Organizational Learning from Performance Feedback_1 pdf

Organizational Learning fromPerformance Feedback A Behavioral Perspective on Innovation and Change Revisiting Cyert and March’s classic 1963 Behavioral Theory of the Firm, Henrich Greve

This page intentionally left blank

Organizational Learning from

Performance Feedback

A Behavioral Perspective on Innovation and Change

Revisiting Cyert and March’s classic 1963 Behavioral Theory of the Firm,

Henrich Greve offers an intriguing analysis of how firms evolve inresponse to feedback about their own performance Based on ideasfrom organizational theory, social psychology, and economics, he ex-plains how managers set goals, evaluate performance, and determinestrategic changes Drawing on a range of recent studies, including theauthor’s own analysis of the Japanese shipbuilding industry, he reports

on how theory fits current evidence on organizational change of risktaking, research and development expenses, innovativeness, investment

in assets, and market strategy The findings suggest that high-performingorganizations quickly reduce their rates of change, but low-performingorganizations only slowly increase those rates Analysis of performancefeedback is an important new direction for research and this book pro-vides valuable insights in how organizational learning interacts withother influences on organizational behaviour such as competitive rivalryand institutional influences

   is Professor of Strategy at the Norwegian School

of Management BI, Norway

Organizational Learning from Performance Feedback

A Behavioral Perspective on Innovation

and Change

Henrich R Greve

  

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São PauloCambridge University Press

The Edinburgh Building, Cambridge  , United Kingdom

First published in print format

Information on this title: www.cambridge.org/9780521818315

This book is in copyright Subject to statutory exception and to the provision ofrelevant collective licensing agreements, no reproduction of any part may take placewithout the written permission of Cambridge University Press

isbn-10 0-511-06982-0 eBook (EBL)

isbn-10 0-521-81831-1 hardback

isbn-10 0-521-53491-7 paperback

Cambridge University Press has no responsibility for the persistence or accuracy of

s for external or third-party internet websites referred to in this book, and does notguarantee that any content on such websites is, or will remain, accurate or appropriate

Published in the United States of America by Cambridge University Press, New York

Contents

2.1 Organizational decision process page 15

3.1 Performance-based adjustment of search and decision

3.2 Possible reactions to performance feedback 613.3 Determinants of response to performance feedback 643.4 Risk as a function of cumulative resources 68

4.4 Format change in response to performance feedback 1194.5 Innovative, satellite, and production change 1206.1 Strategic change with homogeneous capabilities 1626.2 Strategic change with heterogeneous capabilities 164

vii

As a reader of books I always thought the acknowledgment section was apolite gesture to friends and loved ones Now I know that in the course ofwriting a book, one accumulates numerous debts that should be acknowl-edged here I have also found that two seemingly symbolic phrases found

in acknowledgment sections are completely true: Many have helped me,but the remaining flaws are mine only I can only mention those who havehelped directly by reading and commenting on the manuscript, whichmeans that the intellectual debts to researchers whose ideas I have usedare not mentioned here The reference list gives some indication of howmuch I have gained from the work of other researchers

The book owes its existence to James G March, who sent me an emailsaying it was time to write a book reporting the findings on aspiration-levellearning that others and I have made He also supervised the research ondiffusion that started my career, and supplied the idea and funding forcollecting the data on market shares that formed the basis of my firstpaper on aspiration-level learning

Several careful readers have helped me improve the book PinoAudia, Phil Bromiley, Hitoshi Mitsuhashi, Martin Schulz, Zur Shapira,and the referees for Cambridge University Press made challenging anddevelopmental comments on draft chapters Some of their commentscould not be taken care of in this book, but are a valuable cache of ideasfor future work My first paper on aspiration-level decision making ap-

peared in Administrative Science Quarterly, and I am grateful to the editor

Christine Oliver and three anonymous reviewers for helping sharpen the

arguments and noting issues that needed further work Also at ASQ,

Linda Johansen’s editing and advice on style in that and other papershas greatly improved my writing My work on performance feedback

in the shipbuilding industry has benefited from the comments of ChrisAhmadjian, Giovanni Dosi, Raghu Garud, Don Hambrick, Paul Ingram,Theresa Lant, Steve Mezias, Reinhardt Selten, and anonymous review-ers Chris Harrison at Cambridge University Press provided valuableeditorial advice

ix

x Acknowledgments

As all authors do, I owe a great deal of gratitude to my supportivefamily, especially my wife Takako, but also my children Jan and Ryo.Still, I want to dispel the myth that book writing unavoidably means atemporary absence of normal family life I recently asked Jan whether heknew that papa had written a book, and he said no It is possible to dosuch work within nearly normal work hours with some discipline and asupportive school

The data collection has been helped by a string of extremely capableresearch assistants At Stanford University, David Barba coded the radioratings data, and other radio data were coded by Joe, Mary, and Mireyah

At the University of Tsukuba, Shunsuke Iriguchi, Toshinobu Iriguchi,Masanori Osame, and Lisa Shimizu coded the shipbuilding data.The data collection on radio stations was supported by a grant to mefrom the Stanford Center of Organizational Research and a grant to James

G March from the Spencer Foundation The data collection on building was supported by a grant from Japan’s Ministry of Education(administered by the Japan Research Foundation)

ship-1 Introduction

This book is about how organizations react to performance feedback Itpresents a theory of organizations learning from their experience by col-lecting performance measures, creating aspiration levels based on theirown past performance or that of other organizations, and changing orga-nizational activities if the performance is lower than the aspiration level.The mechanism is one of simple self-regulation by attempting to reach agoal not currently met but not seeking, in the short run at least, to go fur-ther than the level that just achieves it Organizations with performancebelow the aspiration level of their managers have higher rates of strategicchange, R&D expenditure, innovation, and investment These activitiesinfluence the performance and risk of the organization, but otherwise theyhave little in common All are affected by the organizational performancebecause managers are willing to try a wide range of strategic actions tosolve a problem of low performance

We can see this reaction to performance feedback reflected in the havior of individual firms After the Japanese car makers had great suc-cesses in the 1980s US auto market, General Motors was still the world’slargest auto maker and the dominant firm in the USA It was doing lesswell than it had in the past, however, with its domestic market share incars falling from 49% in 1980 to below 40% in 1987.1During this period,General Motors implemented a remarkable series of projects to make upfor the perceived shortfall It continued a massive investment program

be-in its factories that had been announced be-in 1979 and aimed to makeGM’s manufacturing more automatized than that of any other car maker.This program would eventually cost $40 billion, making it perhaps thelargest non-government investment program in history GM started col-laborative manufacturing with Toyota in the now-famous NUMMI plant,and took equity positions in foreign car makers such as Suzuki, Isuzu,Nissan, and Daewoo GM supported this push into Asia by building a

1 This paragraph is based on information in three Harvard Business School cases (Badaracco 1988; Green 1993; Keller 1994).

1

2 Organizational Learning from Performance Feedback

Japanese-style supply network complete with equity positions in key pliers and a supplier association, departing from its usual practice of ob-taining supplies internally or from competitive bidding It made the newcar brand Saturn, which was managed by a subsidiary that incorporatedseveral innovative design, manufacturing, and marketing practices, andwas located in Tennessee, outside GM’s Midwest manufacturing belt.The facility investment program was initiated before the falling marketshare had become a palpable problem, but was continued unwaveringlyafter the reduced sales might have suggested that it would lead to excesscapacity The other change activities were initiated after the fall of mar-ket share had become a problem, and seemed to be ways of searching forsolutions to it In particular, both Saturn and the Asian alliances focused

sup-on the small car market, where General Motor’s market share decline wasparticularly pronounced

General Motors is an extremely large organization, so the scale andscope of its change activities would be difficult for others to match Thebasic pattern of changing in response to disappointing performance iswell known across many industries and organizational sizes, however, so

it is clearly not special for GM Intel shifted its market strategy fromcomputer memories to microprocessors after finding itself losing the bat-tle over market share, and thus economies of scale, in each successivegeneration of computer memory (Burgelman 1994) In 1988, the smallJapanese company Nichia Chemical started research on blue LEDs (light-emitting diodes), a technology that had frustrated the development ef-forts of much larger firms, after having entered successive markets withsemiconductor products and found itself beaten by established competi-tors every time (Johnstone 1999) It would eventually become the firstcompany to commercialize a blue LED, and its success in developingthis technology is as remarkable as the fact that such a small companyattempted a research project with so high risk and expense in the firstplace

The routine of searching when the organization is doing poorly but notwhen it is doing well is a central part of managerial lore When the searchfor solutions succeeds it is called a “turnaround,” and it is a milestoneevent in the career of the responsible manager (Dumaine 1990) Whenthe firm fails after searching for solutions but not finding any that work,the search may be referred to as “floundering” and, with the benefit ofhindsight, seen as misdirected or futile (Saporito 1998) These post hocjudgments based on the outcomes obscure the similarity of the behavior:troubled firms seek to change (Bowman 1982), and since the result ofstrategic changes is nearly always uncertain, large gains or losses are bothpossible Turnaround and floundering are different post hoc evaluations,but they start the same way

Introduction 3

Searching for solutions when doing poorly is one side of the coin; theobverse side is the failure of successful organizations to search for ways toimprove This is called the “competence trap” (Levitt and March 1988)

or “paradox of success” (Audia, Locke, and Smith 2000), and a goodindicator of its prevalence is all the talk about the importance of con-tinuous improvement in the managerial literature Rigid adherence to ahigh-price, low-volume strategy with no licensing of the operating sys-tem proved to be Apple’s bane in the late 1980s (Carlton 1997), butthe immediate profits of this strategy were so large that management didnot consider its long-term consequences The strategy conceded so muchmarket share that Intel and Microsoft gained strong footing for launchingtheir Wintel challenge, leaving Apple with a long uphill battle for highermarket share which, as one might expect, it started after the performancefell Such lack of foresight is not a sign of unusual managerial ineptness,but seems common in firms that are doing well A well-known symptom

of competence traps is the late and tepid response of successful firms

to new technologies that threaten their market (Christensen and Bower1996; Cooper and Schendel 1976; Tushman and Anderson 1986).While the cases suggest a general pattern of changing in response to lowperformance, they leave many important details open to question Thefirst issue is what exactly is meant by low performance The feeling ofcrisis in General Motors was triggered by a fall in market share, but GMwas still the largest automaker in the US and the world by a wide margin.This was not good enough for GM’s managers, however, as the experi-ence of being the world’s dominant automaker since the 1920s (Carrolland Hannan 1995a) had left them expecting a higher market share thantheir competitors Similarly, GM’s profits were still high at the time thatmany of its change efforts started, but not as high as they had been Itturns out that there is no clear delineation of high and low performance

on the measures that managers use to evaluate their organizations, onlyrough rules of thumb Managers set their own standards for what level ofperformance is desired Such standards, which will be called aspirationlevels here, are influenced by the organization’s history and its competi-tors’ performance The mechanisms for adjusting aspiration levels are animportant part of research on performance feedback in organizations.The second issue is whether organizational responses to low perfor-mance are as strong as they should be There is ample evidence of orga-nizations failing to change even when their performance is low (Lorenz1994; Meyer and Zucker 1989; Starbuck and Hedberg 1977), contrary

to the suggestion that adversity spurs change (Ocasio 1995) Indeed,General Motors was criticized for passivity in spite of all the changes itmade in response to the fall in market share (Green 1993) Such criticismsometimes seems unfair, but it raises an important point Organizations

4 Organizational Learning from Performance Feedback

may change in response to low performance, but still not change enough

to solve the problem Whether organizations make enough changes or not

is a question of the functional form of the relation from performance toorganizational change The critique that organizations make insufficientchanges in response to low performance does not mean that they donothing, but rather means that organizational failure spurs change lesseffectively than organizational success reduces change As will becomeclear later in the book, improved performance will often cause the rate

of organizational change to drop by a considerable amount, but a rioration of performance of the same size results in a barely perceptibleincrease in the rate of change This asymmetry in the response to successand failure suggests that organizations react conservatively to negativeperformance feedback: managers seem willing to believe that all is welluntil they have been presented with strong proof to the contrary Orga-nizations and individuals have powerful defenses against radical changes(Hannan and Freeman 1977; Kuran 1988), and these make it possiblefor organizations to change without changing enough

dete-Third, one may wonder whether it makes sense for successful nizations to be inert Should managers “leave good enough alone” andonly fix the organization when it is broken? The case for recommendingchanges in successful firms is usually built on environmental changes,such as changes in markets and technologies Environmental changescan cause the competitive strength of a successful firm to erode if it doesnot adapt This argument is true, but it is limited to highly dynamic en-vironments A more general case for changing successful firms can also

orga-be made Managers of successful firms may have ideas for how to presstheir advantage so that the firm can become even more successful Theideas may be untried and risky, but so are changes done in unsuccessfulfirms Why are such ideas often rejected in successful organizations? Theanswer is that the same amount of risk is less appealing to managers ofsuccessful firms than managers of unsuccessful firms Later I will showthat this risk aversion in successful firms makes sense in some competitiveenvironments, but not in others

Finally, there is a question of how general the pattern of changing inresponse to low performance is Case studies are suggestive, but do notprove that performance feedback is a mechanism of change There are somany organizations in the world that it would probably be possible to findcases supporting any theory of why organizations change, including weirdtheories like sunspot cycles.2 To present a strong case for performance

2 There is a theory of sunspot cycles and economic activity It does not suggest that sunspot

cycles directly cause economic cycles, but rather that beliefs in economic cycles that follow sunspot cycles can cause them to happen through behaviors that cause the expectations

to be self-confirming.