Volume 13(5): 677699

ISSN 13505084
Copyright 2006 SAGE
(London, Thousand Oaks, CA
and New Delhi)

Beyond Measuring the Human

Resources Management
Organizational Performance Link:
Applying Critical Realist
Meta-Theory

articles

Anthony Hesketh
Lancaster University, UK

Steve Fleetwood
Lancaster University, UK

Abstract. One response to the possible outsourcing of the human

resource (HR) function is to turn to science and seek to demonstrate an
empirical association between HR practices and increased organizational
performance. This paper critically examines the shortcomings of the
scientific approach by first, reviewing the three distinctive versions of
research on the Human Resources Management (HRM)Performance link
in an attempt to demonstrate their commitment to a common scientific
meta-theory. Second, we use critical realism to demonstrate: (1) that
theoretical underdevelopment and lack of explanatory power are encouraged by the use of an inappropriate scientific meta-theory; (2) the possibility of meta-theorizing the causal connection between HRM and
performance without seeking statistical associations; and (3) how all this is
in-keeping with Institutional theory. Finally, all of this is achieved by
inserting evidence from interviews with HR professionals to demonstrate
not only that they are sceptical of the scientific approach, but also that
they hold views of the world not dissimilar to the critical realist approach
we advocate. Key words. closed systems; event regularities; methodology; ontology; powers; prediction; reflexive performance; science; tendencies; tendential prediction

DOI: 10.1177/1350508406067009

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

http://org.sagepub.com

Organization 13(5)
Articles
Something is definitely going on in human resources management
(HRM). Recent debate over how best to identify and measure the key
human resource practices that generate increased organizational performance has been raging, with contributions from government, practitioner
bodies, the quality press, and consulting houses.1 Never before have
people been our most important asset as they are now. And yet despite
this, the outsourcing of HR functions enjoyed a record year in 2005 with
figures projected to rise even more sharply (Equa Terra, 2005; Davis,
2003; Nelson Hall, 2005).
Faced with the threat of being outsourced, many HR professionals have
pinned their hopes on convincing senior executives that the HR function
adds, as opposed to saps, value. Proof of this comes in the form of
research carried out by academics2 (and more recently HR consultants)
seeking to empirically demonstrate the contribution made by HRM to
organizational performancehenceforth referred to as the HRM
Performance (HRMP) link. There are, however, under-estimated and
unrecognized problems for both HR professionals and empirical
researchers. Let us briefly rehearse (some of) these problems, before
subsequently addressing them in depth.
First, empirical evidence for the existence of an HRMP link is inconclusive. Second, the non-existence of an empirical association between
HRM practices and organizational performance does not entail the nonexistence of some kind of causal connection between them. It could be
the case that a causal connection exists, but the nature of this causality is
more complex than can be captured via the usual statistical techniques.
If, as we believe, this is the case, then concluding that there is no causal
connection, simply because there is no empirical association is misleading in the extreme. Third, even if convincing empirical evidence for the
existence of an HRMP link is eventually found, a statistical association
in, and of itself, constitutes neither a theory nor an explanation.
Although some of the more sophisticated empirical researchers (Guest,
1997, 2001; Boselie et al., 2005) are aware of the lack of theory, even if
they under-estimate the scale of the problem, the lack of explanation
remains totally unrecognized.
These empirical and theoretical problems are caused, at least in part,
by a reluctance to even consider the possibility that they might have their
roots in meta-theoretical problems.3 That is, these empirical and theoretical problems might be caused by the use of what is often referred to as a
scientific approach (Boudreau and Ramstad, 1999: 343; Murphy and
Zandvakili, 2000: 93; Brown, 2004: 40) with its commitment to empirical
research techniques. Indeed, empirical researchers have tended to ignore
meta-theory altogether, strongly implying that theoretical problems will
resolve themselves via more, and better, empirical work (Becker and
Gerhart, 1996; Boudreau and Ramstad, 1999; Guest, 1997, 2001;
McMahan et al., 1999). We disagree, holding the view that although

678

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
empirical research might be a useful starting point for understanding the
HRMP link, it should not be treated as an explanatory end game.
Unfortunately, however, a great deal of this research does not proceed
beyond attempts to find an empirical association between HRM practices
and organizational performance.
Unhappy with the current state of affairs, we decided to adopt two
strands of inquiry. On the one hand, we decided to reflect deeply upon
the current meta-theoretical presuppositions underpinning empirical
research on the HRMP link, to tease out exactly where the problems
might lie. On the other hand, we decided to undertake empirical work to
ascertain what kind of meta-theoretical approaches HR managers actually
think are appropriate for understanding the relationship between HRM
practices and the performance of their organizations and some of the
institutional forces at work shaping these decisions.
The paper has three main sections. First, we briefly review the three
distinctive versions of research on the HRMP link in an attempt to
demonstrate their commitment to a common scientific meta-theory.
Second, we use critical realism to demonstrate: (1) that theoretical
underdevelopment and lack of explanatory power are encouraged by the
use of an inappropriate scientific meta-theory; (2) the possibility of
meta-theorizing the causal connection between HRM and performance
without seeking statistical associations; and (3) how our analysis is in
keeping with Institutional theory. In all three sections, we provide
evidence from interviews with HR professionals to demonstrate not only
that they are sceptical of the scientific approach, but also that they
(largely) hold views of the world not dissimilar to the critical realist
approach we advocate.4
Our empirical work is primarily based on over 70 interviews with HR
directors and senior managers in over 30 organizations in the UK, Europe
and the USA (Fleetwood and Hesketh, 2006). As will become clear below,
central to our work on the HRMP link is the critically realist notion of a
structured or relational ontology [for an example of this in organizational
research settings, see Fairclough (2005)]. Access to the deeper realm of
underlying causal mechanisms, structures, powers and tendencies
requires a more discursively reflexive approach to data collection [e.g.
Alvesson and Skolberg (2000)]. Our interviews followed a similar pattern, asking respondents to: (1) summarize their career histories and
current roles and (2) articulate what they understand about the nature of
the link between HRM techniques and organizational performance; followed by (3) an exploration of the foundations on which their assumptions of how HRM influences organizational performance rest (Fleetwood
and Hesketh, 2006).

Black Boxes, Russian Dolls and Rubiks Cubes

Research on the HRMP link has a long history. Although authors such as
Woodward (1965) addressed the issue four decades ago, many point to

679

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
the seminal papers of Osterman (1994), Pfeffer (1994) and especially
Huselid (1995) as providing the solid foundations (cf. Paauwe, 2004:
534). This universalistic mode (Delery and Doty, 1996) proposes the
adoption of best practice or high performance work systems which are
not only universally applicable and successful, but should also be
evaluated or measured in terms of top-level organizational performance
(e.g. market-based measures such as Tobins q or accounting measures
such as gross rate of return on capital or GRATE), for which various
accounting analyses can be imported and applied to the field of HRM.
Prescribed HR practices differ, ranging from Pfeffers (1995) 13 interrelated practices for managing people, through Huselids (1995) high
performance work systems, to more recent attempts such as that of
Ellinger et al. (2002) to relate learning concepts to organizational performance. No matter what variation we find in the independent variables,
the dependent variables are nearly always the same, namely variations in
the publicly reported top-level financial performance of organizations.
Configurational theories differ from the assumptions held by universalistic researchers insofar as an internal consistency between HRM practices is viewed as essential to unlocking enhanced organizational
performance (e.g. Delery and Doty, 1996). In many ways, the configurational approach is an extension of contingency theory, which advocates
the utilization of different types of HR practices in accordance with the
different strategic positions adopted by organizations (e.g. Schuler and
Jackson, 1987). Nevertheless, the dependent variable of actual financial
performance is seen as the yardstick against which HR techniques,
whether universalistic, configurational, contingent, vertical or horizontally integrated, etc., and their implementation are to be measured.
Whereas the universalistic mode is often described as the black box
approach to examining the HRMP link because of its examination of the
link between various inputs (HR policies and practices) and outputs
(profits, financial performance), the contingency modes contribution to
the HRMP link is perhaps best represented by the metaphor of a Russian
doll in which the outer doll of enhanced organizational performance is
contingent on the fitting together of the inner dolls of different HR
practices. The configuration mode is illustrated by the Rubiks cube
analogy wherein a shift in the strategic orientation of an organization
(Schuler and Jackson use the examples of innovation, qualityenhancement, cost reduction, etc.) requires a concomitant shift in emphasis on specific (if sometimes wholly different) HRM practices.5

Common Meta-Theory and Common Problems

The slight variations in these three approaches, and variations in the
dependent and independent variables used, should not blind us to one
very important point: they all adopt the same scientific approach,
which can be styled as follows: HR (or high performance work) practices
and organizational performance are quantified using appropriate metrics

680

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
and measurements, generating empirical proxies as data. Various statistical techniques (e.g. regression, analysis of variance, correlation, structural equation modelling, factor analysis and so on) are then employed on
these data to test hypotheses, typically, to the effect that certain (bundles
of) HRM practices, under certain conditions, are associated with organizational performance.6
Whatever merits the scientific approach possesses, it is vital that we
explicitly acknowledge one fundamental limitation: an empirical association between HRM practices and organizational performance does not
constitute an explanation of this association. A regression equation, for
example, even one with good predictive power, has no explanatory
power. Variations in the proxy measurements of HR do not explain
variations in the financial performance of organizations.
Although some researchers are perfectly aware of these limitations, and
perhaps recognize the need for this scientific approach to be used
alongside other methods (methodological pluralism, methodological triangulation, and so on), others may not be aware. Kinnie et al. (2005: 19),
for example, refer to the explanatory power of the models apparently
without even considering the possibility that regression models do not
possess such power. This is obviously a contentious issue, so let us
elaborate a little.
In the lingua franca of statistics, to explain is to use the independent
variables (often misleadingly referred to as explanatory variables) to
account for some proportion of the variance in the dependent variable.
This is, of course, a very specific way of using the term explanation and
in this lingua franca it is quite unobjectionable. But outside this lingua
franca, it is problematic because it does not actually explain why the
independent variables account for some proportion of the variance in the
dependent variable. However useful it might be to know that X1, X2, and
X3 explain 75% of the variance in Y, neither the equation itself, nor the
empirical data that constitute the variables, give us any idea why this is
the case. There is more to a bone fide explanation than explaining some
proportion of the variance in a dependent variable (cf. Lipton, 1993;
Ruben, 1992). There are, however, four more practical objections to the
exclusive use of the scientific approach in current research into the
HRMP link. Let us deal with each of these in turn.
First, as is now recognized by some empirical researchers, the data are
inconclusive. Rather than provide a long list of references, we refer the
reader to two recent surveys Wall and Wood (2005: 454) and Boselie et al.
(2005: 812).
Second, the recommendations do not work in practice. This is evident
from the fact that scores of firms employing best-practice HRM techniques are not enjoying the differential in their performance used to
underscore the case for the scientific approach to the HRMP link.
Crucially, when stepping into the practical domain, many HR managers

681

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
find little resonance between the scientific claims surrounding statistical association and causality with their own organizations everyday
praxis. This is a point we shall return to below when we present our own
empirical findings.
Third, HR practices that are associated with increased performance in
one industry, occupation, location or time appear not to be associated
with increased performance in other industries, occupations, locations or
times. Indeed, to the best of our knowledge, none of the key empirical
studies has been replicated, using the same model but new data from a
later time period (Fleetwood and Hesketh, 2006). Following Sayer (2000),
we have philosophical objections to the epistemological viability of
apparent variations in the capacity of HR to drive organizational performance across different industrial sectors.
Finally, the social world is not only far more complex than this
scientific approach presupposes, but HR managers also know it is far
more complex, and implicitly reject the meta-theoretical presuppositions
of the scientific approach. If this were not the case, then HR managers
would be able to take any one of the scores of studies claiming to have
identified a relationship between some bundle of HRM practices and
performance, perhaps in a similar industry, introduce this bundle, and
expect a similar increase in performance. That experience rarely lives up
to such prediction raises important questions.
One such study often cited in defence of the scientific approachs
acknowledgement of this complexity claims to have used extensive field
work that revealed which HR policies differentiated mass and flexible
production systems most clearly (MacDuffie, 1995: 203). Crucially, the
loss of social complexity that goes with the transformation of qualities to
quantities as social phenomena are transformed into variables that facilitate statistical manipulation, leads to an inevitable trade-off:
I selected for measurement only practices that could potentially be implemented in any plant in the international sample, thus excluding practices
that are exclusively associated with one particular company or country.
(MacDuffie, 1995: 203; emphasis added)

This operationalization of HR practices represents for us an unsatisfactory trade-off in the analysis of the HRMP link. This is emphatically
recognized by the extensive literature devoted to the significance of
unique organizational practices and (often intangible) resources represented in the resource based view (Barney, 1991). Consequently, the
complex organizational processes and practices thought crucial to
enhanced performance might be overlooked in an attempt to find a statistical common denominator. But nor is it just a case of merely identifying
and then quantifying the right variables. Significant questions remain as to
whether the scientific approach reifies the significance of some (quantifiable) variables at the expense of (unquantifiable) social practices at lower
levels within organizations. We are not alone in this observation. In a
recent review of the HRMP literature, Pauwee (2004: 55) notes: to date

682

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
there is very little research that peels back the onion and describes the
process through which HRM systems influence the principal intermediate variables that ultimately affect firm performance.
In sum, whatever the merits of using the scientific approach to carry
out research into the HRMP link, it simply does not reach far enough
inside organizations to explain what is going on therein. Far from seeing
the pursuit of statistical associations as an end game, as it is often
presented in academic journals and (especially) the more generic publications aimed at HR professionals, we interpret it as a point of departure. At best, it tells us there is a connection to be explained and, at this
point, the search for a deeper understanding of the relationship between
HRM and organizational performance must begin. Indeed, a recent observation by one of the leading researchers in the field (Paauwe, 2004: 367)
has called for this very intellectual exercise, arguing for such development to be couched in terms of Institutional theory suggesting that we
need a theory to assess the relationship between a set of HRM policies
and practices and to explore how these relate, interact, or are influenced
by the context? To do this, however, we need to go beyond this
scientific approach and its notion of causality as mere regularity or
association. To build on the observations revealed via statistical analysis,
we need to consider the contribution meta-theory can make to our
understanding of the link between HRM and organizational performance.
And this brings us to critical realism.

Going Beyond Science: Critical Realist Meta-Theory

This section sketches critical realist meta-theory to provide a way of
meta-theorizing the connection between HRM and performance, without
reducing this to statistical association, and to do so in ways that are
totally in keeping with Institutional theory.
Establishing the connection between critical realism and Institutional
theory is difficult not only because there are many versions of Institutionalism, but also because critical realism, as a philosophy of science,
inhabits a meta-theoretical domain, whereas Institutionalism inhabits a
theoretical domain: the role of meta-theory is to interrogate the presuppositions of any theory. Nielsen (2001) provides an excellent overview of the main Institutional theories or approaches (e.g. American
neo-Institutionalism; evolutionary economics; new economic sociology;
historical Institutionalism; the cognitive-institutional approach and new
Institutionalism in political science) before going on to demonstrate how
five of the following six approaches are compatible with critical
realismthe latter being the odd one out because of its commitment to
positivism. He then summarizes the key theoretical assumptions of these
approaches in four points which are shared by critical realists. First, they
focus on the role of institutions (and social structures) such as habits,
routines, and norms in the coordination of behaviour. Second, human
agency is seen as purposeful, or intendedly rational, endowed with some

683

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
freedom to deliberate or choose in accordance with individual psychology rather than, on one hand, as irrational, automatic rule-followers or
totally encapsulated in an externally defined role or, on the other hand,
rational in the sense of isolated maximizing economic man. Third, they
emphasize the constitutive importance of the cultural and cognitive
framework. Finally, they recognize the central and pervasive role of
power and conflict.
One reason to doubt the affinity between critical realism and Institutionalism, is that critical realists tend not to use the term institution.
This is, however, largely because they tend to favour the term social
structures. This difference is largely semantic because there is general
agreement that institutions are a specific kind of social structure. Institutions according to Hodgson, are the kind of structures that matter most
in the social realm . . . (Hodgson, 2005: 2; see also Lawson, 2005). All
this suggests that the critical realist meta-theory may well provide an
appropriate meta-theoretical underpinning to Institutionalist theory. This
raises the following question: how might Institutionalist theory utilize
critical realist meta-theory?
All theory (consciously and explicitly or unconsciously and implicitly)
employs meta-theory. If a meta-theory is not consciously or explicitly
selected, then a meta-theory will be implicitly or unconsciously presupposed anyway. It seems entirely reasonable to suggest that appropriate meta-theory should be carefully considered and selected such that
it is consistent with the theory rather than simply allow theory to be
informed by meta-theoretical happenchance. Indeed, Nielsens paper
demonstrates that errors creep into Institutionalist theory precisely
because inappropriate meta-theory is implicitly or unconsciously presupposed. Let us consider an example closer to home.
Working within the Institutionalist tradition, Ramsay et al. (2000)
consider the possibility that the enhanced organizational performance
associated with high performance work systems is caused by work
intensification, offloading of tasks, controls and increased job strain
something they refer to as the Labour Process model. The conclusion of
their paper, however, runs into a contradiction, arguably, meta-theoretical
in nature. None of the models they test provide an adequate account of
the outcomes of high performance work systems. Although they reject the
possibilities that the data are problematic because they are based on the
respected WERS 98 data, and that the statistical models are problematic
because they encapsulate established measures that have been tested
previously, they overlook the possibility that this entire scientific
approach is wanting. Indeed, they recognize that:
The statistical models of the relationships between employees responses to
HPWS [high performance work systems] practices, employee response and
organizational outcome used in the analysis are perhaps too simplistic to
capture the complex reality of the implementation of the operation of
HPWS . . . It is quite plausible that outcomes flowing from managerial

684

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
innovation are much less determinate than our models apprehend.
(Ramsay et al., 2000: 52120)7

We suggest that conscious application of critical realism would have

sharpened their (existing) insights, and discouraged them from employing meta-theoretical apparatus that they themselves doubt. If Institutional
theory can benefit from the meta-theoretical clarity that critical realism
brings, what exactly is it that critical realism offers?

Critical Realism
Critical realism is a meta-theory rooted in ontology. Critical realists have
made headway in recent years in developing a social ontology, namely a
set of very abstract statements about the way the social world is (its being
if you like) and, from this, have derived methodological commitments.
Let us start with ontology and work our way to methodology.
The social world consists of human agents and social structuresby
which we mean institutions, mechanisms, resources, rules, conventions,
habits, procedures, and so on. This has been discussed since the 1980s in
social theory under the agencystructure framework, although critical
realists have done much to develop this framework recently (Archer, 1995;
Stones, 2005). Notably, critical realists emphasize the transformational
nature of the social world, whereby agents draw upon social structures
(etc.) and, in so doing, reproduce and transform these same structures.
As is now well known, critical realists make great play of open and
closed systems. Systems are defined as closed when they are characterized by event regularities, and open when they lack event regularity,
where an event regularity is styled as: whenever events x1, x2, . . . xn then
event y or in stochastic form, whenever events x1, x2, . . . xn on average,
then event y on average. Event regularities are also styled as: y = f(x1 . . .
xn) and form the basis upon which any mathematical or econometric
specification is derived. However, event regularities, and hence closed
systems, are extremely rare phenomena, especially in the social world.
Presupposing closure when the social world appears to be open, therefore, initiates a series of problems that we cannot go into here (cf.
Lawson, 1997; Fleetwood, 2001). Does empirical research on the HRMP
link presuppose a closed system? In a word, yes. To suggest, as the
literature overwhelmingly tries to, that some HRM practices are statistically associated with increased performance, is to assume regularity
and hence closure.

Critical Realism at Work

It is almost always the case that physical entities, such as information
technology, and social entities, such as organizations, unions, high performance work systems, and so on, exist as clusters of components that
endow them with powers. An organization consists of a cluster of social
structures, institutions, mechanisms, rules, resources, conventions, habits procedures, etc., along with the human agents that activate them.

685

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
Because the entity usually does whatever it does in virtue of the interaction of the totality of causal components, we need a term to refer to
them as a whole. The term we use is a generative ensemble.
We can think of the firm as a generative ensemble that enables or
causes the production of goods and services. Or we can think of the
workplace, the shopfloor, the work-system or the team, as a complex web
of interlocking generative ensembles, sub-configurations, sub-subconfigurations, and so on. Just-in-time production, for example, is possible because the sub-configuration that enables inventory interlocks with
the sub-configuration that enables distribution within the plant. The subconfiguration that causes distribution within the plant interlocks with the
sub-sub-configuration that consists of the maintenance of fork-lift trucks.
Much depends upon the questions we are asking, and the level of
abstraction we are using. Certain business processes manifest themselves
as sub-configurations more readily than others. Rarely, however, do such
configurations, sub-configurations and sub-sub configurations lend themselves to measurement. This complexity is routinely overlooked by the
HRMP literature that often utilizes simplistic and overarching HR
structures as proxies, thereby ignoring the influential and complex
underlying causal mechanisms at work in the social processes underpinning such HR work practices.
From our interviews with senior HR managers and function directors
(now totalling over 70), we found clear parallels between our notion of
generative ensemble with the ways in which practitioners understood
and conceptualized their day-to-day activities:
When Im working on the implementation of [HR] processes I find myself
thinking about the orchestra analogy. You know, like a conductor. The
conductor plays no instrument and makes no sound, but without a
conductor youre just going to get a cacophony of noise rather than
something which conforms to a score. Your [High Performance Work
Systems] are the score which describes, or denotes, what the thing should
sound like in the end, and the implementation [of HR processes] is the
conductor. And if youre trying to evaluate different sections of the
orchestra and how well do they work, the answer is it doesnt all depend
on the score. If you have a score that actually doesnt have any brass in it,
you cant say the brass section is pretty crap because theyre not making
much noise. The score simply doesnt call for that. When one is trying to
evaluate the whole impact of HR one very much has to take into consideration the whole score. Some things are very much in the fray, and others
arent called for very much. You will bring in things in operations which
can make a difference, and then afterwards the effect might be quite
muted. (HR Director; emphasis added)

Powers
Entities (including humans and social structures) possess powers, that is,
dispositions, capacities and potentials to cause certain things, but not
others. Gunpowder, for example, has the power to explode, but not to

686

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
speak a language: humans have this power. Now powers are rather
complex things. Powers may be possessed with or without being exercised and may be exercised with or without being actualized; hence, the
italicized section in the above interview extract. We would add to this
observation some additional points.
First, a power is possessed by an entity in virtue of its internal makeup, and this power endures whether or not it is exercised or actualized,
and therefore endures irrespective of any outcomes it generates. When a
power endures in this sense, it can be said to act transfactually. An
employee has the power to work effectively in the sense that she may
work in a highly productive, highly value-adding manner. Or she may
not. When exercised, certain HRM practices may have the power to foster
certain high value-adding behaviours. Or they may not. The point is the
power endures even if it remains unexercised and un-actualized. One
CEO, this time of a major consulting house, emphatically concurred with
this view:
I think the question you ask encapsulates the problem we face. If a
sceptical CEO needs to be convinced about our proposition that the
performance of their infrastructure can be improved, Im frankly, not
interested. If they are seeking some proof of concept that is null until
proven, then Im already on my way out of the door. I really have not got
the time, patience, or inclination to go through that sort of meander down
to Damascus, frankly. If youre looking for a well articulated scientific
model with evidence, then, unless you are prepared to listen to marketing
hyperbole, you wont find it.

Second, and following on from the previous point, a power exercised is a

possessed power that has been triggered, and is generating an effect in an
open system. Due to interference from the effects of other exercised
powers, however, we can never know a priori, what the outcome of any
particular power will be. An exercised power may act transfactually.
Consequently, it is not the existence of certain powers or structures of
aggregated powers often labelled as key HR practices in the HRMP
literature that cause enhanced performance. Rather, the actualization of
these powers is itself contingent on other, underlying powers and structures that may, or may not, be actualized at any particular time and/or
place. Understanding the processes through which certain HR practices
or strategies enable certain powers represents the Achilles heel of current
HRMP literature, and, in many ways, represents the starting point in our
empirical analysis of HRs role in explaining the performance of people,
their teams and the organizations to which they belong.
Third, a power actualized is an exercised power generating its effect
and not being deflected or counteracted by the effects of other exercised
powers. An actualized power does not act transfactually but factually in
the sense that the power generates its effect.
Finally, people can have personal powersa reference to agents
subjective and reflexive formulation of personal projects (Archer, 2003: 5).

687

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
Powers are possessed by workers in virtue of their biological, physiological, psychological and social make upalthough it is important to
note that these levels are irreducible to one another so we cannot simply
reduce social behaviour to biology. Unlike most animals, humans do not
just execute genetically pre-programmed tasks; they conceive these tasks
firstalthough there may be a complex and recursive process between
conception and execution. The power of conception is of crucial importance here because it consists of the powers of imagination, ingenuity and
creativity that conceived of the Pyramids, the Guggenheim, the cart, the
MIR space station, surgical tools, nuclear weapons, and the HR business
partner. These same powers of imagination, ingenuity and creativity are
also exercised in the conception of less grandiose endeavours such as
finding better ways of producing a rivet, writing a programme or engaging
in a telephone conversation. HRM practices such as performance management, retention, job design, and especially engagement, along with
schemes to increase employee participation and empowered employees,
are designed to unleash and harness the powers of imagination, ingenuity, customer service and creativity that workers bring with them to the
work place. If workers did not have these powers there would be no point
whatsoever in even contemplating HRM practices. Indeed, the sub-title of
Pfeffers (1994) book Unleashing the Power of the Workforce captures the
point beautifully. The fact that HRM has not succeeded in unlocking
workers powers does not mean they do not exist: something could be
counteracting these powers. Or, alternatively, these powers may well
have been unleashed: that theyor their impactcannot be statistically
captured does not preclude their existence.

Tendencies
The term tendency refers to a force. Metaphorically speaking, a force
drives, propels, pushes, thrusts, asserts pressure and so on. Perhaps the
most important point to note about a tendency is that it refers not to any
outcome or result of some acting force, such as a regularity or pattern in
the resulting flux of events, but to the force itself. A tendency should not,
therefore, be (mis)understood as some kind of rough and ready event
regularity or law, nor as a stochastic regularity.8
Now, to write that a configuration has a tendency to x, does not mean
that it will x. In an open system, configurations do not exist in isolation
from one other, rather there is a multiplicity of such configurations each
with their own tendencies and these tendencies converge in some space
time location. The relation between configuration and tendency might be
characterized as follows. The configuration does not always bring about
certain effects, but it has a tendency to cause them. Hence, it acts transfactually. Configurations continue to cause the flux of events, irrespective of
the conditions under which they are said to operate. We do not say of a
transfactually acting configuration that it would bring about certain events
if certain conditions prevail, or ceteris paribus. Rather, the configuration

688

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
tends to cause certain events, period. Configurations continue to enable the
flux of events irrespective of any events that ensue. A transfactually acting
configuration does not depend for its action upon the patterns of events
that it causes: it continues to enable social processes, whether the
ensuing events are constantly or non-constantly conjoined.
Each configuration (and sub-configuration) generates tendencies and
these tendencies can counteract, and augment, one another in complex
ways. The sub-configuration that constitutes a workforce with the tendency to resist control, co-exists simultaneously with the subconfiguration that constitutes a management team with the tendency to
assert control. The outcome, however, depends upon the relative
strengths of the tendencies. Notice, then, that it is the generative ensemble as a totality, and not any of its individual components that generates
the powers and, therefore, the tendencies the generative ensemble has.
Notice also that statistical analysis cannot deal with tendencies because
it can only deal with phenomena that are actually manifesting their
generative powers, whence they can be observed in action. We can, for
example, statistically analyse the tendency of gunpowder to explode, but
we cannot statistically analyse the powers of imagination, ingenuity and
creativity possessed by a workforce.
In the following example, this HR director illustrates our point about
the interplay between powers and tendencies, their configurations, subconfigurations and sub-sub-configurations, while at the same time alluding to the transfactuality of HRs capacity to work only when in the right
hands and minds derived from experience. In short, it is not the actual
insertion of certain HR systems and structures, but understanding
through experience and reflexivity how they work in light of the subtle
complexities, and processes by which societal expectations of appropriate organizational action influence the structuring and behaviour of
organizations in given ways (Dacin, 1997: 48) that is crucial:
All of this stuff about High Performance Work Systems to my mind misses
the point. Its a bit like taking a Formula 1 motorcar and asking somebody
with a conventional driving licence to race it round Silverstone. They
know that the car has gears, a clutch, a steering wheel and brakes, but that
person has no idea about the capability of the car, its nuances and what
you have to do in order to get the best out of that car. That is not to say that
they will not be able to drive it around the track. But you have to know
what youre doing with it, have the experience, and understand the whole
process of racing the car. There are all sorts of complications that you have
to bring into the equation. You have a team of mechanics to support you,
they need to know when youre coming in in order to prepare. [. . .] A lot of
HR is like bad racing car driving. Theyve got the kit but it doesnt always
work because you have to know when to use some things and not others.
Its about experience, not textbooks telling you what to do. Everybody
wants to isolate the it and do something with it. Its much more subtle
and powerful at the same time. (HR Director)

689

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles

From Causality to ExplanationReflexive Performance

Complex causality refers to a situation where the cause of an event is not
assumed simply to be the event(s) that preceded it (which would be some
kind of simple causality), but rather is the entire conflux of interacting
causal phenomena. The causality of the lamps illumination, for example,
is the nature of the glass, the gas, the filament, the wire, the switch, the
plug, the electricity, as well as the finger that flicked the switch. None of
these events in isolation causes light. Rather, it is their configuration in a
particular form that causes in a complex way their aggregated powers to
generate light.
Complex causality is connected to what we call robust explanation.
Providing a history of a phenomenon, and hence explaining it, could be
interpreted to mean giving information about the underlying mechanisms
and structures, along with (if we are dealing with social phenomena) the
human agency that reproduces and transforms these mechanisms and
structures. That is, explanation could be based upon complex causality. If
and when causality is complex, then explanation cannot be reduced
merely to giving information about a succession of events, but rather,
requires information about the entire conflux of interacting causal phenomena beyond that captured even by sophisticated techniques utilized
in the multiple analysis of variance (MANOVA). Information about the
entire conflux of interacting causal phenomena is necessary for a robust
explanation. Information about the nature of the glass, the gas, the
filament, the wire, the switch, the plug, the electricity, as well as the
finger that flicked the switch, all add to the richness of the explanation
and are therefore not superfluous but absolutely necessary. There is little
doubt that most of us would recognize this information immediately as
constituting a very rich, or robust, explanation because it would (at the
very least go some way to) answering the question Why?.
People and HR processes, of course, do not behave in similar ways to
glass, gas, filaments, wires, switches and electricity. So let us consider
what would constitute a robust explanation. If and when a particular
outcome is complex in its enablement, then explanation is irreducible
merely to giving information about a succession of events. A robust
explanation of the increase in, for example, the performance of any given
HR processes and their capacity to drive business performance requires
two kinds of information. First, a robust explanation requires, what we
might call, hermeneutic information. That is, information relating to the
way the relevant agents (i.e. stakeholders) interpret, understand, make
sense of, the workplace and thereby initiate action. The following example
from our interviews sets these observations into institutional context:
That thing about numbers is quite interesting because they are not a
differentiator in their own right. All they are is an indicator that the
sausage machine is churning out what it is supposed to be churning out.
Once the numbers settle down, after the initial blips, we dont even need to
see them. They dont drive the performance of HR. [. . .] Its much more

690

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
detailed and complicated. So we come to measurement from a very
different perspective. (HR Director)

Second, a robust explanation requires information about a significant (but

not infinite) set of interacting and causal phenomena through which
agents initiate this action. This might, for example, include information
on the social, political, economic and spatial environment of the industry
and/or the firm; the composition of the team; the experiences and wishes
of individuals comprising the team; the nature of the new jobs, tasks and
skills (if any); the relationship between team members, the line managers
and corporate strategy; the nature of control in the firm; and the nature of
any synergies (or dis-synergies) created by the interaction of these enabling causal phenomena, and so on. The following provides an example
of this from our interviews:
When I want to know, Well, what do we actually in fact do? I have to lift
up the stones and have a look underneath. Many in HR are almost
exclusively reactive; put the structures in place and see what happens. You
cant just buy things or people in and help organizations through major
change programmes. You have to work at it, to better understand whats in
there in front of you. (HR Director)

Clearly, these examples are not exhaustive but they are indicative of
many of the accounts we heard. Furthermore, all of the causal phenomena identified in the examples we provide could be broken down into
their components as we try to provide more detailed information.9
Needless to say, a great deal of this information will be irreducibly
qualitative. All this information adds to the richness of the explanation
and is therefore not superfluous but absolutely necessary. There is little
doubt that most of us would recognize this information immediately as
constituting a robust explanation because it would (at least to some
degree) answer the question Why?.

Explanation and Tendential Prediction

Because of the openness of social systems, events cannot be inductively
predicted, or predicted as deductions from axioms, assumptions and
lawsas in the deductive method sketched above. But the social structures, institutions, mechanisms, rules, resources, etc., that human agents
draw upon in order to initiate action, can be retroduced and their
operation uncovered and explained. Retroduction consists:
. . . in the movement, on the basis of analogy and metaphor amongst other
things, from a conception of some phenomenon of interest to a conception
of some totally different type of thing, mechanism, structure or condition
that, at least in part, is responsible for the given phenomenon. If deduction
is illustrated by the move from the general claim that all ravens are black
to the particular inference that the next one seen will be black, and
induction by the move from the particular observation of numerous black
ravens to the general claim that all ravens are black, retroductive or
abductive reasoning is indicated by a move from the observation of

691

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
numerous black ravens to a theory of a mechanism intrinsic . . . to ravens
which disposes them to be black. (Lawson, 1997: 24)

For critical realists, then, retroduction replaces induction and deduction

as modes of inference, and explanation replaces prediction as the key
objective of science. To the extent that we can successfully retroduce to
the causal structures (etc.) that govern some observation (and there is no
gain-saying the difficulty of this), we have a theory that explains this
observation. To the extent we have a theory and an explanation, we have
an understanding of the tendencies generated by these structures (etc.).
To the extent that we understand these tendencies, we can make claims
about how the structures (etc.) are likely to govern the actions of the
human agents that draw upon them. We hesitate to call this a prediction
because the term is now so entwined in scientistic discourse that it is
almost impossible to untangle it and give it another meaning. Nonetheless, it is a prediction of some kind, albeit heavily qualified, and we call
it tendential prediction.
How might this work in the case of Human Resource Management?10 If
we can successfully retroduce to the social structures (etc.) that, when
drawn upon by workers and managers, cause bundles of HR practices to
(say) increase organizational performance, then we have a theory with an
explanation of organizational performance. Such a theory with an explanation would allow us to understand the tendencies generated when
workers and managers engage with HR practices and social structures
(etc.). If we understand these tendencies we can make tendential predictions. We might, for example, be able to understand the tendencies
generated by the exercise of human labour power to activate workers
powers for creative, imaginative, ingenious, self motivated and self
directed action, as well as the countertendencies generated by the alienation, exploitation and commodification of human labour power. We
might therefore be able to assess the efficacy of tendencies and countertendencies, and make a tendential prediction about the likelihood of
specific bundles of HR practices increasing organizational performance.
This process involves what we refer to as reflexive performance.

Reflexive Performance
According to one of the leading critical realists, Archer (2003), individuals often engage in a specific agential enterprise which she refers to as
a project. When deliberating over a project, agents engage in a process of
reflexive determination, which refers to the personal process undertaken by individuals in identifying the causal mechanisms linking structure to their agency. What we call reflexive performance, involves agents
not only identifying the structures, institutions, mechanisms, rules,
resources, etc., deemed to enable or constrain wider organizational
performance, but also constructing strategies to use them to pursue their
personal goals. No social structure (etc.) is constraining or causal tout
court. Whether social structures enable or constrain depends upon the

692

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
nature of the relationship between it and the individual drawing upon it.
Whether or not their [i.e. structures] causal power is to constrain or to
enable is realized, and for whom they constitute the constraints or
causalites, depends upon the nature of the relationship between them
and agential projects (Archer, 2003: 8; original emphases).
Reflexive performance, affords two critical observations in examining
the HRMP link: namely, the transfactuality of HR practices and their
impact upon wider organizational performance, and, second, the contingency, and hence variability, in the capacity of certain HR practices to
generate the outcomes expected. Let us deal with each of these in turn.
In terms of transfactuality, open systems and lack of event regularity
means that the structures and practices of high performance work
systems may generate enhanced organizational performance, or they
may not. This goes a long way to explaining the wide variations in
organizational performance of organizations with similar HR structures,
processes and practices. In Archers critical realist language, we have to
recognize that it is essential to distinguish between the existence of
structural properties and the exercise of their causal powers (Archer,
2003: 7). This in turn relies on the reflexive determination of those HR
professionals and other employees engaged in the processes thought to be
performance enhancing. Therefore, to predict outcome y as the direct
result of implementing high performance work system x negates (or at
least fails to consider) the role of individuals in (not) adopting or aligning
certain prescribed HR processes and practices with their, and others,
agential projects.
For Archer, this process of explanation is about the interplay between
the subjective world of agents and the objective and independent world
of social structures and institutions. All this takes place in a process
labelled discursive penetration. Individuals, be they HR professionals,
managers or other employees, must all diagnose their own situations,
identify their own interests and align the situations in which they find
themselves in some way with their own agential projects. For Archer, at
all three of these points, individuals are fallible in terms of misdiagnosis
of their situations, the misidentification of their own interests and of
course the misjudgement of what they deem appropriate action. All these
processes require an internal conversation:
The answer to this is held to be via the internal conversation. This is the
modality through which reflexivity towards self, society and the relationship between them is exercised. In itself it entails just such things as
articulating to ourselves where we are placed, ascertaining where our
interests lie and adumbrating schemes of future action. (Archer, 2003: 9)

The internal conversation of the HR directors we have presented above

reveals to us what they believe to be the case and it is causally efficacious, at least to a degree, because their thinking is deemed to be
efficaciousthey have the power to implement practices they deem
appropriate. The conversation then becomes one between what might be

693

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
labelled self knowledge (what one knows and thinks), societal knowledge
(knowing what others think), a stance (the adoption of a position of what
one wants to achieve) and the legitimacy of this agential project underpinned by the accompanying explanations to justify ones project to
oneself. Here, at last, we move away from scientific notions of causality
(as event regularity) to the genuine interplay between subjective properties (discursive penetration through the internal conversation) and objective structures (organizational successes, failures, good performance,
bad performance, etc.). People, their diagnoses and their actions, then,
are as influential and important for analysis, as the structures and
institutions they necessarily engage with. That said, institutions and their
environments play a major role in shaping how the causal role of HR is
perceived and explained. Moreover, even when individuals, and in the
case of the example below, very powerful individuals have one view of
how the HRMP link should be understood, institutional isomorphism
takes over:
Im alarmed at the time devoted to financials by HR. [. . .] There are
measures that we are immersed in, dissecting every dollar for the CEO
whose focus is on the score and on the numbers. Nothing else here appears
to matter. (HR Director)

In sum, critical realism provides us with three meta-theoretical insights.

First, it provides us with a sophisticated understanding of the limitations
of the scientific approach, and therefore sound reasons to move beyond
it. Second, it provides us with one possible alternative, namely, the
causal-explanatory methoda method more suited to the openness of the
social world in general, and one identified by Institutional theorists in
particular. Third, it provides us with a notion of reflexive performance
where we seek to identify the enabling causal configurations at work,
through the discursive penetration of HR professionals internal conversations. Understanding the underlying processes which enable or
constrain certain processes is far more important than the adoption of
some form of scientific thinking in which outcome y is guaranteed
through the implementation of x.

Conclusion
Although the scientific approach, with its commitment to empirical
research techniques, might be a useful starting point for understanding
the HRMP link, it is currently treated as the end game. We suggest that
critical realism, by offering a more fruitful meta-theory, can play a
significant part in developing an understanding to the causal role played
by HR in organizational performance. But we do not live in a social
vacuum; this much has been revealed to us in our ongoing work with HR
directors, who may well (implicitly) subscribe to much of the metatheory we propose here, but have to implement it in the real-world
institutions in which our research took place. It is for this reason that we

694

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
leave the final word to one of our respondents, Chief Operating Officer
and Vice President of HR of one of the worlds largest financial houses:
HR plays, or I should say, HR should play a fundamental role in helping
the organization achieve its mission. Ive always said, you know, I got the
questions when I announced I was moving into HR, Carlos, wait a second,
what are you doing? Have you lost your marbles going into HR? My
response to this was, Guys, Im moving into the most important function
of the bank, and they all laughed. For me its very simple, what drives
business performance? Its all about people, right? And whos focused on
people? HR. [. . .] The challenge is finding the linkage [between HRM and
performance] and for this you need to move to the next level of understanding whats going on.

Notes
1

6
7

8
9

Accenture (2004); Bukowitz et al. (2004); Donkin (2002, 2003, 2004);

Kingsmill (2003); Overell (2002); PricewaterhouseCoopers (2003); Saratoga
(2005); Thomas et al. (2003); Witzel (2004); Wright (2001); Watson Wyatt
(2001/2: 5 and 1).
Rather than provide the usual (long) list of references, we simply point the
reader to the survey by Wall and Wood (2005) of the 25 most commonly cited
articles that appeared in reputable journals.
We use meta-theory as a portmanteau term to refer to philosophy of science,
ontology, epistemology, methodology, causality and research techniques. For
a discussion in the HR context, see Pauwee (2004).
For an introduction to critical realism, see Ackroyd and Fleetwood (2000);
Archer et al. (1998); Carter and New (2004); Danermark et al. (1997);
Fleetwood and Ackroyd (2004); Lawson (1997, 2003); Reed (2001); and Sayer
(1994, 2000). Several other writers in the Institutionalist tradition have made
use of critical realism without feeling the need to elaborate upon it (e.g.
Delbridge, 2000: 149; Beynon et al., 2002: 346).
We acknowledge that some researchers may not find these summaries
satisfactory. Indeed, we refer to these perspectives only because researchers
deploy them to represent the different schools of thought within the HRMP
link field of research they identify with. An evaluation of their adequacy is
long overdue and, in many ways, our paper represents the initiation of this
process.
The preoccupation with quantitative analysis is demonstrated by Hoobler
and Brown Johnston (2005: 668) and Boselie et al. (2005: 70).
Even leading Institutionalists Powell and DiMaggio (1991) find it difficult to
break entirely with the scientific approach. This is exemplified by their
desire to seek inductively based predictions.
For an elaboration of tendencies, see Fleetwood (2001).
Although the list of what could, in principle, be included in a robust
explanation could expand to include everything all the way back to the big
bang, in practice, social scientists usually avoid this problem by making use
of abstraction (i.e. making judgements about which factors to include and
exclude). Sometimes researchers get it wrong. This problem is no different
than deciding upon which variables to include and exclude (cf. Runde,
1998).

695

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
10

This sketch is, of course, at a very high level of abstraction. The point is to
illustrate how we might go about using the meta-theoretical concepts known
to critical realism to generate a tendential prediction.

References
Accenture (2004) The High Performance Workforce Study, Accenture Publication, 178. New York, NY: Accenture.
Alvesson, M. and Skolberg, K. (2000) Reflexive Methodology: New Vistas for
Qualitative Research. London: Sage.
Ackroyd, S. and Fleetwood, S. (2000) Realist Perspectives on Organization and
Management. London: Routledge.
Archer, M. (1995) Realist Social Theory: The Morphogenetic Approach. Cambridge: Cambridge University Press.
Archer, M. (2003) Structure, Agency and the Internal Conversation. Cambridge:
Cambridge University Press.
Archer, M. Bhaskar, R. Collier, A. Lawson, T. and Norrie, A. (1998) Critical
Realism: Essential Readings. London: Routledge.
Barney, J. (1991) Firm Resources and Sustainable Competitive Advantage,
Journal of Management 17(1): 99120
Becker, B. and Gerhart, B. (1996) The Impact of Human Resource Management on
Organizational Performance: Progress and Prospects, Academy of Management
Journal 30(4): 779801.
Beynon, H. Grimshaw, D. Rubery, J. and Ward, K. (2002) Managing Employment
Change: The New Realities of Work. Oxford: Oxford University Press.
Boselie, P. Dietz, G. and Boon, C. (2005) Commonalities and Contradictions in
HRM and Performance Research, Human Resource Management Journal 15(3):
6794.
Boudreau, J. and Ramstad, P. (1999) Human Resource Metrics: Can Measures Be
Strategic, Research in Personnel and Human Resources Management Supplement 4: 7598.
Brown, D. (2004) Capital Vetters, People Management 30 September: 3841.
Bukowitz, W. Williams, R. and Mactas, E. (2004) Human Capital Measurement,
Industrial Research Institute MayJune: 439.
Carter, B. and New, C. (2004) Making Realism Work: Realist Social Theory and
Empirical Research. London: Routledge.
Dacin, M. T. (1997) Isomorphism in Context: the Power and Prescription of
Institutional Norms, Academy of Management Journal 40(1): 4681.
Danermark, B. Ekstrom, M. Jakobsen, L. and Karlsson, J. (1997) Explaining
Society: Critical Realism in the Social Sciences. London: Routledge.
Davis, S. (2003) Is This The End of HR?, in M. Effron, R. Gandossy and M.
Goldsmith (eds) Human Resources in the 21st Century. New York, NY: Wiley.
Delbridge, R. (2000) Life on the Line in Contemporary Manufacturing: The
Workplace Experience of Lean Production and the Japanese Model. Oxford:
Oxford University Press.
Delery, J. E. and Doty, D. H. (1996) Modes of Theorizing in Strategic Human
Resources Management: Tests of Universalistics, Contingency and Configurational Performance Predictions, Academy of Management Journal 39:
80235.

696

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
Donkin, R. (2002) Measuring the Worth of Human Capital, Financial Times 7
November.
Donkin, R. (2003) Use a Foot Rule To Measure Your Best Assets, Financial Times
31 October.
Donkin, R. (2004) The Rise of People Power, Financial Times 29 April.
Ellinger, A. Ellinger, A. Yang, B. and Horton, S. (2002) The Relationship Between
the Learning Organization Concept and Firms Financial Performance: An
Empirical Assessment, Human Resource Development Quarterly 13(1): 521.
EquaTerra (2005) Equaterra Releases Quarterly Outsourcing Surveys identifying
Healthy, Yet Slowing Growth; HR and IT Are Strong Performers, Press Release
15 July 2005. New York, NY: EquaTerra.
Fairclough, N. (2005) Discourse Analysis in Organization Studies: The Case for
Critical Realism, Organization Studies 26(6): 91539.
Fleetwood, S. (2001) Causal Laws, Functional Relations and Tendencies, Review
of Political Economy 13(2): 20120.
Fleetwood, S. and Ackroyd, S. (2004) Realism in Action in Organization and
Management Studies. London: Routledge.
Fleetwood, S. and Hesketh, A. (2006) The Performance of HR. Cambridge;
Cambridge University Press.
Guest, D. (1997) Human Resource Management and Performance: A Review and
Research Agenda Process, International Journal of Human Resources Management 8(3): 26376.
Guest, D. (2001) Human Resource Management: When Research Confronts
Theory, International Journal of Human Resources Management 12(7):
1092106.
Hodgson, G. (2005) What Are Institutions?, Journal of Economic Issues 40(1):
125.
Hoobler, J. and Brown Johnston, N. (2005) An Analysis of Current HRM Publications, Personnel Review 33(6): 66576.
Huselid, B. (1995) The Impact of Human Resource Management Practices on
Turnover, Productivity, and Corporate Financial Performance, Academy of
Management Journal 38(3): 63572.
Kingsmill, D. (2003) Accounting for People, Report of the Task Force on Human
Capital Management. London: DTI.
Kinnie, N., Hutchinson, S., Purcell, J., Rayton, B. and Swart, J. (2005) Satisfaction
with HR Practices and Commitment to the Organization: Why One Size Does
Not Fit All, Human Resource Management Journal 15(4): 929.
Lawson, T. (1997) Economics and Reality. London: Routledge.
Lawson, T. (2005) What Is an Institution, Journal of Economic Issues 37(1):
175-207.
Lipton, P. (1993) Inference To The Best Explanation. London: Routledge.
MacDuffie (1995) Human Resorce Bundles and Manufacturing Performance:
Organizational Logic and Flexible Production Systems in the World Auto
Industry, Industrial and Labor Relations Review 48(2): 197221.
McMahan, G. Virick, M. and Wright, P. (1999) Alternative Theoretical Perspectives for Strategic Human Resource Management Revisited: Progress, Problems, and Prospects, Research in Personnel and Human Resources
Management Supplement 4: 99122.

697

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Organization 13(5)
Articles
Murphy, T. and Zandvakili, S. (2000) Data and Metrics Driven Approach to
Human Resource Practices: Using Customers, Employees and Financial Metrics, Human Resource Management 39(1): 91105.
Nelson Hall (2005) BPO Quarterly Report Q4 2005. Press Release. London: Nelson
Hall.
Nielsen, K. (2001) Institutional Approaches in the Social Sciences: Typology,
Dialogue and Future Challenges, Journal of Economic Issues 35(2): 50517.
Osterman, P. (1994), How Common is Workplace Transformation and How Can
We Explain Who Does It?, Strategic Management Journal 18: 697713.
Overall, S. (2002) The Metric System of Performance, Financial Times 31
October.
Pauwee, J. (2004) Human Resource Management and Organizational Performance. Oxford: Oxford University Press.
Pfeffer, J. (1994) Competitive Advantage Through People: Unleashing the Power
of the Workforce. Boston, MA: Harvard University Press.
Pfeffer, J. (1995) Sustaining Competitive Advantage Through the Effective Management of People, Academy of Management Executive 9(1): 5569.
Powell, W. and DiMaggio, P. J. (eds) (1991) The New Institutionalism in Organizational Analysis. Chicago, IL: Chicago University Press.
Powel, W. and Di Maggio, P. (1991) Introduction, in W. Powel and P. DiMaggio
(eds) The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press.
PriceWaterhouseCoopers (2003) Sustaining Value Through People, PriceWaterhouseCoopers Publication, 126. London: PriceWaterhouseCoopers.
Ramsay, H. Scholarios, D. and Harley, B. (2000) Employees and High Performance Work Systems: Testing Inside the Black Box, British Journal of Industrial
Relations 38(4): 50131.
Reed, M. (2001) Organization, Trust and Control: A Realist Analysis, Organization Studies 22(2): 20128.
Ruben, D.-H. (1992) Explaining Explanation. London: Routledge.
Runde, J. (1998) Assessing Causal Economic Explanations, Oxford Economic
Papers 50: 151172.
Saratoga (2005) Key Trends in Human Capital: A Global Perspective, PriceWaterhouseCoopers publication, 119. New York, NY: PriceWaterhouseCoopers.
Sayer, A. (1994) Methodology in Social Science: A Realist Approach. London:
Routledge.
Sayer, A. (2000) Realism and Social Science. London: Sage.
Schuler, R. and Jackson, S. (1987), Linking Competitive Strategies with Human
Resource Practices, Academy of Management Executive 1: 20913.
Stones, R. (2005) Structuration Theory. Basingstoke: Palgrave Macmillan.
Thomas, R. Cheese, P. and Benton, J. (2003) Human Capital Development,
Accenture Research Note 1: 14.
Wall, T. and Wood, S. (2005) The Romance of Human Resource Management and
Business Performance, and the Case for Big Science, Human Relations 58(4):
42962.
Watson Wyatt (2001/2) Human Capital Index: Human Capital As a Lead
Indicator of Shareholder Value, pp. 15. London.
Witzell, M. (2004) A Better Feel for the Intangibles, Financial Times 29 April.
Woodward, J. (1965) Industrial Organization: Theory and Practice. Oxford:
Oxford University Press.

698

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015

Beyond Measuring the HRMOrganizational Performance Link

Anthony Hesketh and Steve Fleetwood
Wright, P. (2001) Avery Dennisons Strategic HR Metrics: Measure Now, Perfect
Later, Strategic HR Review Launch issue: 224.

Anthony Hesketh teaches HR in the Management School at Lancaster University. His

research focuses on capturing the impact of transformational outsourcing on
organizational performance. Recent books include The Mismanagement of Talent
(Oxford University Press, 2004) with Phil Brown and Understanding the Performance of HR (Cambridge University Press, 2006) with Steve Fleetwood. He is the
Director of the new Centre for Performance-Led HR at Lancaster. Address:
Department of Management Learning, Lancaster University Management School,
Lancaster LA1 4YX, UK. [email: A.Hesketh@lancaster.ac.uk]
Steve Fleetwood teaches employment relations and HRM at undergraduate and postgraduate levels. His interests are in work, labour and employment, Marxist
political economy, labour markets and philosophy of science, especially critical
realism. His major publications include: Hayeks Political Economy, the Socio
Economics of Order (Routledge, 1999); Critical Realism in Economics: Development and Debate (Routledge, 1999); Critical Realist Applications in Organisation and Management Studies (Routledge, 2004) with Stephen Ackroyd;
Critical Realism and Marxism (Routledge, 2002) with Andrew Brown and John
Roberts; and Realist Perspectives on Organisation and Management (Routledge,
2000) with Stephen Ackroyd. Address: Department of Organisations, Work &
Technology, Lancaster University Management School, Lancaster LA1 4YX, UK.
[email: s.fleetwood@lancaster.ac.uk]

699

Downloaded from org.sagepub.com at Copenhagen University Library on February 19, 2015