Professional Documents
Culture Documents
I,.n the past, organization theory has been tivity. "Job enrichment" and "democratic manage-
mainly concerned with what might be called ment" are labels commonly applied to these
"organization for production" that is, with sys- emphases in organizational design. 2
tems that use the services of substantial numbers The human relations movement did not limit
of employees to generate, more or less con- itself to redesigning organizations to achieve the
tinuously, some kind of output or "product." 1 traditional goals of organizational effectiveness
The normative theory of organization, aimed at and efficiency. It also raised fundamental value
enhancing organizational efficiency and effective- questions of whether organizations should be
ness, traditionally paid special attention to two designed in these terms; or whether, on the other
problems: how to divide up the work for its hand, there should be a deliberate sacrifice of
efficient performance and in such a way as to keep effectiveness and efficiency in order to make work
the needs for coordination of the parts within itself a rewarding and enjoyable part of human
manageable bounds; and how to construct and life. This value issue is sometimes obscured in the
maintain mechanisms for coordinating the several literature when the assumption is made, implicitly
organizational parts especially authority mech- or explicitly, that "the happy employee is the
anisms. productive employee." That assumption tends to
Research on "human relations" in organi- prevail through the early writings on human
zations, beginning on a substantial scale in the relations; but has been frequently questioned or
1930s, turned attention in organizational design to attacked in recent years by the social critics of the
the linkage between the individual as organization New Left.
member and the pattern of organizational activity. I do not intend to enter into that controversy
The principal normative concern here was to here. I raise the issue because I think the debate
create organizational environments in which em- has been carried out on premises that are less and
ployees would be motiviated to join the organ- less valid in today's organizations, and will be still
ization, to remain in it, and to contribute less valid in the organizations of the future. The
vigorously and effectively to its goals. As a result attack on work in organizations as dehumanizing
of theory and empirical research in human rela- generally takes as its model of the organization a
tions, the factory and office came to be viewed as system engaged in mass repetitive processing of
relatively impoverished human environments- materials or symbols the assembly line or a room
starving both The human mind and the human full of clerks or draftsmen. Chaplin's movie,
emotions. Proposals were advanced for the re- "Modern Times," exaggerates only slightly the
design of organizations in ways that would make portrait of the factory given in the human rela-
work intellectually more stimulating (or less bor- tions literature.
ing), and give the worker a greater feeling of But with the introduction of highly automated
participation in the decisions governing his ac- machinery, and particularly with the introduction
MAY/JUNE 1973
APPLYING INFORMATION TECHNOLOGY 269
of mechanized information processing equipment, of the same economic activity, first viewed as
the assembly line becomes a rather rare form of goods-producing activity, then as as a service-pro-
organization of production, as does the repetitive ducing activity: that is, producing bouses and
unautomated clerical process. The human opera- bousing respectively.
tive or clerk is less often a cog in the ongoing A house is a tangible commodity that can be
production process, or even the first-level control- manufactured and distributed through the usual
ler of that process. He is more and more an market mechanisms; housing is a bundle of services
observer, moderator, maintenance man, and repair- provided by a dwelling in the context of a
man for a nearly autonomous process that can neighborhood, with schools, streets, shopping
carry on for significant intervals of time without facilities, and a pattern of social interaction among
direct human intervention. the inhabitants. However complex it may be to
The question of whether "the happy employee define the qualities of a house, narrowly conceived
is the productive employee" still needs asking as a structure, it is far more complex to define the
under these new conditions of employment. But it qualities of housing, in the sense of a situation that
is not obvious that the question will have the same creates and supports a pattern of social activity.
answer it had in the past. The problems of Related to the tendency of organizations in
cognitive impoverishment and alienation may be our society to broaden the definition of their goals
replaced by quite new problems or may disappear from the production of tangible commodities to
altogether. My own assessment based on direct the production of bundles of services that may or
observation and the studies of others of highly may not be associated with tangible commodities,
automated work situations is that the new environ- is a tendency to broaden their concern for the
ment will be, for most workers, a more pleasant externalities associated with their activities. Ex-
and humane environment than the old. We will ternalities are simply those consequences of action
need a great deal more experience with this new that are not charged, through the existing market
environment, however, to provide solid evidential mechanisms, to the actors. The classical example is
foundations for that optimism. the factory smoke whose social costs have not
But I am getting a bit ahead of my story, for I generally been paid by the consumers of the
have not yet said why I think organizations are factory's product.
changing in the directions I have mentioned. It may be that organizations producing services
usually have more and larger externalities associat-
The Post-Industrial Society ed with their activities than organizations produc-
ing goods; it may be that we are simply becoming
Peter Drucker has used the phrase "post-indus- more sensitive, in our society to the indirect
trial society" to describe the emerging world in consequences of organizational activity directed
which manufacturing production, and the ac- toward specialized goals; it may be that, with the
tivities associated with it, plays a much less central growth of population and technology, the actual
role than it did in the world of the past century. interdependencies of organizations, and hence the
Organizations in the post-industrial society provide externalities they cause, are becoming more exten-
services, many of them intangible, more than they sive and significant Whatever the reasons and all
manufacture things. Already, a large part of the three of those mentioned probably contribute to
economic activity of our society consists in provid- the trend organizational decision making in the
ing services for education, health, and leisure-time organizations of the post-industrial world shows
activities. every sign of becoming a great deal more complex
Providing services tends to pose different than the decision making of the past. As a
organizational problen s from producing tangible consequeiice of. this fact, 'the decision-making
goods. It is usually more difficult to define process, rather than the processes contributing
appropriate output measures for service organi- immediately and directly to the production of the
zations than for organizations that produce tang- organization's final output, will bulk larger and
ible commodities. Whatever problems are present larger as the central activity in which the organi-
in measuring the quality of goods are magnified zation is engaged.
greatly in measuring the quality of services. The In the post-industrial society, the central prob-
point can be illustrated by comparing two versions lem is not how to organize to produce efficiently
MAY/JUNE 1973
270 PUBLIC ADMINISTRATION REVIEW
(although this will always remain an important Two Requirements of Organizational Design
consideration), but how to organize to make
decisions that is, to process information. Until The division of labor is quite as important in
recent years, decision making was exclusively a organizing decision making as in organizing pro-
human activity; it .involved processes going on duction, but what is being divided is different in
inside the hum in head and symbolic communica- the two cases. From the information processing
tion among humans. In our present world, decision point of view, division of labor means factoring
making is shared between the human and mech- the total system of decisions that need to be made
anized components of man-machine systems, the into relatively independent subsystems, each one
machines being those devices we call computers. of which can be.; designed with only minimal
The division of labor between the human and concern for its interactions with the others. The
computer components in these systems has division is necessary because the processors that
changed steadily over the past 20 years, and we are available to organizations, whether humans or
can expect it to continue to change as the computers, are very limited in their processing
sophistication of computer technology and capacity in comparison with the magnitude of the
particularly computer programming or "software" decision problems that organizations face. The
technology grows. number of alternatives that can be considered; the
The anatomy of an organization viewed as a intricacy of the chains of consequences that can be
decision-making and information-processing sys- traced all of these are severely restricted by the
tem may look very different from the anatomy of limited capacities of the available processors.
the same organization viewed as a collection of Any division of labor among decisional sub-
people. The latter viewpoint, which is the tradi- systems creates externalities, which arise out of
tional one, focuses attention on the groupings of the interdependencies among the subsystems that
human beings that is, the departmentalization. are ignored. What is wanted is a factorization that
The former viewpoint, on the other hand, focuses minimizes these externalities and consequently
on the decision-making process itself that is, upon permits a maximum degree of decentralization of
the flows and transformations of symbols. If we final decision to the subsystems, and a maximum
carve an organization, conceptually, into sub- use of relatively simple and cheap coordinating
systems on the basis of the principal components devices like the market mechanism to relate each
into which the decision-making process divides, we of the decisional subsystems with the others.
may, and probably will, arrive at a very different Not only must the size of decision problems
dissection than if we carve it into its departmental handled by organizations be reduced to manage-
and sub-departmental components. Moreover, the able proportions by factorization, but the number
greater the interdependencies among the depart- of decisions to be processed must be limited by
mental components, the greater will be the dif- applying good principles of attention management.
ference in these two ways of conceptualizing the Attention management for an organization means
organization. exactly what it means for an individual human
Both of these viewpoints are useful and even being: processing capacity must be allocated to
essential in arriving at sound designs for organi- specific decision tasks, and if the total capacity is
zations. In this analysis, I shall emphasize the less not adequate to the totality of tasks, then priori-
conventional point of view and shall discuss the ties must be set so that the most important or
decision-making process disembodied, so to speak, critical tasks are attended to.
from the flesh-and-blood (or glass and metal, as The information-processing systems of our
the case may be) decision makers who actually contemporary world swim in an exceedingly rich
carry out this process. Instead of watching a man soup of information, of symbols. In a world of this
or computer as information reaches him and he kind, the scarce resource is not information; it is
processes it and transmits new information in his processing capacity to attend to information.
turn, we will watch information as it flows from Attention is the chief bottleneck in organizational
one man or computer to another and is transform- activity, and the bottleneck becomes narrower and
ed in the course of flow. This approach, if it has narrower as we move to the tops of organizations,
no other advantages (though I believe it does), will where parallel processing capacity becomes less
give us a fresh look at the design of organizations. easy to provide without damaging the coordinating
MAY/JUNE 1973
APPLYING INFORMATION TECHNOLOGY 271
function that is a prime responsibility of these disenchantment that followed the initial attempts
levels. at constructing such systems. First, the data
The richness of the informational environment processing and data storage tasks proved much
and the scarcity of attention have many conse- larger and more complex than had been imagined.
quences for organizational design, some of which Perhaps more crucial, it became less and less clear
will be developed presently. At this point, only a just how the data were to enter into the decision-
couple of further comments need to be made. making process, or indeed to just what decisions
First, the difficulty of coping with the infor- they were relevant.
mation-rich environment is compounded by the fact The lesson is clear. There is no magic in
that most information relevant to top-level and "comprehensiveness." It may be sufficient motive
long-run organizational decisions typically origi- to climb a mountain "because it is there," but the
nates outside the organization, and hence in forms mere existence of a mass of data is not a sufficient
and quantities that are beyond its control. This reason for collecting it into a single, comprehen-
means that the organization must have an "inter- sive information system. Indeed, the problem is
face" for ingesting such information selectively, quite the opposite: of finding way of factoring
and for translating it into formats that are compat- decision problems in order to relate the seveial
ible with its internal information flows and sys- components to their respective relevant data
tems. sources. Analysis of the decision-making system
Second, if attention is the scarce resource, and its data requirements must come first; only
then it becomes particularly important to distin- then can a reasonable approach be made to
guish between problems for decision that come defining the data systems that will support the
with deadlines attached (real-time decisions), and decision-making process.
problems that have relatively flexible deadlines. The history of management information
Rather different system designs are called for to systems has been nearly the same as the history of
handle these different kinds of decisions. municipal data banks. In the enthusiasm to make
In summary, the inherent capacity limits of use of the enormous power of computers, there
information-processing systems impose two re- was a tendency, in designing such systems, to take
quirements on organizational design: that the the existing source records as starting point and to
totality of decision problems be factored in such a try to give top management access to all this
way as to minimize the' interdependence of the information. The question was not asked, or not
components; and that the entire system be so asked with sufficient seriousness, whether top
structured as to conserve the scarce resource, management either wanted or needed such infor-
attention. The organizational design must provide mation, nor whether the information that top
for interfaces to handle information that orginates management needed and should want could in fact
outsideo the organization, and special provision be derived from these particular source records.
must be made for decisions that have particular The systems were not designed to conserve the
time limits associated with them. critical scarce resource the attention of managers
Applying these basic design requirements and they tended to ignore the fact that the
makes it easy to see the fallacy in some recent, and information most important to top managers
more or less abortive, approaches to the improve- comes mainly from external sources and not from
ment of information systems: municipal data the internal records that were immediately acces-
banks, and management information systems. sible for mechanized processing.
There was great enthusiasm, only a couple of years Thus many of the efforts to design infor-
ago, for developing comprehensive data banks for mation systems for municipalities and corpora-
metropolitan areas these data banks to incorpor- tions fell into the fallacy of thinking that "more
ate in a single system all of the myriad pieces of information is better." They took over, implicitly,
information about land and its uses, and about the assumptions of a past society where informa-
people and their activities that are generated by tion rather than attention was the scarce resource.
the operations of urban government. As the result
of servcral attempts to construct such systems, the Characteristics of the Technology
enthusiasm has been much moderated, and several
incipient undertakings of this kind have been Good design requires bringing the desired ends
abandoned. There were several reasons for the into effective relation with the available means. To
MAY/JUNE 1973
272 PUBLIC ADMINISTRATION REVIEW
design effective decision-making organizations, we our other information systems and place sole
must understand the structure of the decisions to reliance on the telephone and the vast distributed
be made; and we must understand the decision- memory with which it connects us. I am even a
making tools at our disposal, both human and little reluctant to mention this particular tech-
mechanical men and computers. nique publicly, because if it attains general popu-
The Human Components larity, we will all be busy answering the phone,
and the trick will be spoiled for those of us who
In our fascination with the new capabilities use it extensively. It does provide, however, a
that computers offer us, we must be careful not to useful illustration of how we must think about
forget that our human decision makers have some information-processing systems including human
pretty remarkable qualities too. Each human systems their components and interconnections,
decision maker is provided with a sizeable memory if we are to design them well. We must learn to
that is stocked cumulatively over a long period of characterize them in terms of the sizes of their
years with various kinds of relevant and irrelevant memories, the ways in which those memories are
information and skills. Each is able to com- indexed, their processing rates, and the rapidity
municate in natural language with his fellows, with.which they can respond. The human compon-
either in direct face-to-face settings or by remote ents of information systems are just as describable
devices like the telephone. as the machine componets, and since World War II
Suppose, for example, that we were interested we have, learned a great deal, through
in designing an organization that would lead us to psychological research, about the parameters of
the most expert source of information in the the human system.
United States about any particular question that Our new and growing understanding of infor-
happened to arise. Now this expert information is mation processing enables us to look at familiar
stored both in human heads and in books. More- processing systems man and telephone in new
over, the information in books is also indexed in ways. It also introduces us to new kinds of
human heads, so that usually the most expeditious systems, which we put under the general label of
way to find the right book is to ask a human who "computers," that have capabilities of the most
is an expert on the subject the book deals with. varied kinds.
Not only are books indexed in human heads,' but
people are also. Taking these resources into ac- The Computer as Memory
count, the most powerful information processing The computer is, first of all, a memory. I have
system for carrying out this search task is the already expressed my qualms about confusing the
aggregate of memory that is distributed among design of an information-collecting system with
200 million human heads, together with the the design of an information-processing system.
telephone system that links these distributed The fault, of course, is not in collecting informa-
memories. On receipt of the inquiry, I pick up the tion (although that may be costly in itself); it is in
phone and call the person, among my demanding the scarce attention of decision makers
acquaintances, whose field of expertness is as close to the information that has been collected. Mem-
as possible (it need not be very close at all). I ask ories, as components of information-processing
him, not for the answer to the question, but for systems, need to be viewed as stores of potential
the name of the person in his circle of acquain- information, which, if indexed effectively, can
tance who is closest to being an expert on the become available at a reasonable cost whenever it
topic. I repeat the process until I have the is needed as input to a decision-making process.
information I want. It will be a rare instance when Consider a man who has collected a library of
more than three or four calls are required. 30,000 books. Even if he reads one book a day a
Suppose that the question is whether whales pretty good clip it will take him 100 years to read
have spleens. (I can't imagine why we want to through all the shelves. We may even consider it a
know, but this example is as good as any other.) I bit ostentatious of him to have collected more
call a biologist, who refers me to an ichthyologist, books than he can possibly read as though he
who refers me to a specialist on whales, who either were trying, ro impress us with his learning.
knows the answer or can refer me to the book However, we must not be too hasty in judging
where I will find it. him. If his library is properly indexed, then our
I don't mean to propose that we junk all of collector has potential access to any of the
MAY/JUNE 1973
APPLYING INFORMATION TECHNOLOGY 273
MAY/JUNE 1973
274 PUBLIC ADMINISTRATION REVIEW
although the need for them has been fairly obvious man-readable version, and at an insignificant
for some years. The tardiness of response to the incremental cost. Hence, we may look forward to
need is evidence both ofothe novelty of the a time in the near future when the written word
modelling technology and the novelty of looking will be almost universally available in both mach-
at organization as a collection of decision systems ine-readaole and man-readable editions. Once
MAY/JUNE 1973
APPLYING INFORMATION TECHNOLOGY 275
severe handicap for almost any individual mem- informed participation have not gone unnoticed.
ber of the audience, we could devise a lecture that Informed participation in decision is a problem
would be more appropriate than the one actually not only for the part-time contributors to the
screened. In communicating information through system the voters but for elected representatives
pictures it has the further disadvantages of its and high-level administrators as well. It is at the
inability to abstract, to generalize, or to sample core of the familiar organizational problem of the
properly from complex populations of events. It relation of experts to laymen.
possesses no analytic capabilities other than those
An Example of Poorly Informed Participation
provided by the commentator.
To point to these severe deficiencies of The phosphate detergent fiasco is only the
television as a source of politically relevant infor- most recent of innumerable instances of the
mation is not to deny its effects on the political difficulties of settling technical issues with wide
system. The most obvious of these effects is its public involvement. A problem existed: excessive
power to focus the attention of an entire society algal bloom in lakes. It was known that this
for a period of time on a particular set of condition only incurred in waters having a high
events whether these be trips to the moon, riots, level of organic nutrients; phosphate, in particular,
or summit conferences. In general, television in was highly suspect as culprit. (Clear scientific
forms very little about such events, but it may evidence on the precise mechanisms of
succeed in evoking strong emotions simultaneously eutrophication, and the precise conditions under
from millions of viewers, and consequently in which it will or won't occur is nearly non-
rearranging the public agenda. existent.) A substantial fraction of the phosphate
in water is contributed by household detergents.
Politics as Information Processing There exist non-phosphate or low-phosphate deter-
gents. Therefore but the "therefore," however
This last example, as well as the earlier obvious it may seem, left unanswered most of the
example of energy policy, are illustrative of a critical questions: Would removal of phosphate
broader point: that our political institutions are from household detergents do any good if other
organizations, and that what we have said about phosphate sources remained? Was phosphate, in
the design of information-processing systems for fact, the principal culprit? What are the alternative
organizations applies fully to the design of the methods, other than changing' detergents, for
decision-making components of the political eliminating phosphate from lakes? What are the
system. Nowhere is the problem of attention properties of the alternative detergents, and what
management and the conservation of attention of consequences would their use have? What would
greater importance than in the political process. be the relative costs of various courses of action?
The themes of alienation of the electorate, and As our previous analysis should lead us to
the need to "return government to the people," predict, the mass media had a large impact in
which have been prominent in recent public focusing public attention on the eutrophication
discussion of our institutions are, of course, not problem, and in creating insistent demands for
new to the American political scene. Cynicism foreceful action. There is no difficulty in picturing
about the political process is deeply ingrained in on television a dead fish or (at least on color
the American culture. For the most part, proposals television) an algae-green lake. What the mass
for political reform have focused on strengthening media could not do was to remove the scientific
controls over the professional full-time unknowns from the problem, or provide a systems
participants in the system, elected or appointed analysis of alternatives. As a result of the pressure
(e.g., recall elections, presidential primary), or on for action, the federal government came close to
creating new channels of direct popular participa- banning phosphate from detergents, but unfor-
tion in decision making (e.g., the initiative and tunately had not done so by the time some of the
referendum, official provision for citizen difficulties and dangers of the alternatives had
participation in program administration). become known.
As high technology has come to play a larger My intent here is not to suggest what the
role in our society, and in the provision of solution of the problem should have been, but to
governmental services in particular, the growing use these events to point to the weaknesses of our
difficulties of assuring that participation will be current decision-making processes in political mat-
MAY/JUNE 1973
276 PUBLIC ADMINISTRATION REVIEW
tcrs where complex technical issues are involved. Thus, the ABM and SST decision processes
Most of the current proposals to remedy these were informed by analyses that met three essential
weaknesses are of a traditional kind: create new criteria: comprehensiveness, technical sophistica-
organizations to provide the technical expertise tion, and pluralism. Satisfaction of these criteria
that is now lacking. The two most prominent does not guarantee that correct decisions are
proposals are the creation or strengthening of reached in a world of uncertainty and conflict of
organizations that would represent the consumer interest, that guarantee can seldom be provided. It
and the creation of organizations for technology does tend to assure that the decisions reached are
assessment, these latter to be responsible to the such as reasonable men of good will could arrive
legislative body. at. And it is the guarantee of due process, not the
So far as these proposals go, they seem to me guarantee of infallibility, that democratic institu-
sound, but 1 would warn that we should not tions require.
expect too much from them. In particular, in In making this plea for more and better systems
talking about "consumer organizations," we must analyses, and more and better institutional ar-
remember that that is precisely what a government rangements for carrying them out, I should not
is created to be, and thus new organizations of this want to be thought to be urging centralization of
kind are no more exempt from the iron law of any and all decisions. An equally important
bureaucracy than are existing institutions (as direction for improvement of decision processes is
experience with trade unions, political parties, and to instill new vigor into market mechanisms as an
the American Medical Association should have essential component of a decentralized society.
taught us). Our growing awareness of the externalities, partic-
The Contribution of Systems Analysis ularly environmental costs, that the price system
has ignored in the past has caused something of a
If magic is to be performed, it will not come crisis of confidence in market mechanisms as social
through the mere creation of new organizations. regulators and decision makers. Where externalities
The problem is not primarily one of control but of are present, there is an alternative, however, to
information not one of enforcing virtue but of centralization: incorporating the externalities in
discovering-jhe path of virtue. We do not need new the pricing process. A tax on oxides of sulphur
organizations so much as we need new decision emitted into the atmosphere is an example of such
processes. To illustrate my meaning, let me cite a measure. To attack such a proposal as "a license
some happier examples than the phosphate debate to pollute" is to substitute sloganizing for problem
of recent public discussion. solving. We need not only to solve the problems
The struggles over the anti-ballistic missile and we perceive, but, in solving them, to minimize the
the supersonic transport are, to my mind, ex- drain on our scarce resources. Using the price
amples of what we may hope for in the way of system as motivator, where it is applicable, is one
informed discussion of highly technical and com- of the most powerful means available to us for
plex issues. This does not mean that the correct enlisting a multitude of energies and minds into
decisions were necessarily reached. I have no more the search for least-cost solutions to our problems.
infallible means for deciding that than did the
disputants at the time of the debate. Honest and Conclusion
reasonable men could and did take either side of
either question. But what distinguished these The major problems of governmental (and
particular debates was that both sides were armed corporate and educational) organization today are
with sophisticated analyses based on man-years of not problems of departmentalization and co-
careful study supported by quantitative models. ordination of operating units. Instead, they are
For this reason, it was possible for the layman, problems of organizing information storage and
with a reasonable expenditure of time, to under- information processing not problems of the
stand where the differences lay which disagree- division of labor, but problems of the factorization
ments about assumptions were responsible for the of decision making. These organizational problems
divergent conclusions reached. Moreover, for each are best attacked, at least to a first approximation,
of the decisions there was not a single analysis but by examining the information system in abstrac-
several, prepared by protagonists that had dif- tion from agency and department structure.
ferent sets of interests and different viewpoints. With the rapid development of information-
MAY/JUNE 1973
APPLYING INFORMATION TECHNOLOGY 277
processing technology, the corporate and public puters and systems design tools in organizations
decision-making processes are becoming immense- will be found in the last half of The Shape of
ly more sophisticated and rational than they were Automation (Harper & Row, 1965) and in chapter
in past eras. If we require any proof for this, we 3 of The Sciences of the Artificial (M.I.T. Press,
need only compare the ABM debate (regardless of 1969).
whether we like its outcome) with any debate on 1 have proposed a taxonomy of species of
the Acropolis reported by Thucydides or, for that information in my essay "Research for Choice," in
matter, with any debate in the U.S. Congress in W. R. Ewald, Jr. (ed.), Environment and Policy:
the first half of this century. The Next Fifty Years (Indiana University Press,
With the development of information-proces- 1968, pp. 360-380), and have drawn out the
sing technology, we have a growing capacity to implications of attention scarcity for technology
consider interactions and tradeoffs among alter- assessment in "Designing Organizations for an
natives and consequences; to cumulate our under- Information-Rich World," published in Manin
standing of fragments of the whole problem by Greenberger (ed.), Computers, Communications,
embedding these fragments in comprehensive and the Public Interest (Johns Hopkins Press,
models. 1971, pp. 37-52).
Barbara Ward and others have pointed out to
us that the largest crises in our world today are Notes
crises of aspirations. The population problem is as
old as Man. What is new about it today is that we 1. It takes a bit of stretching to include military
are resolved not to accept a gloomy outcome, but organizations in this category of "production organiza-
to deal with it. For centuries, Man's actions have tions," and regulatory bodies do not really fit at all.
been creating all kinds of unintended and unex- However, most of the literature, empirical and theoret-
ica, of organization theory in recent years has focused
pected consequences. He could live in good con- on industrial organizations and governmental agencies
science with his actions to the extent that he was that provide public services, both of which categories
unaware of these consequences. Today, we can do fit. I shall make no attempt here to discuss
trace minute and indirect effects of our behavior: regulatory agencies. Military organizations, however,
the relation of smoking to cancer, the relation of have a much larger component of information pro-
cessing in their total pattern of activity than do, say,
the brittleness of eagles' eggs to the presence of typical manufacturing organizations. Hence it is proba-
DDT in the environment. With this new ability to bly no accident that the concern with organizing for
trace effects, we feel responsible for them in a way effective information processing first made its appear-
we previously did not. The intellectual awakening ance on a large scale in military organizations, and that
is also a moral awakening. they have been and still arc the initial testing grounds
'The new problems created (or made visible) by for most modern information processing technology.
2. Theb names of Lewin, Roethlisberger, Likcrt, Mac-
our new scientific knowledge are symptoms of
Gregor, and Argyris will identify the general range of
progress, not omens of doom. They demonstrate approaches and emphases within the human relations
that Mankind now possess the analytic tools that movement. Many other names could, of course, be
are basic to understanding his problems basic to added to the list, which is only illustrative.
understanding the human condition. 3. One evidence of the degree of novelty of the com-
puter's capabilities is the resistance it evokes from
Of course, to understand problems is not those who refuse to see in it anything more than an
necessarily to solve them. But it is the essential enlarged desk calculator. Not since the Darwinian
first step. The new information technology that controversy of the past century have we seen such a
we are creating enables us to take that step. passionate defense of the uniqueness of man against
claims of kinship by systems that don't belong to his
species.
Some Further Reading 4. We have., now had a generation's experience with
decision models for economic policy. The construction
and testing of such models in the United States has
The premises about the nature of organizations been carried out in considerable part by nongovern-
upon which this essay rests are developed and mental agencies the Cowics Foundation for Research
in Economics, and The Brookings Institution, for
documented in Administrative Behavior (Macmil- example. Since the day when President Nixon declared
lian, 1947, 1957) and in the two final chapters of himself to be a Keyncsian, their impact on government
March and Simon, Organizations (Wiley, 1958). decisions could no longer be in doubt, although the
Fuller discussions of the potentialities of com- impact certainly preceded that declaration by a decade
MAY/JUNE 1973
278 PUBLIC ADMINISTRATION REVIEW
1313 EAST 60TH STREET CHICAGO. ILLINOIS 60637 2830 Mary St., Falls Church, Va. 22042
(312) 324 3400 Tel. 703 560-6900
CHICAGO • WASHINGTON
'New
Implementation
Jeffrey L. Pressman and Aaron B. Wildavsky
Most studies of policy-making focus on the formulation of policy. Less well known are the
hazards involved in implementing public policies. Unique in its focus on implementation,
this volume examines the period after the goals have been set the problems arising from
the divisive funding system as well as a host of disagreements over the "details" of imple-
mentation.
LC: 72-78949 225 pages $7.50
from California
University of California Press Berkeley 94720