Organization Culture

["Will Glass-Husain"]

Hi,

As part of a much bigger modeling project, I am looking for descriptions of
different ways to model the influence of organization culture on
performance. A difficulty we have been facing is that everyone seems to
agree that organization culture is important, but no one has been able to
clearly articulate causal relationships.

Our goal with this part of the model is to let the user play the role of
management and take initiatives regarding culture as part of an effort to
improve overall performance. (using the model as a "what if" tool).

There seem to be two approaches we could take on this. One is to assume
that organizations contain a culture as an element in itself. We could
model (for example) the degree to which two cultures are aligned in a
merger. The other is to look at specific cultural elements (for example an
organization that puts a high value on customer-relations or a high value on
team collaboration) and describe how this would impact organizational
performance. The difficulty with this is that modeling at too specific
level limits the choices of our users. (i.e. we have to model every
intervention we wish to allow the user to make).

Id appreciate any insights from other folks who have struggled with this.

Best, WILL
From: "Will Glass-Husain" <wilgl@powersim.com>

["Gunkler, John"]

First, a reference to two articles: a "structural equation modeling"
research article and an attempt to model "inertia" (resistance to
change), then some thoughts on the subject.

You might check: Marcoulides, George A. and Ronald A. Heck,
"Organizational culture and performance: Proposing and testing a model,"
1993 (Vol. 4, No. 2, May), Organization Science, pp. 209-225.

Marcoulides and Heck propose a list of variables (warning: This
looks a lot like the "laundry list" method we SD folks disdain) that
may have causal relationships to each other and to organizational
performance; then they "test" the model by interviewing a sample
of 392 people (using linear programming methods, LISREL to test
statistical relationships.)

Also in the same issue of Organization Science -- Gresov, C., H.
Haveman, and T. Oliva, "Organizational design, inertia and the dynamics
of competitive response," pp. 181-208.

At least youll see an attempt to identify variables and that can be
useful.

My thoughts on what youre doing, in general:

1. I have been consulting in the areas of strategic thinking and
organizational change for nearly 20 years, working primarily with
Fortune 1000 companies. I believe that organizational culture has "the
last word" to say about organizational performance -- it is that
powerful. It is especially powerful when an organization tries to
change. Failure to deal explicitly with the culture dooms most change
efforts to failure (i.e., short-term annoyance but no long-term
improvement.)

2. Much of the literature about organizational culture is
pseudo-anthropological. That is, it attempts to draw static pictures of
culture as if the organization were no longer a living thing. I dont
care, in my work, about how prettily someone can dissect an organization
-- identifying tens or hundreds of "important" variables and factors
that, in some fashion not usually well described, "determine" or
categorize the culture. All that Im interested in is dynamic variables
-- what factors, when changed, result in long-term change in the
organization. Where do we push to make something happen? How can one
mobilize the energies and motivation of people to do things differently
in the face of natural resistance to any change? I dont want, nor do I
need, to "characterize" an organizations culture, nor to categorize it,
classify it, nor put it into some static taxonomy. I only want to know
how to make it change in goal-directed ways.

3. Therefore, I believe most of what you will find written about
organizational culture will be of little use (and, in fact, may be very
distracting) to your efforts to model it as a dynamic factor.

4. The Gresov et al. article, while overkilling their methodology
(catastrophe theory), at least attempts to look at dynamic variables.
They identify organizational design as the "splitting factor" which, by
reaching or not reaching a point of "singularity," determines whether
the organization will change or not. Organizational design is
conceptualized on a scale from flexible to inflexible (resistant to
change) based on such things as organizational structure, effectiveness
of the competitor intelligence system, technology, level of internal
stress, age and maturity of decision makers, etc. They try to predict
"intensity of competitive response" (i.e., how much change will occur)
from this "design" factor plus a "competitive pressure" construct.

5. I can recommend practitioner articles (with anecdotal support but no
research) that match my personal experiences in helping organizations
change, if that would be useful. I believe that there are only about
half a dozen critical "factors" to work on in order to make culture
change succeed. They arent as sophisticated as the factors cited in
the two articles above, but perhaps theyre more at the level you want
for you model. Please contact me if you want to discuss this further.

From: "Gunkler, John" <Jgunkler@satmansys.com>

[JohnMcNeil]

Will,
You are asking the wrong people.
Your problem falls into two parts.
First is working out what your mental model is and what the assumptions are
that you want to test. For that, see an organizational specialist, a priest
or your mentor.
Second is modeling it. At that stage you can talk to system dynamicists.
They are no more skilled at structuring your thinking than you are.
John
From: JohnMcNeil <JohnMcNeil@aol.com>

["Shayne Gary"]

Will,
Your initial query and some of the responses posted to the list have
mentioned the use of stock variables for different aspects of
culture, but I would also look for managerial policies which
characterize the culture of the organization. Managements choice of
what information to look at and how to interpret that information is
very much a cultural issue. As an example, an acquired business can
find itself in either a competitive culture in which senior
management evaluates business units on objective financial criteria
and allocates corporate resources based on that data alone, or a
cooperative culture in which senior management evaluates business
units using more subjective criteria and emphasizes resource sharing
and skill transfer among business units. Some businesses would
flourish under the competitive culture while others would flourish
under the cooperative culture. These are clearly just two examples
of different organizational cultures.

Best of luck in your modeling work...



Shayne Gary
London Business School
Tel: 44-171-262-5050 ext 3646
Fax: 44-171-724-7875
sgary@lbs.ac.uk

[GBHirsch]

Hi Will,

In response to your query about modeling organizational culture, Id suggest
that you find variables that describe elements of culture that are independent
of particular interventions, but can change over time as a result of various
interventions. As an example, Ive been working on a model of innovation in
schools with a group of educators (a topic youve had a lot of experience
with). Key variables weve identified include Trust Between School and
Community, Teacher Motivation, and Experience with Innovation. Values of
these variables can be chosen to reflect cultures in different schools
(receptive to vs. hostile to innovation). These variables will affect the
impact of different strategies for implementing innovation and also change in
response to those strategies and their impacts.

The question in modeling any organizations culture would seem to be: "What
are the aspects of culture that have the greatest effect on the organizations
ability to take action?" (as opposed to incidental characteristics that have
little bearing on organizational performance). Answering this question would
allow you to select variables that have the greatest effect on success of
interventions and also identify the aspects of culture that need to be changed
to make desirable strategies succeed. Changing culture is difficult, so one
would want to concetrate on the most critical aspects.

Hope this is helpful.

Gary Hirsch
GBHirsch@aol.com

[Jeff Morrow]

Perhaps some of the difficulty you have stems from individual firms
cultural idiosyncrasies: all of the interesting dynamics may not be
abstract enough to build a generic model with tunable parameters. A good
example of including specific cultural elements can be found in John
Stermans (et al.) model of Analog Devices implementation of TQM.

Having started to build SD models of product development processes and
finding Ed Roberts, Tarek Abdel-Hamids, and Dave Fords work dauntingly
impressive (and already done), I shifted toward trying to find
quantitative, empirical ways to characterize aspects of PD culture.

Ive had some success applying something called consensus theory, a fairly
well developed approach from cognitive anthropology. Done properly,
consensus theory can identify the cognitive domain elements shared by a
cohort, indicate which members of the cohort have more or less cultural
wisdom, estimate canonic cultural wisdom, and quantify inter-subject
correlation. These are all useful for model building, particularly the
identification of what anthropologists call "omniscient observers," the
cohort members who really know whats going on.

Maybe the way to think about modeling culture should include an up front,
empirical investigation of the culture of interest as a given part of the
model building process. This does complicate things and anthropology
takes time. Toward minimizing the burden I have built some PC-mediated
versions of cognitive anthropology tools which allow ad hoc,
self-administration and automate data collection. Demo versions are
available for Windows95 and NT4.0.

Jeff Morrow
From: Jeff Morrow <jfmorrow@u.washington.edu>
morrow:loess 17044 Westside Hwy SW Vashon, WA 98070 206.463.1855
AlliedSignal PO Box 97001 Redmond, WA 98073 425.885.8702

[Anthony Gill]

Id appreciate any insights from other folks who have struggled with this.
Like goodwill and branding this is not a easy topic to tackle. I
believe that your first task is to agree a workable definition on What
is organization culture? You may be surprised at how many different
answers you get within the same organization. That will then guide you
through the rest of the process. It could turn out to be a political
minefield as you start to surface aspect about possible power bases.

The pragmatic definition of Organization Culture I find useful is: The
way things happen / are done around here. This is refering to
leadership / management style (command-and-control v empowerment or
usually somewhere in between); does the organization structure
facilitate or hinder good working relationships; how are roles defined;
how work are processes carried out; what drives promotion up the
corporate ladder; how is the ladder defined; the impact of reward
systems; the role of status (who gets the perks and why?); office layout
(who is closest and furtherest from the seats of power and why?); dress
code; tolerance of mistakes made by people at all levels; how people
learn (not only from mistakes); team working or lack of it ie, its not
my job to do that; attitudes to expense claims (is there a will I get
away with that? mentality; what motivates everyone; attitutes towards
customers and other employees; do it right first time versus someone
else can sort it out mentality; etc etc. As we know these are all
interconnected in some way. Organization culture is a unique facet of
every organization.

Will, your problem is to pick the key drivers of organization culture (I
would suggest that these are likely to be slightly different for each
organization), identify how they are related and find appropriate ways
of measuring soft variables or their proxy measures. Naturally these to
some extent will be determined by the purpose of the larger model.
Probably best to keep it simple and avoid getting caught up in the
politics. Clever design of the measures (dont state the obvious problem
should it exist) will get people thinking. That is success in my book.


Tony Gill phone: +44 (0)1295 812262
Phrontis Limited
Beacon House fax: +44 (0)1295 812511
Horn Hill Road
Adderbury email: TonyGill@phrontis.com
Banbury
OXON. OX17 3EU URL: http://www.phrontis.com/
U.K.