has altered his views
about the value of the case study. Hans Eysenck (1976, 9), who originally did
not regard the case study as anything other than a method of producing anec-
dotes, later realized that ‘sometimes we simply have to keep our eyes open and
look carefully at individual cases – not in the hope of proving anything, but
rather in the hope of learning something!’ Proof is hard to come by in social sci-
ence because of the absence of ‘hard’ theory, whereas learning is certainly pos-
sibly. More recently, similar views have been expressed by Charles Ragin,
Howard Becker, and their colleagues in explorations of what the case study is
and can be in social inquiry (Ragin and Becker 1992).
As for predictive theory, universals, and scientism, the study of human
affairs is thus at an eternal beginning. In essence, we have only specific cases
and context-dependent knowledge. The first of the five misunderstandings
about the case study – that theoretical (context-independent) knowledge is
more valuable than practical (context-dependent) knowledge, and case studies
can only provide the latter – can therefore be revised as follows:
Predictive theories and universals cannot be found in the study of human affairs.
Concrete, context-dependent knowledge produced by case studies is therefore more
valuable than the vain search for predictive theories and universals.
Cases as ‘Black Swans’
The view that one cannot generalize on the basis of a single case is usually con-
sidered to be devastating to the case study as a scientific method. This second
misunderstanding about the case study is typical among proponents of the nat-
ural science ideal within the social sciences. Yet even researchers who are not
normally associated with this ideal may be found to have this viewpoint.
According to Anthony Giddens, for example,
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
123
Research which is geared primarily to hermeneutic problems may be of generalized
importance in so far as it serves to elucidate the nature of agents’ knowledgeability,
and thereby their reasons for action, across a wide range of action-contexts. Pieces of
ethnographic research like … say, the traditional small-scale community research of
fieldwork anthropology – are not in themselves generalizing studies. But they can
easily become so if carried out in some numbers, so that judgements of their typicality
can justifiably be made (1984, 328).
It is correct that one can generalize in the ways Giddens describes, and that
often this is both appropriate and valuable. But it would be incorrect to assert
that this is the only way to work, just as it is incorrect to conclude that one can-
not generalize from a single case. It depends upon the case one is speaking of,
and how it is chosen. This applies to the natural sciences as well as to the study
of human affairs (see also Platt 1992; Ragin and Becker 1992).
For example, Galileo’s rejection of Aristotle’s law of gravity was not based
upon observations ‘across a wide range,’ and the observations were not ‘carried
out in some numbers.’ The rejection consisted primarily of a conceptual exper-
iment and later on of a practical one. These experiments, with the benefit of
hindsight, are self-evident. Nevertheless, Aristotle’s view of gravity dominated
scientific inquiry for nearly two thousand years before it was falsified. In his
experimental thinking, Galileo reasoned as follows: if two objects with the
same weight are released from the same height at the same time, they will hit
the ground simultaneously, having fallen at the same speed. If the two objects
are then stuck together into one, this object will have double the weight and
will according to the Aristotelian view therefore fall faster than the two individ-
ual objects. This conclusion operated in a counter-intuitive way for Galileo. The
only way to avoid the contradiction was to eliminate weight as a determinant
factor for acceleration in free fall. And that was what Galileo did. Historians of
science continue to discuss whether Galileo actually conducted the famous
experiment from the leaning tower of Pisa, or whether it is simply a myth. In
any event, Galileo’s experimentalism did not involve a large random sample of
trials of objects falling from a wide range of randomly selected heights under
varying wind conditions, and so on, as would be demanded by the thinking of
the early Campbell and Giddens. Rather, it was a matter of a single experiment,
that is, a case study, if any experiment was conducted at all. (On the relation
between case studies, experiments, and generalization, see Lee 1989; Wilson
1987; Bailey 1992; Griffin et al. 1991). Galileo’s view continued to be subjected
to doubt, however, and the Aristotelian view was not finally rejected until half
a century later, with the invention of the air pump. The air pump made it pos-
sible to conduct the ultimate experiment, known by every pupil, whereby a
Bent Flyvbjerg
124
SOSIOLOGISK TIDSSKRIFT
coin or a piece of lead inside a vacuum tube falls with the same speed as a
feather. After this experiment, Aristotle’s view could be maintained no longer.
What is especially worth noting in our discussion, however, is that the matter
was settled by an individual case due to the clever choice of the extremes of
metal and feather. One might call it a critical
case: for if Galileo’s thesis held for
these materials, it could be expected to be valid for all or a large range of mate-
rials. Random and large samples were at no time part of the picture. Most crea-
tive scientists simply do not work this way with this type of problem.
Carefully chosen experiments, cases, and experience were also critical to the
development of the physics of Newton, Einstein, and Bohr, just as the case
study occupied a central place in the works of Darwin, Marx, and Freud. In
social science, too, the strategic choice of case may greatly add to the generaliz-
ability of a case study. In their classical study of the ‘affluent worker,’ John
Goldthorpe et al. (1968, 1969) deliberately looked for a case that was as favora-
ble as possible to the thesis that the working class, having reached middle-class
status, was dissolving into a society without class identity and related conflict
(see also Wieviorka 1992). If the thesis could be proved false in the favorable
case, then it would most likely be false for intermediate cases. Luton, a prosper-
ous industrial center with companies known for high wages and social stability
– fertile ground for middle-class identity – was selected as a case, and through
intensive fieldwork the researchers discovered that even here an autonomous
working-class culture prevailed, lending general credence to the thesis of the
persistence of class identity. Below we will discuss more systematically this type
of strategic sampling.
As regards the relationship between case studies, large samples, and discover-
ies, W. I. B. Beveridge (1951; here quoted from Kuper and Kuper eds. 1985, 95)
observed immediately prior to the breakthrough of the quantitative revolution in
the social sciences: ‘[M]ore discoveries have arisen from intense observation than
from statistics applied to large groups.’ This does not mean that the case study is
always appropriate or relevant as a research method, or that large random sam-
ples are without value (see also the Conclusions below). The choice of method
should clearly depend on the problem under study and its circumstances.
Finally, it should be mentioned that formal generalization, be it on the basis
of large samples or single cases, is considerably overrated as the main source of
scientific progress. Economist Mark Blaug (1980) – a self-declared adherent to
the hypothetico-deductive model of science – has demonstrated that while
economists typically pay lip service to the hypothetico-deductive model and to
generalization, they rarely practice what they preach in actual research. More
generally, Thomas Kuhn has shown that the most important precondition for
science is that researchers possess a wide range of practical skills for carrying
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
125
out scientific work. Generalization is just one of these. In Germanic languages,
the term ‘science’ (
Wissenschaft
) means literally ‘to gain knowledge.’ And formal
generalization is only one of many ways by which people gain and accumulate
knowledge. That knowledge cannot be formally generalized does not mean
that it cannot enter into the collective process of knowledge accumulation in a
given field or in a society. A purely descriptive, phenomenological case study
without any attempt to generalize can certainly be of value in this process and
has often helped cut a path toward scientific innovation. This is not to criticize
attempts at formal generalization, for such attempts are essential and effective
means of scientific development. It is only to emphasize the limitations, which
follows when formal generalization becomes the only legitimate method of sci-
entific inquiry.
The balanced view of the role of the case study in attempting to generalize
by testing hypotheses has been formulated by Eckstein:
[C]omparative and case studies are alternative means to the end of testing
theories, choices between which must be largely governed by arbitrary or
practical, rather than logical, considerations …
[I]t is impossible to take seri-
ously the position that case study is suspect because problem-prone and comparative
study deserving of benefit of doubt because problem-free (1975, 116, 131, emphasis
in original; see also Barzelay 1993, 305 ff.).
Eckstein here uses the term ‘theory’ in its ‘hard’ sense, that is, comprising
explanation and prediction. This makes Eckstein’s dismissal of the view that
case studies cannot be used for testing theories or for generalization stronger
than my own view, which is here restricted to the testing of ‘theory’ in the
‘soft’ sense, that is, testing propositions or hypotheses. Eckstein shows that if
predictive theories would exist in social science, then the case study could be
used to test these theories just as well as other methods.
More recently, John Walton (1992, 129) has similarly observed that ‘case
studies are likely to produce the best theory.’ Eckstein observes, however, the
striking lack of genuine theories within his own field, political science, but
apparently fails to see why this is so:
Aiming at the disciplined application of theories to cases forces one to state theories
more rigorously than might otherwise be done – provided that the application is truly
‘disciplined,’ i.e., designed to show that valid theory compels a particular case inter-
pretation and rules out others. As already stated, this, unfortunately, is rare (if it
occurs at all) in political study. One reason is the lack of compelling theories (1975,
103–4).
Bent Flyvbjerg
126
SOSIOLOGISK TIDSSKRIFT
The case study is ideal for generalizing using the type of test that Karl Popper
called ‘falsification,’ which in social science forms part of critical reflexivity. Fal-
sification is one of the most rigorous tests to which a scientific proposition can
be subjected: if just one observation does not fit with the proposition it is con-
sidered not valid generally and must therefore be either revised or rejected.
Popper himself used the now famous example of, ‘All swans are white,’ and
proposed that just one observation of a single black swan would falsify this
proposition and in this way have general significance and stimulate further
investigations and theory-building. The case study is well suited for identifying
‘black swans’ because of its in-depth approach: what appears to be ‘white’ often
turns out on closer examination to be ‘black.’
Finding black swans was an experience with which I became thoroughly
familiar when I did my first in-depth case study, of urban politics and planning
in the city of Aalborg, Denmark (Flyvbjerg 1998a). For instance, in university I
had been trained in the neoclassical model of 'economic man,' competition,
and free markets. As I dug into what happened behind closed doors in Aalborg,
I found that economic man does not live here. The local business community
were power mongers who were busy negotiating illicit deals with politicians
and administrators on how to block competition and the free market and create
special privileges for themselves. The neoclassical model was effectively falsified
by what I saw in Aalborg. Similarly, the model of representative democracy,
which on the surface of things appears to apply, and by law is supposed to
apply in Aalborg and Denmark, was strangely absent in the deep detail of the
case. Here I found a highly undemocratic, semi-institutionalized way of making
decisions, where leaders of the business community and of the city government
had formed a secret council, which effectively replaced the democratically
elected city council as the place where important decisions on urban politics
and planning were made. My colleagues in third-world nations, who appear to
hold less illusions about markets and democracy than academics in the first
world, get a good laugh when I tell my Aalborg stories. They see that, after all,
we in the North are not so different; we are third-world, too.
We will return to falsification in discussing the fourth misunderstanding of
the case study below. For the present, however, we can correct the second mis-
understanding – that one cannot generalize on the basis of a single case and that
the case study cannot contribute to scientific development – so that it now reads:
One can often generalize on the basis of a single case, and the case study may be cen-
tral to scientific development via generalization as supplement or alternative to other
methods. But formal generalization is overvalued as a source of scientific develop-
ment, whereas ‘the force of example’ is underestimated.
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
127
Strategies for Case Selection
The third misunderstanding about the case study is that the case method is
claimed to be most useful for generating hypotheses in the first steps of a total
research process, while hypothesis-testing and theory-building is best carried
out by other methods later in the process. This misunderstanding derives from
the previous misunderstanding that one cannot generalize on the basis of indi-
vidual cases. And since this misunderstanding has been revised as above, we
can now correct the third misunderstanding as follows:
The case study is useful for both generating and testing of hypotheses but is not lim-
ited to these research activities alone.
Eckstein – contravening the conventional wisdom in this area – goes so far as to
argue that case studies are better for testing hypotheses than for producing
them. Case studies, Eckstein (1975, 80) asserts, ‘are valuable at all stages of the
theory-building process, but most valuable at that stage of theory-building
where least value is generally attached to them: the stage at which candidate
theories are tested.’ Testing of hypotheses relates directly to the question of
'generalizability,' and this in turn relates to the question of case selection.
Here generalizability of case studies can be increased by the strategic selection
of cases (on the selection of cases, further see Ragin 1992; Rosch 1978). When
the objective is to achieve the greatest possible amount of information on a given
problem or phenomenon, a representative case or a random sample may not be
the most appropriate strategy. This is because the typical or average case is often
not the richest in information. Atypical or extreme cases often reveal more infor-
mation because they activate more actors and more basic mechanisms in the sit-
uation studied. In addition, from both an understanding-oriented and an action-
oriented perspective, it is often more important to clarify the deeper causes
behind a given problem and its consequences than to describe the symptoms of
the problem and how frequently they occur. Random samples emphasizing rep-
resentativeness will seldom be able to produce this kind of insight; it is more
appropriate to select some few cases chosen for their validity.
Table A summarizes various forms of sampling. The
extreme case
can be well-
suited for getting a point across in an especially dramatic way, which often
occurs for well-known case studies such as Freud’s ‘Wolf-Man’ and Foucault’s
‘Panopticon.’ In contrast, a
critical case
can be defined as having strategic impor-
tance in relation to the general problem. For example, an occupational medi-
cine clinic wanted to investigate whether people working with organic solvents
suffered brain damage. Instead of choosing a representative sample among all
those enterprises in the clinic’s area that used organic solvents, the clinic strate-
Bent Flyvbjerg
128
SOSIOLOGISK TIDSSKRIFT
gically located a single workplace where all safety regulations on cleanliness,
air quality, and the like, had been fulfilled. This model enterprise became a crit-
ical case: if brain damage related to organic solvents could be found at this par-
ticular facility, then it was likely that the same problem would exist at other
enterprises which were less careful with safety regulations for organic solvents.
Via this type of strategic choice, one can save both time and money in research-
ing a given problem. Another example of critical case selection is the above-
mentioned strategic selection of lead and feather for the test of whether differ-
ent objects fall with equal velocity. The selection of materials provided the pos-
sibility to formulate a generalization characteristic of critical cases, a generaliza-
tion of the sort, ‘If it is valid for this case, it is valid for all (or many) cases.’ In its
negative form, the generalization would be, ‘If it is not valid for this case, then
it is not valid for any (or only few) cases.’
Table A: Strategies for the Selection of Samples and Cases
Type of Selection Purpose
A. Random selection To avoid systematic biases in the sample. The sample’s
size is decisive for generalization.
1. Random sample To achieve a representative sample which allows for
generalization for the entire population.
2. Stratified sample To generalize for specially selected sub-groups within
the population.
B. Information-oriented selection To maximize the utility of information from small
samples and single cases. Cases are selected on the basis
of expectations about their information content.
1. Extreme/deviant cases To obtain information on unusual cases, which can be
especially problematic or especially good in a more clo-
sely defined sense, for instance policy fiascos/successes.
2. Maximum variation cases To obtain information about the significance of various
circumstances for case process and outcome; e.g., three
to four cases which are very different on one dimension:
size, form of organization, location, budget, etc.
3. Critical cases To achieve information which permits logical deductions
of the type, ‘if this is (not) valid for this case, then it
applies to all (no) cases.’
4. Paradigmatic cases To develop a metaphor or establish an exemplar for the
domain which the case concerns.
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
129
How does one identify critical cases? This question is more difficult to answer
than the question of what constitutes a critical case. Locating a critical case
requires experience, and no universal methodological principles exist by
which one can with certainty identify a critical case. The only general advice
that can be given is that when looking for critical cases, it is a good idea to look
for either ‘most likely’ or ‘least likely’ cases, that is, cases which are likely to
either clearly confirm or irrefutably falsify propositions and hypotheses. This is
what I thought I was doing when planning the Aalborg case study mentioned
above (Flyvbjerg 1998a). I was mistaken, however, but to my chagrin I did not
realize this until I was halfway through the research process. Initially, I concei-
ved of Aalborg as a 'most likely' critical case in the following manner: if ratio-
nality and urban planning were weak in the face of power in Aalborg then,
most likely, they would be weak anywhere, at least in Denmark, because in
Aalborg the rational paradigm of planning stood stronger than anywhere else.
Eventually, I realized that this logic was flawed, because my research of local
relations of power showed that one of the most influential 'faces of power' in
Aalborg, the Chamber of Industry and Commerce, was substantially stronger
than their equivalents elsewhere. This had not been clear at the outset because
much less research existed on local power relations than research on local
planning. Therefore, instead of a critical case, unwittingly I ended up with an
extreme case in the sense that both rationality and power were unusually
strong in Aalborg, and my case study became a study of what happens when
strong rationality meets strong power in the arena of urban politics and plan-
ning. But this selection of Aalborg as an extreme case happened to me, I did
not deliberately choose it. It was a frustrating experience when it happened,
especially during those several months from when I realized I did not have a
critical case until it became clear that all was not lost because I had something
else. As a case researcher charting new terrain one must be prepared for such
incidents, I believe.
A model example of a ‘least likely’ case is Robert Michels’s (1962) classical
study of oligarchy in organizations. By choosing a horizontally structured grass-
roots organization with strong democratic ideals – that is, a type of organization
with an especially low probability of being oligarchical – Michels could test the
universality of the oligarchy thesis; that is, ‘If this organization is oligarchic, so
are most others.’ A corresponding model example of a ‘most likely’ case is W. F.
Whyte’s (1943) study of a Boston slum neighborhood, which according to
existing theory should have exhibited social disorganization, but in fact showed
quite the opposite (see also the articles on Whyte’s study in
Journal of Contempo-
rary Ethnography
, vol. 21, no. 1, 1992).
Cases of the ‘most likely’ type are especially well suited to falsification of
Bent Flyvbjerg
130
SOSIOLOGISK TIDSSKRIFT
propositions, while ‘least likely’ cases are most appropriate to tests of verifica-
tion. It should be remarked that a most likely case for one proposition is the
least likely for its negation. For example, Whyte’s slum neighborhood could be
seen as a least likely case for a hypothesis concerning the universality of social
organization. Hence, the identification of a case as most or least likely is linked
to the design of the study, as well as to the specific properties of the actual
case.
A final strategy for the selection of cases is choice of the
paradigmatic case
.
Thomas Kuhn has shown that the basic skills, or background practices, of natu-
ral scientists are organized in terms of ‘exemplars’ the role of which can be
studied by historians of science. Similarly, scholars like Clifford Geertz and
Michel Foucault have often organized their research around specific cultural
paradigms: a paradigm for Geertz lay for instance in the ‘deep play’ of the Bal-
inese cockfight, while for Foucault, European prisons and the ‘Panopticon’ are
examples. Both instances are examples of paradigmatic cases, that is, cases that
highlight more general characteristics of the societies in question. Kuhn has
shown that scientific paradigms cannot be expressed as rules or theories. There
exists no predictive theory for how predictive theory comes about. A scientific
activity is acknowledged or rejected as good science by how close it is to one or
more exemplars; that is, practical prototypes of good scientific work. A paradig-
matic case of how scientists do science is precisely such a prototype. It operates
as a reference point and may function as a focus for the founding of schools of
thought.
As with the critical case, we may ask, ‘How does one identify a paradigmatic
case?’ How does one determine whether a given case has metaphorical and
prototypical value? These questions are even more difficult to answer than for
the critical case, precisely because the paradigmatic case transcends any sort of
rule-based criteria. No standard exists for the paradigmatic case because it sets
the standard. Hubert and Stuart Dreyfus see paradigmatic cases and case studies
as central to human learning. In an interview with Hubert Dreyfus (author’s
files), I therefore asked what constitutes a paradigmatic case and how it can be
identified. Dreyfus replied:
Heidegger says, you recognize a paradigm case because it shines, but I’m afraid that
is not much help. You just have to be intuitive. We all can tell what is a better or
worse case – of a Cézanne painting, for instance. But I can’t think there could be any
rules for deciding what makes Cézannne a paradigmatic modern painter … [I]t is a
big problem in a democratic society where people are supposed to justify what their
intuitions are. In fact, nobody really can justify what their intuition is. So you have
to make up reasons, but it won’t be the real reasons.
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
131
One may agree with Dreyfus that intuition is central to identifying paradig-
matic cases, but one may disagree it is a problem to have to justify one's intui-
tions. Ethnometholodogical studies of scientific practice have demonstrated
that all variety of such practice relies on taken-for-granted procedures that feel
largely intuitive. However, those intuitive decisions are accountable, in the
sense of being sensible to other practitioners or often explicable if not immedi-
ately sensible. That would frequently seem to be the case with the selection of
paradigmatic cases. We may select such cases on the basis of taken-for-granted,
intuitive procedures but are often called upon to account for that selection.
That account must be sensible to other members of the scholarly communities
of which we are part. This may even be argued to be a general characteristic of
scholarship, scientific or otherwise, and not unique to the selection of paradig-
matic social scientific case studies. For instance, it is usually insufficient to jus-
tify an application for research funds by stating that one’s intuition says that a
particular research should be carried out. A research council ideally operates as
society’s test of whether the researcher can account, in collectively acceptable
ways, for his or her intuitive choice, even though intuition may be the real, or
most important, reason why the researcher wants to execute the project.
It is not possible consistently, or even frequently, to determine in advance
whether or not a given case – Geertz' cock fights in Bali, for instance – is para-
digmatic. Besides the strategic choice of case, the execution of the case study
will certainly play a role, as will the reactions to the study by the research com-
munity, the group studied, and, possibly, a broader public. The value of the case
study will depend on the validity claims which researchers can place on their
study, and the status these claims obtain in dialogue with other validity claims
in the discourse to which the study is a contribution. Like other good crafts-
men, all that researchers can do is use their experience and intuition to assess
whether they believe a given case is interesting in a paradigmatic context, and
whether they can provide collectively acceptable reasons for the choice of case.
Finally, concerning considerations of strategy in the choice of cases, it
should be mentioned that the various strategies of selection are not necessarily
mutually exclusive. For example, a case can be simultaneously extreme and
critical, as explained above for Michels’s (1962) study of oligarchy in organiza-
tions. The interpretation of such a case can provide a unique wealth of infor-
mation, because one obtains various perspectives and conclusions on the case
according to whether it is viewed and interpreted as one or another type of
case.
Bent Flyvbjerg
132
SOSIOLOGISK TIDSSKRIFT
Do Case Studies Contain a Subjective Bias?
The fourth of the five misunderstandings about case-study research is that the
method maintains a bias toward verification, understood as a tendency to con-
firm the researcher’s preconceived notions, so that the study therefore becomes
of doubtful scientific value. Diamond (1996, 6), for example, holds this view.
He observes that the case study suffers from what he calls a ‘crippling draw-
back,’ because it does not apply ‘scientific methods,’ by which Diamond under-
stands methods useful for ‘curbing one’s tendencies to stamp one’s pre-existing
interpretations on data as they accumulate.’
Francis Bacon (1853, xlvi) saw this bias toward verification, not simply as a
phenomenon related to the case study in particular, but as a fundamental
human characteristic. Bacon expressed it like this
The human understanding from its peculiar nature, easily supposes a greater degree
of order and equality in things than it really finds. When any proposition has been
laid down, the human understanding forces everything else to add fresh support and
confirmation. It is the peculiar and perpetual error of the human understanding to
be more moved and excited by affirmatives than negatives.
Bacon certainly touches upon a fundamental problem here, a problem, which
all researchers must deal with in some way. Charles Darwin (1958, 123), in his
autobiography, describes the method he developed in order to avoid the bias
toward verification:
I had … during many years followed a golden rule, namely, that whenever a pub-
lished fact, a new observation or thought came across me, which was opposed to my
general results, to make a memorandum of it without fail and at once; for I had
found by experience that such facts and thoughts were far more apt to escape from the
memory than favorable ones. Owing to this habit, very few objections were raised
against my views, which I had not at least noticed and attempted to answer.
The bias toward verification is general, but the alleged deficiency of the case
study and other qualitative methods is that they ostensibly allow more room
for the researcher’s subjective and arbitrary judgment than other methods:
they are often seen as less rigorous than are quantitative, hypothetico-deduc-
tive methods. Even if such criticism is useful, because it sensitizes us to an
important issue, experienced case researchers cannot help but see the critique
as demonstrating a lack of knowledge of what is involved in case-study
research. Donald Campbell and others have shown that the critique is falla-
cious, because the case study has its own rigor, different to be sure, but no less
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
133
strict than the rigor of quantitative methods. The advantage of the case study is
that it can ‘close in’ on real-life situations and test views directly in relation to
phenomena as they unfold in practice.
According to Campbell, Ragin, Geertz, Wieviorka, Flyvbjerg, and others,
researchers who have conducted intensive, in-depth case studies typically
report that their preconceived views, assumptions, concepts, and hypotheses
were wrong and that the case material has compelled them to revise their
hypotheses on essential points. The case study forces upon the researcher the
type of falsifications described above. Ragin (1992, 225) calls this a ‘special fea-
ture of small-
N
research,’ and goes on to explain that criticizing single-case
studies for being inferior to multiple case studies is misguided, since even sin-
gle-case studies ‘are multiple in most research efforts because ideas and evi-
dence may be linked in many different ways.’
Geertz (1995, 119) says about the fieldwork involved in most in-depth case
studies that ‘The Field’ itself is a ‘powerful disciplinary force: assertive,
demanding, even coercive.’ Like any such force, it can be underestimated, but
it cannot be evaded. ‘It is too insistent for that,’ says Geertz. That he is speaking
of a general phenomenon can be seen by simply examining case studies, like
Eckstein (1975), Campbell (1975), and Wieviorka (1992) have done. Campbell
(1975, 181–2) discusses the causes of this phenomenon in the following pas-
sage:
In a case study done by an alert social scientist who has thorough local acquaintance,
the theory he uses to explain the focal difference also generates prediction or expecta-
tions on dozens of other aspects of the culture, and he does not retain the theory
unless most of these are also confirmed … Experiences of social scientists confirm this.
Even in a single qualitative case study, the conscientious social scientist often finds no
explanation that seems satisfactory. Such an outcome would be impossible if the cari-
cature of the single case study … were correct – there would instead be a surfeit of
subjectively compelling explanations.
According to the experiences cited above, it is falsification and not verification,
which characterizes the case study. Moreover, the question of subjectivism and
bias toward verification applies to all methods, not just to the case study and
other qualitative methods. For example, the element of arbitrary subjectivism
will be significant in the choice of categories and variables for a quantitative or
structural investigation, such as a structured questionnaire to be used across a
large sample of cases. And the probability is high that (1) this subjectivism sur-
vives without being thoroughly corrected during the study and (2) that it may
affect the results, quite simply because the quantitative/structural researcher
Bent Flyvbjerg
134
SOSIOLOGISK TIDSSKRIFT
does not get as close to those under study as does the case-study researcher and
therefore is less likely to be corrected by the study objects ‘talking back.’
According to Ragin:
this feature explains why small-N qualitative research is most often at the forefront of
theoretical development. When N’s are large, there are few opportunities for revising
a casing [that is, the delimitation of a case]. At the start of the analysis, cases are
decomposed into variables, and almost the entire dialogue of ideas and evidence
occurs through variables. One implication of this discussion is that to the extent that
large-N research can be sensitized to the diversity and potential heterogeneity of the
cases included in an analysis, large-N research may play a more important part in
the advancement of social science theory (1992, 225; see also Ragin 1987, 164–71).
Here, too, this difference between large samples and single cases can be under-
stood in terms of the phenomenology for human learning discussed above. If
one thus assumes that the goal of the researcher’s work is to understand and
learn about the phenomena being studied, then research is simply a form of
learning. If one assumes that research, like other learning processes, can be
described by the phenomenology for human learning, it then becomes clear
that the most advanced form of understanding is achieved when researchers
place themselves within the context being studied. Only in this way can
researchers understand the viewpoints and the behavior, which characterizes
social actors. Relevant to this point, Giddens states that valid descriptions of
social activities presume that researchers possess those skills necessary to par-
ticipate in the activities described:
I have accepted that it is right to say that the condition of generating descriptions of
social activity is being able in principle to participate in it. It involves ‘mutual knowl-
edge,’ shared by observer and participants whose action constitutes and reconstitutes
the social world (1982, 15).
From this point of view, the proximity to reality, which the case study entails,
and the learning process which it generates for the researcher will often consti-
tute a prerequisite for advanced understanding. In this context, one begins to
understand Beveridge’s conclusion that there are more discoveries stemming
from the type of intense observation made possible by the case study than from
statistics applied to large groups. With the point of departure in the learning
process, we understand why the researcher who conducts a case study often
ends up by casting off preconceived notions and theories. Such activity is quite
simply a central element in learning and in the achievement of new insight.
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
135
More simple forms of understanding must yield to more complex ones as one
moves from beginner to expert.
On this basis, the fourth misunderstanding – that the case study supposedly
contains a bias toward verification, understood as a tendency to confirm the
researcher’s preconceived ideas – is revised as follows:
The case study contains no greater bias toward verification of the researcher’s precon-
ceived notions than other methods of inquiry. On the contrary, experience indicates
that the case study contains a greater bias toward falsification of preconceived notions
than toward verification.
The Irreducible Quality of Good Case Narratives
Case studies often contain a substantial element of narrative. Good narratives
typically approach the complexities and contradictions of real life. Accordingly,
such narratives may be difficult or impossible to summarize into neat scientific
formulae, general propositions, and theories (Benhabib 1990, Rouse 1990,
Roth 1989, White 1990, Mitchell and Charmaz 1996). This tends to be seen by
critics of the case study as a drawback. To the case-study researcher, however, a
particularly ‘thick’ and hard-to-summarize narrative is not a problem. Rather,
it is often a sign that the study has uncovered a particularly rich problematic.
The question, therefore, is whether the summarizing and generalization, which
the critics see as an ideal, is always desirable. Nietzsche (1974, 335 [§ 373]) is
clear in his answer to this question. ‘Above all,’ he says about doing science,
‘one should not wish to divest existence of its
rich ambiguity
’ (emphasis in origi-
nal).
In doing the Aalborg study, I tried to capture the rich ambiguity of politics
and planning in a modern democracy. I did this by focusing in-depth on the
particular events that made up the case and on the minutiae that made up the
events. Working with minutiae is time consuming, and I must concede that
during the several years when I was toiling in the archives, doing interviews,
making observations, talking with my informants, writing, and getting feed-
back, a nagging question kept resurfacing in my mind. This is a question bound
to haunt many carrying out in-depth, dense case studies: 'Who will want to
learn about a case like this, and in this kind of
detail
?' I wanted the Aalborg case
study to be particularly dense because I wished to test the thesis that the most
interesting phenomena in politics and planning, and those of most general
import, would be found in the most minute and most concrete of details. Or to
put the matter differently, I wanted to see whether the dualisms general-spe-
cific and abstract-concrete would metamorphose and vanish if I went into suffi-
Bent Flyvbjerg
136
SOSIOLOGISK TIDSSKRIFT
ciently deep detail. Richard Rorty has perceptively observed that the way to re-
enchant the world is to stick to the concrete. Nietzsche similarly advocates a
focus on 'little things.' Both Rorty and Nietzsche seem right to me. I saw the
Aalborg case as being made up of the type of concrete, little things they talk
about. Indeed, I saw the case itself as such a thing, what Nietzsche calls a dis-
creet and apparently insignificant truth, which, when closely examined, would
reveal itself to be pregnant with paradigms, metaphors, and general signifi-
cance. That was my thesis, but theses can be wrong and case studies may fail. I
was genuinely relieved when, eventually, the strategy of focusing on minutiae
proved to be worth the effort.
Lisa Peattie (2001, 260) explicitly warns against summarizing dense case
studies: 'It is simply that the very value of the case study, the contextual and
interpenetrating nature of forces, is lost when one tries to sum up in large and
mutually exclusive concepts.' The dense case study, according to Peattie is more
useful for the practitioner and more interesting for social theory than either
factual 'findings' or the high-level generalizations of theory.
The opposite of summing up and 'closing' a case study is to keep it open.
Here I have found the following two strategies to work particularly well in
ensuring such openness. First, when writing up a case study, I demur from the
role of omniscient narrator and summarizer. Instead, I tell the story in its diver-
sity, allowing the story to unfold from the many-sided, complex, and some-
times conflicting stories that the actors in the case have told me. Second, I
avoid linking the case with the theories of any one academic specialization.
Instead I relate the case to broader philosophical positions that cut across spe-
cializations. In this way I try to leave scope for readers of different backgrounds
to make different interpretations and draw diverse conclusions regarding the
question of what the case is a case of. The goal is not to make the case study be
all things to all people. The goal is to allow the study to be different things to
different people. I try to achieve this by describing the case with so many facets
– like life itself – that different readers may be attracted, or repelled, by different
things in the case. Readers are not pointed down any one theoretical path or
given the impression that truth might lie at the end of such a path. Readers will
have to discover their own path and truth inside the case. Thus, in addition to
the interpretations of case actors and case narrators, readers are invited to
decide the meaning of the case and to interrogate actors' and narrators' inter-
pretations in order to answer that categorical question of any case study: ‘What
is this case a case of?’
Case stories written like this can neither be briefly recounted nor summa-
rized in a few main results. The case story is itself the result. It is a 'virtual real-
ity,' so to speak. For the reader willing to enter this reality and explore it inside
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
137
and out the payback is meant to be a sensitivity to the issues at hand that can-
not be obtained from theory. Students can safely be let loose in this kind of
reality, which provides a useful training ground with insights into real-life prac-
tices that academic teaching often does not provide.
If we return briefly to the phenomenology for human learning we may
understand why summarizing case studies is not always useful and may some-
times be counterproductive. Knowledge at the beginner’s level consists pre-
cisely in the reduced formulas which characterize theories, while true expertise
is based on intimate experience with thousands of individual cases and on the
ability to discriminate between situations, with all their nuances of difference,
without distilling them into formulas or standard cases. The problem is analo-
gous to the inability of heuristic, computer-based expert systems to approach
the level of virtuoso human experts, even when the systems are compared with
the experts who have conceived the rules upon which these systems operate.
This is because the experts do not use rules but operate on the basis of detailed
case-experience. This is
real
expertise. The rules for expert systems are formu-
lated only because the systems require it; rules are characteristic of expert
sys-
tems
, but not of real human
experts
.
In the same way, one might say that the rule formulation which takes place
when researchers summarize their work into theories is characteristic of the
culture of research, of researchers, and of theoretical activity, but such rules
are not necessarily part of the studied reality constituted by Bourdieu’s (1977,
8, 15) ‘virtuoso social actors.’ Something essential may be lost by this summa-
rizing – namely the possibility to understand virtuoso social acting, which, as
Bourdieu has shown, cannot be distilled into theoretical formulae – and it is
precisely their fear of losing this ‘something,’ which makes case researchers
cautious about summarizing their studies. Case researchers thus tend to be
skeptical about erasing phenomenological detail in favor of conceptual clo-
sure.
Ludwig Wittgenstein shared this skepticism. According to Gasking and Jack-
son, Wittgenstein used the following metaphor when he described his use of
the case study approach in philosophy:
In teaching you philosophy I’m like a guide showing you how to find your way
round London. I have to take you through the city from north to south, from east to
west, from Euston to the embankment and from Piccadilly to the Marble Arch. After
I have taken you many journeys through the city, in all sorts of directions, we shall
have passed through any given street a number of times – each time traversing the
street as part of a different journey. At the end of this you will know London; you will
be able to find your way about like a born Londoner. Of course, a good guide will
Bent Flyvbjerg
138
SOSIOLOGISK TIDSSKRIFT
take you through the more important streets more often than he takes you down side
streets; a bad guide will do the opposite. In philosophy I’m a rather bad guide (1967,
51).
This approach implies exploring phenomena firsthand instead of reading maps
of them. Actual practices are studied before their rules, and one is not satisfied
by learning only about those parts of practices that are open to public scrutiny;
what Erving Goffman (1963) calls the ‘backstage’ of social phenomena must be
investigated, too, like the side streets which Wittgenstein talks about.
With respect to intervention in social and political affairs, Abbott (1992, 79)
has rightly observed that a social science expressed in terms of typical case nar-
ratives would provide ‘far better access for policy intervention than the present
social science of variables.’ MacIntyre (1984, 216) similarly says, ‘I can only
answer the question ‘What am I to do?’ if I can answer the prior question ‘Of
what story or stories do I find myself a part?’’ Several observers have noted that
narrative is an ancient method and perhaps our most fundamental form for
making sense of experience. (Novak 1975, 175; Mattingly 1991, 237; see also
Abbott 1992, Arendt 1958, Carr 1986, Ricoeur 1984, Fehn et al. 1992, Rasmus-
sen 1995, and Bal 1997).
To MacIntyre (1984, 214, 216), the human being is a ‘story-telling animal,’
and the notion of a history is as fundamental a notion as the notion of an action.
In a similar vein, Mattingly (1991, 237) points out that narratives not only give
meaningful form to experiences we have already lived through. They also pro-
vide us a forward glance, helping us to anticipate situations even before we
encounter them, allowing us to envision alternative futures. Narrative inquiries
do not – indeed, cannot – start from explicit theoretical assumptions. Instead,
they begin with an interest in a particular phenomenon that is best understood
narratively. Narrative inquiries then develop descriptions and interpretations of
the phenomenon from the perspective of participants, researchers, and others.
Labov (1966, 37-9) writes that when a good narrative is over ‘it should be
unthinkable for a bystander to say, ‘So what’?’ Every good narrator is continu-
ally warding off this question. A narrative that lacks a moral that can be inde-
pendently and briefly stated, is not necessarily pointless. And a narrative is not
successful just because it allows a brief moral. A successful narrative does not
allow the question to be raised at all. The narrative has already supplied the
answer before the question is asked. The narrative itself is the answer (Neha-
mas 1985, 163–64).
A reformulation of the fifth misunderstanding, which states that it is often
difficult to summarize specific case studies into general propositions and theo-
ries, thus reads as follows:
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT
139
It is correct that summarizing case studies is often difficult, especially as concerns case
process. It is less correct as regards case outcomes. The problems in summarizing case
studies, however, are due more often to the properties of the reality studied than to the
case study as a research method. Often it is not desirable to summarize and generalize
case studies. Good studies should be read as narratives in their entirety.
It must again be emphasized that despite the difficulty or undesirability in sum-
marizing case studies, the case study method in general can certainly contribute
to the cumulative development of knowledge; for example, in using the princi-
ples to test propositions described above under the second and third misunder-
standings.
Conclusions
Today, when students and colleagues present me with the conventional wis-
dom about case-study research – for instance that one cannot generalize on the
basis of a single case or that case studies are arbitrary and subjective – I know
what to answer. By and large, the conventional wisdom is wrong or mislead-
ing. For the reasons given above, the case study is a necessary and sufficient
method for certain important research tasks in the social sciences, and it is a
method that holds up well when compared to other methods in the gamut of
social science research methodology.
When students ask me for reference to a good book on how to carry out
case study research in practice, I usually recommend Robert Stake's (1995)
The
Art of Case Study Research
. If the students are intellectually curious, I suggest they
also read Charles Ragin and Howard Becker's (1992)
What is a Case?
Both books
are first-rate and fit well with the views presented in this article.
Let me reiterate, however, that the revision of the five misunderstandings
about case study research described above, should not be interpreted as a rejec-
tion of research which focuses on large random samples or entire populations;
for example, questionnaire surveys with related quantitative analysis. This type
of research is also essential for the development of social science; for example,
in understanding the degree to which certain phenomena are present in a
given group or how they vary across cases. The advantage of large samples is
breadth, while their problem is one of depth. For the case study, the situation is
the reverse. Both approaches are necessary for a sound development of social
science.
Here as elsewhere, the sharp separation often seen in the literature
between qualitative and quantitative methods is a spurious one. The separa-
tion is an unfortunate artifact of power relations and time constraints in grad-
Bent Flyvbjerg
140
SOSIOLOGISK TIDSSKRIFT
uate training; it is not a logical consequence of what graduates and scholars
need to know to do their studies and do them well. In my interpretation, good
social science is opposed to an either/or and stands for a both/and on the ques-
tion of qualitative versus quantitative methods. Good social science is prob-
lem-driven and not methodology-driven, in the sense that it employs those
methods which for a given problematic best help answer the research ques-
tions at hand. More often than not, a combination of qualitative and quantita-
tive methods will do the task best. Fortunately, there seems currently to be a
general relaxation in the old and unproductive separation of qualitative and
quantitative methods.
This being said, it should nevertheless be added that the balance between
case studies and large samples is currently biased in favor of the latter in social
science, so biased that it puts case studies at a disadvantage within most disci-
plines. In this connection, it is worth repeating the insight of Thomas Kuhn
that a discipline without a large number of thoroughly executed case studies is
a discipline without systematic production of exemplars, and that a discipline
without exemplars is an ineffective one. In social science more good case stud-
ies could help remedy this situation.
Notes
1 The quote is from the original first edition of the
dictionary (1984). In the third edition (1994), a
second paragraph has been added about the case
study. The entry is still highly unbalanced,
however, and still promotes the mistaken view that
the case study is hardly a methodology in its own
right, but is best seen as subordinate to investiga-
tions of larger samples
Literature
Abbott, Andrew (1992) ‘What Do Cases Do? Some
Notes on Activity in Sociological Analysis’, in
Charles C. Ragin and Howard S. Becker (eds), What
is a Case? Exploring the Foundations of Social Inquiry.
Cambridge: Cambridge University Press, 1992, pp.
53–82.
Abercrombie, Nicolas, Hill, Stephen and Turner, Bryan
S. (1984) Dictionary of Sociology. Harmondsworth:
Penguin, 1984.
Arendt, Hannah (1958) The Human Condition. Chicago:
The University of Chicago Press, 1958.
Bacon, Francis (1853) Novum Organum, in The Physical
and Metaphysical Works of Lord Bacon, book 1. Lon-
don: H. G. Bohn.
Bailey, Mary Timney (1992) ‘Do Physicists Use Case
Studies? Thoughts on Public Administration
Research’, Public Administration Review, 52(1):47–54.
Bal, Mieke (1997) Narratology: Introduction to the Theory
of Narrative. Toronto: University of Toronto Press,
second edition.
Barzelay, Michael (1993) ‘The Single Case Study as
Intellectually Ambitious Inquiry’, Journal of Public
Administration Research and Theory, 3(3):305–318.
Benhabib, Seyla (1990) ‘Hannah Arendt and the
Five Misunderstandings About Case-Study Research
SOSIOLOGISK TIDSSKRIFT 141
Redemptive Power of Narrative’, Social Research,
57(1):167–196.
Beveridge, W. I. B. (1951) The Art of Scientific Investiga-
tion. London: William Heinemann.
Blaug, Mark (1980) The Methodology of Economics: Or
How Economists Explain. Cambridge: Cambridge Uni-
versity Press.
Bourdieu, Pierre (1977) Outline of a Theory of Practice.
Cambridge: Cambridge University Press.
Campbell, Donald T. (1975) ‘Degrees of Freedom and
the Case Study’, Comparative Political Studies,
8(1):178–191.
Campbell, Donald T. and Stanley, J. C. (1966) Experi-
mental and Quasi-Experimental Designs for Research.
Chicago: Rand McNally.
Carr, D. (1986) Time, Narrative, and History. Blooming-
ton: Indiana University Press.
Christensen, C. Roland with Hansen, Abby J. (eds)
(1987) Teaching and the Case Method. Boston: Har-
vard Business School Press.
Cragg, Charles I. (1940) ‘Because Wisdom Can’t be
Told’, Harvard Alumni Bulletin. Harvard Business
School Reprint 451-005, pp. 1–6.
Darwin, Charles (1958) The Autobiography of Charles
Darwin. New York: Norton.
Diamond, Jared (1996) ‘The Roots of Radicalism’, The
New York Review of Books, 14 November, pp. 4–6.
Dogan, Mattei and Pelassy, Dominique (1990) How To
Compare Nations: Strategies in Comparative Politics.
Chatham: Chatham House, second edition.
Dreyfus, Hubert and Stuart with Athanasiou, Thomas
(1986) Mind Over Machine: The Power of Human Intui-
tion and Expertise in the Era of the Computer. New
York: Free Press.
Eckstein, Harry (1975) ‘Case Study and Theory in
Political Science’, in Greenstein, Fred J. and Polsby,
Nelson W. (eds), Handbook of Political Science, vol. 7.
Reading, MA.: Addison-Wesley, pp. 79–137.
Eysenck, H. J. (1976) ‘Introduction’, in Eysenck (ed.),
Case Studies in Behaviour Therapy. London: Routledge
and Kegan Paul.
Fehn, Ann, Hoestery, Ingeborg and Tatar, Maria (eds)
(1992) Neverending Stories: Toward a Critical Narratol-
ogy. Princeton: Princeton University Press.
Flyvbjerg, Bent (1998a) Rationality and Power: Democracy
in Practice. Chicago: The University of Chicago Press.
Flyvbjerg, Bent (1998b) ‘Habermas and Foucault:
Thinkers for Civil Society?’, British Journal of Sociol-
ogy, 49(2):208–233.
Flyvbjerg, Bent (2001) Making Social Science Matter: Why
Social Inquiry Fails and How It Can Succeed Again.
Cambridge: Cambridge University Press.
Flyvbjerg, Bent (2002) 'Bringing Power to Planning
Research: One Researcher’s Praxis Story'. Journal of
Planning Education and Research, 21(4):353–366.
Gasking, D. A. T. and Jackson, A. C. ‘Wittgenstein as a
Teacher’, in Fann, K. T. (ed.) (1967) Ludwig Wittgen-
stein: The Man and His Philosophy. Sussex: Harvester
Press, pp. 49–55.
Geertz, Clifford (1995) After the Fact: Two Countries, Four
Decades, One Anthropologist. Cambridge, MA: Har-
vard University Press.
Giddens, Anthony (1982) Profiles and Critiques in Social
TheoryK. Berkeley: University of California Press.
Giddens, Anthony (1984) The Constitution of Society:
Outline of the Theory of Structuration. Cambridge: Pol-
ity Press.
Goffman, Erving (1963) Behavior in Public Places: Notes
on the Social Organization of Gatherings. New York:
The Free Press.
Goldthorpe, John, Lockwood, David, Beckhofer,
Franck and Platt, Jennifer (1968–9) The Affluent
Worker, vols. 1–3. Cambridge: Cambridge University
Press.
Griffin, Larry J., Botsko, Christopher, Wahl, Ana-
Maria and Isaac, Larry W. (1991) ‘Theoretical Gen-
erality, Case Particularity: Qualitative Comparative
Analysis of Trade Union Growth and Decline’, in
Ragin, Charles C. (ed.), Issues and Alternatives in
Comparative Social Research. Leiden: E. J. Brill, pp.
110–136.
Kuper, Adam and Kuper, Jessica (eds) (1985) The
Social Science Encyclopedia. London: Routledge and
Kegan Paul.
Labov, William and Waletzky, Joshua (1966) ‘Narra-
tive Analysis: Oral Versions of Personal Experience’,
in Essays on the Verbal and Visual Arts: Proceedings of
the American Ethnological Society. Seattle: American
Ethnological Society, pp. 12–44.
Lee, Allen S. (1989) ‘Case Studies as Natural Experi-
ments’, Human Relations, 42(2):117–137.
MacIntyre, Alasdair (1984) After Virtue: A Study in
Bent Flyvbjerg
142 SOSIOLOGISK TIDSSKRIFT
Moral Theory. Notre Dame: University of Notre
Dame Press, second edition.
Mattingly, Cheryl (1991) ‘Narrative Reflections on
Practical Actions: Two Learning Experiments in
Reflective Storytelling’, in Schön, Donald A. (ed.),
The Reflective Turn: Case Studies in and on Educational
Practice. New York: Teachers College Press, pp. 235–
257.
Michels, Robert (1962) Political Parties: A Study of the
Oligarchical Tendencies of Modern Democracy. New
York: Collier Books.
Mitchell, Richard G. Jr. and Charmaz, Kathy (1996)
‘Telling Tales, Writing Stories: Postmodernist
Visions and Realist Images in Ethnographic Writ-
ing’, Journal of Contemporary Ethnography,
25(1):144–166.
Nehamas, Alexander (1985) Nietzsche: Life as Literature .
Cambridge, MA: Harvard University Press, 1985.
Novak, M. (1975) ‘“Story” and Experience’, in Wig-
gins, J. B. (ed.), Religion as Story. Lanham, MD: Uni-
versity Press of America.
Peattie, Lisa (2001) 'Theorizing Planning: Some Com-
ments on Flyvbjerg's Rationality and Power', Interna-
tional Planning Studies, 6(3):257–262.
Platt, Jennifer (1992) ‘“Case Study” in American
Methodological Thought’, Current Sociology, 40(1):
17–48.
Ragin, Charles C. (1987) The Comparative Method: Mov-
ing Beyond Qualitative and Quantitative Strategies. Ber-
keley: University of California Press.
Ragin, Charles C. (1992) ‘“Casing” and the Process of
Social Inquiry’, in Ragin, Charles C. and Becker,
Howard S. (eds), What is a Case? Exploring the Foun-
dations of Social Inquiry. Cambridge: Cambridge Uni-
versity Press, pp. 217–226.
Ragin, Charles C. and Becker, Howard S. (eds) (1992)
What is a Case? Exploring the Foundations of Social
Inquiry. Cambridge: Cambridge University Press.
Rasmussen, David (1995) ‘Rethinking Subjectivity:
Narrative Identity and the Self’, Philosophy and
Social Criticism, 21(5–6):159–172.
Ricoeur, Paul (1984) Time and Narrative. Chicago: The
University of Chicago Press.
Rosch, Eleanor (1978) ‘Principles of Categorization’, in
Rosch, Eleanor and Lloyd, Barbara B. (eds), Cogni-
tion and Categorization. Killsdale, NJ: Lawrence Erl-
baum, pp. 27–48.
Roth, Paul A. (1989) ‘How Narratives Explain’, Social
Research, 56(2):449–478.
Rouse, Joseph (1990) ‘The Narrative Reconstruction
of Science’, Inquiry, 33(2):179–196.
Stake, Robert E. (1995) The Art of Case Study Research.
Thousand Oaks: Sage.
Walton, John ‘Making the Theoretical Case’, in Ragin,
Charles C. and Becker, Howard S. (eds), What is a
Case? Exploring the Foundations of Social Inquiry. Cam-
bridge: Cambridge University Press, pp. 121–137.
White, Hayden (1990) The Content of the Form: Narrative
Discourse and Historical Representation. Baltimore: The
Johns Hopkins University Press.
Whyte, W. F. (1943) Street Corner Society: The Social
Structure of an Italian Slum. Chicago: The University
of Chicago Press.
Wieviorka, Michel (1992) ‘Case Studies: History or
Sociology?’ in Ragin, Charles C. and Becker,
Howard S. (eds), What is a Case? Exploring the Foun-
dations of Social Inquiry. Cambridge: Cambridge Uni-
versity Press, pp. 159–172.
Wilson, Barbara (1987) ‘Single-case Experimental
Designs in Neuro-Psychological Rehabilitation’,
Journal of Clinical and Experimental Neuropsychology,
9(5):527–544.
