MANAGEMENT CONTROL SYSTEMSPerformance Measurement, Evaluation and Incentives

Second Edition

Kenneth A, MerchantUniversity of Southern California

Wim A. Van der StedeLondon School of Economics

Lffir

Prentice HallFINANCIAL Th,tES

An impriil of P eatson EducationHarlow, England . London ' New York . Boston . san Francisco . Toronlo

Sydney. Tokyo . Singapore. Hong Kong .Seoul. Taipei. New DelhiCape Town . Madrid . Mexico City . Amsterdam ' Munich . Paris. Mian

"@@64wrw

MANAGEMENT AND CONTROL

\ fianagemenr conrrol is a critical function in organizations. Management contlolIYlfaitures can lead to large financial losses, r'eputation damage, and possibly even to

organizational failure. Here are some recent examples:

o In Aplil 2005, employees at the 75-year-old California-based not-for'-proirt Gemological

Institute of America (GIA), the world's largest grader of diamonds, were accused ofaccepting bribes fi'om large diamond dealers to inflate diarnond grades. Large diamond

dealers rvouid submit proportionally high bids, often 20 to 30qa highel than prevailing bids

fol lough stones. knowing that they would be able to sell these stones at a profit because

they bribed GIA staff to get a higher-than-deserved grade. A small differ-ence in grade

can mean a huge difference in price, often hundreds of thousands of dollars on larger dia-

monds. The size of the blibes is unknown, but the probe into the allegations mentions cash,

theatel tickets, and other gifts. What is known, however, is that the blibes gave the large

dealers enough of a financial edge to control the market and reap excess profits. As such,

the scandal reverberated thloughout the $80 billion diamond-jewelry industry around the

world, as many customels overpaid for their diamor.rds and many diamond dealers, particu-larly srnaller ones, were forced to leave the industry or were considering it.'

o In 1995, venerable UK bank Baling Brothers, founded in 1817, declared bankruptcy. The

bankruptcy was caused by losses on unautholized trades of futules contracts rnade by a

Singapole-based trader named Nicholas Leeson. Betbre his trades were stopped, Leeson'slosses totaled neariy $1.1 billion, more than t'"vice Baring Bank's capitaiization, A Bankof England investigation into the causes of the losses found major weaknesses in Baring'scontrol systems, including lack of segregation of duties. lack of position limits, and con-

fused lines of management responsibiiity.t When Baring failed, Leeson was two days sholtof reaching his twenty-eighth birthday .In2002, a simiiar case at Allied Irish Banks led tohuge currency-trading losses at its Baltimole unit, Allfirst Financial. Lack of adequate riskcontrols and lack of independent confirmation of trades left the bank wide open to fi'audu-

lentll, qengszled tradin-s losses of $691 miliion over a five-year period. The headline ofThe Wall Street Journal alticle suggested that "Lax Controls May Explain Trading Loss at

Allied Irish."r

o In 2001 a keystloke error by an employee at Lehman Brothels Holdings, Inc. in Londoncost the firm $6 million in trading losses. The erlor changed a customer's f30 million stock

sell order into one valued at about f300 million. Before the error was detected, stocks had

fallen sharply, and Lehman was liable for the damages caused by the error'. The error was

believed to be unintentional.r

c In 2002, a former National Atchives' empiol'ee admitted that he stole dozens of histor-ical documents between 1996 and 1999 simply by putting them in his briefcase. The

Chapter 1 . Management and Control

documents included auto-eraphed pictures of Apollo astronauts. plesidential paldons

signed by Abraharn Lincoln, and slai,e tl'ade materials. u,hich he sold to collectors formore than $200,000. Investi_qators were tipped off to the theft when a t-ederal workerbecame suspicious of an itern offered for sale online. Clearly, the National Archives and

Records Administration did not have good controls in piace.'

o In 2002, two clerical workels at the Laguna Niguei. California-based service center of theUS Immiglation and Naturalization Service (INS) were accnsed of destroying thousands

of ir-nmiglation documents, includin-g visa applications, passports. and other papefs.

According to the probe. tl.re clelks started shreddin-s unprocessed paperwork in early 2002after an inventory revealed a processin-9 backlog of about 90,000 documents. A monthlater, in March 2002, the backlo-s was repor-ted to be zero. The shredding allegedly rventon fbr about anothel'month to keep the backlog at zero, until INS officials discovered the

shredding spree durin-e an evenin-e shift.6 This exanrple illustrates that money is not alwaysthe motive for wrongdoin-q.

The examples described above show the impofiance of having good management con-tlol systems (MCSs) and the types of problems - thefts, fi'auds. and unintentional errors

- they can address.However, adding more controls does not aiways lead to better controi. Some MCSs

in common use often stifle initiative, creativity, and innovation. Here is an example of a

situation with a control-versus-initiative tensron:

In 2003, the European Union's (EU) antifi'aud office discor,ered an allegedly "vast enterpriseof looting" at Eurostat, the statistical service of the European Commission. The plobe focusedon secret birnk accounts in which seniol mana_eers at Eurostat allegedly funneled an estimated€900,000 of EU taxpal,ers' cash to contractors, inchrding companies that they themselves hadhelped set up, by artificially inflating the value of the contracts or by creating fictitious con-tracts. Some noted that this rvas just a confirmation of the popular plejudice that tlie "BLr.rssels

bureaucracy" is rife with col'ruption, lax financial controls, complacencv, and clonyisrn, a repu-tation that the European Commission had eamed in the late 1990s when several other con'up-tion scandals broke. Horvever, othels argued that it rvas not certain that the accounts set up

by the Eurostat officials lvere used for the personal enrichment of those involved, at least notinitially. They argued instead that these accounts may originally have been set up to giveEulostat a way to pay for Lesearch quickly without going through the Commission's cumber-some procedures. Ironicalll,, while the Cornmission has elaborate plocedules to prevent finan-cial fraud. these plocedures may not only have ploved insufficient, they may actually havemade the ploblem worse. Due to the tortuous fonn-filling that is required for fundin,e requests,the numbel of bureaucratic hoops fund requestels have to jump thlou_sh to get anythingapproved, and the notoriously slo,,r, delivery of the funds, comr.nission officials and staff mayhave got used to cutting corners and finding "creative" ways to speed up the plocess. But e\/enthough there n.right be a strong suspicion that the seclet accounts $/ere at first intended to seleiegitimate purposes, they may have been abused as time went on. While the jury was out on thevalidity of the conjectures on each side of the argument, some argued that perhaps the mostessential problem at the Commission was its lack of autlture of responsibilitl'.1

It is wideiy accepted that good MCSs are important. Understanding and comparing theviews in the books and articles written on managernent control is difficult, holvever,becanse much of the MCS language is imprecise. The term "control," as it applies to a

management function, does not have a universally accepted definition. An old, narrowview of an MCS is that of a simple cybernetic system involving a single feedback loop,analogous to a tlrermostat. Thermostats include a singie feedback loop: they lneasure thetemperature, colnpare those rneasurements with the desired standarcl, and. if necessary,take a corrective action (turn on. or off. a fumace or air conditioner). In an MCS feedback

Management and control

loop, managers lleasure perfomtance, compal'e that lneasurement with a pleset perform-ance standard, and, ifnecessary, take corrective actions.

This book, however, like many other writings on management control, takes a broaderview.8 It recognizes that many management controls in common use, such as directsupervision, employee-hiring standards, and codes of conduct, do not focus on measured

perfonnance. They focus instead on encouraging, enabling. or, sometimes, forcingemplol,ees to act in the organization's best interest. This book also recognizes that solne

nanagenrent controls arc proactire, rather than reactive. Ploactive means that the con-

trols ale designed toprevent ploblerns before the organization suffers any adverse effectson perfolmance. Examples of proactive controls include planning processes, requiredexpenditure approvals, computer passwords, and segregation of ernployees' duties.Management control, then, includes all the devices or systems managers use to ensuretliat the behaviors and decisions of theil employees ale consistent with the organization'sobjectives and strategies. The systems themselves al'e commonly referred to as the

motlctgentent control st,.!/e,r?.t (MCSs),Desrgned properly, MCSs influence employees' behaviors in desirable ways and,

consequently, increase the probability that the organization will achieve its goals. Thus,the primarylrnctiort of management control is to influence behaviors in desirable ways.The bene.fit of management control is the increased probability that the olganization'sobjectives will be achieved.

MANAGEMENT AND GONTROL

Management control is the back end of the management process. This can be seen fromthe varior,rs ways in which the broad topic of management is disagglegated.e

Management

The management literature includes many definitions of management. All relate to theprocesses of organizing resources and directing activities fol the purpose of achievingorganizational objectives.

Inevitably, those who study and teach management have broken the subject intosmaller', more manageable elements. Table 1.1 shows the most prominent classificationschemes. The first column identifies the basic management functions: product (or service)development, operations (manufacturing products or performing services), malketing/sales (finding buyers and making sure the products and sen,ices fulfill customer needs),and finance (raising money). Virtually every managernent school offers courses focusedon only one, or only part of one, of these management functions,

Tnsrr 1.1 Different ways of breaking down the broad area of management intosmaller elements

Functions Besources Processes

Product (or service) developmenlOperationsMarketing/salesFinance

P6^^16

MoneyMachineslniormation

Objective settingStrategy lormulationManagement control

So!rce K A Nlerchant Mode,nManagementConlrotSyslems:TexlandCases(UpperSadd'e River. NJ: PrenliceHall,1998), p 3

Chapter 1 Management and Control

The second column of Table 1.1 identifies the major types of resources with whichmanagers must wort: people, money, machines, and information. Management schoolsalso offer collrses or-eanized using this classification. These courses are often called

human resource mana-qement. accounting and finance, production, and informatiortsystems, respectively.

The term monogement corttrol appeal's in the thrrd column of Table 1.1, which sep-

arates the management functions along a proce.rs continuum involving objective setting,strategy formulation. and management control. Control is the back end of the manage-ment process. Many management courses, includin-s business policy, strategic manage-ment, and management control systems. fbcus on eiements of the mana-eement process.

To focus on lnanagement control, we must distinguish it from obiective setting and strat-egy formulation.

0bjective setting

Knorvled-ee of objectires is a necessary plerequisite for the design of any MCS and,indeed, for any purposeful activities. Objectives do not have to be quantified and do nothave to be financial, such as 20o/o annual leturn on equity. A not-for-profit ol-ganization'sprimary objective might be to provide sJrelter tbl homeless people, for exampie. In any

organization. however, employees must have some understanding of what the organiza-tion is trvin-e to accomplish. Othelwise no one could claim that any of the employees'actions are purposive, and no one could ever support a claim that the organization was

s uccessful.In most organizations. the objectives at'e known. That is not to say that ali employees

always agree unaninrously as to how to balance their or-sanizations' r'esponsibilities to allof theil stakeholders (including owners, debtors, employees. suppliers, customers, and

the society at large). They ralely do. But earl1, in their histories, organizations developcomplomise mechanisms to resolve conflicts among stakeholders and reach some levelof agreement about the objectives thelr vTiil pLrrsue.

Strategy formulation

Stt'ategies define how or'_eanizatious should use their resources to meet their objectives.We can vierv strategies as constraints that organizations place on their employees so thatthey will focus their activities on what their olganizations do best; particularly in areas

where they have an advantage ovel their competitors. Well-conceived strategies, whichresult from analyzing the organizations'strengths and weaknesses in the malketplace,

-euide employees in successfully pursuing their olganizations' objectives.r0Strategies can be specified folmally or left lar-{el1, unspecified. Many organizations

develop formal strategies through systematic, often elaborate, plannin.e processes (rvhichwe discuss in Chapter 8). Othet' organizations do not have fonnal. u,t'itten stlategies:instead they try to respond to opporturities that present themselves.r' Major elements ofthese latter organizations' strategies emerge from a series of interactions between man-

a-qemert, employees, and the environntent; from decisions made spontaneously; and

frorn local experimentation designed to learn u,hat activities lead to the greatest success.

Nonetheless, if some decision-making consistency exists, a stl'ategy can be said to have

been formed, regaldless of whethel managers planned ol e\/en intended that particulalconsistency. lr

Judging frorn employees' actions, it is sometirnes difficult to identify an organization'sstrategy. Spontaneous decisions sometirnes conflict dilectly with the organizations'

Management and control

rr-rfl]lrl stl'ategic statelnents, ltot because of mana-qement contlol probletns but because

rhe ftrlntal stlategic statements have become obsolete and employees have decided torrke rrctions that are better than the folnal strategy sLlggests. In the early 1980s, Intel's\rite(l plan was to be a major playel in rnerroLy chips (as rvell as n-ricroprocessors),

but in 198-5 it exited frorn the dynarlic random access lnelnory (DRAM) business. Iuretfospect, Andy Grove, Intel's Chief Executive Officer' (CEO), observed that the com-pan\' \\'as "fooled by its orvn strate-qic rhetolic." Its marketin-e. pricing, and investmentclecisiorrs as eally as 1983 made it clear that some key employees had made a decisionro letl'eat florn rneuror'1,chips.rr The point is that the actual strategy an organizationeuacts rnay be different tiorn its fonnal strategic statelnellts.

Not eveu the rnost elaborate strate-qic visions and statenents ale cor.nplete to the pointu hele they detail every desired action and contemplate s1,s11, possible contingetrcy.Ho'uveveL, fol purposes of designing N{CSs, it is r,rseful to have stlate-eies that are as

specific and detailed as possible, if those strate-qies are well thou-ght out and can be keptcut'l'ent. The formai stlate-9ic statements make it easier for ilrana-qenent both to identifythe feasible lnana,qenent control alternatives and to implement them effectively. Thelranagelneltt controls can be targeted to the organization's critical success factors, suchas developing new products, keeping costs down, or enhancing rnarket share, rather thanaiming more generally at improvin_u profitability. Formal sfrate-9ic statements ale notr.nandator'5,for rlanagernent control purposes, hou,eveL. N4an1'organizatiotrs with lalgellleurergent strate_sies have effective MCSs, aithough their control alternatives are oftenmole limited.

Management control vs. strategic control

In the bloadest seuse. contLol S),Stens can be viewed as having trvo basic functions:stlategic control and managernent contlol . Strategic cotttrol involves nanagers address-ing the questiorr: Is onl strate-u1, valicl? Or', more applopriatell, in changing environments,they ask: ls or"rr stlategy still valid, and if not, how should it be changed? All firms rrustbe concerned with stlategic control issues, but the concern that a strategy may havebecome obsolete is obviously greatet'in firms operating in more d),r.rarnic environments.''

Manctgentertt corttol involves addressing the -eeneral question: Are oul employeeslikely to bel.rave applopriatell,? Thls question can be decomposed into several parts.First. do oul emplo,vees undel'stand u,hat we expect of them? Second, rvill the1, work con-sistentiy hard and try to do what is expected of ther-n; that is, will they implement theorgauization's strategy as intended? Third, ale they capable of doing a good job? Finalll,,if the answer to any of these questions is no, what can be done to solve the managenentcontrol ploblerns? All organizations who nrust lely on their employees to accomplishorganizational objectives must deal u,ith these basic management control issues.

The tools for addt'essing strategic and rnanagement control issues are quite different.Managers addlessin-s stretegi(' control issues have a focus prirnarily extelnal to the

organization; they examine the industly and their or'-uanization's place in it. They thir*about horv the organization, with its pafiicular combination of strengths, weaknesses,opportunities, and limitations, can compete with the other fir'rls in its indnstry. Managersacldressirtg ntettctgetnetlt corttrol issues, on the other hand, have plimalily au intemalfocus; they think about how they can influence enployees'behaviors in desired lvays.

This book focuses on lnana-genent control. In nost cornpanies. focusing on irnpror,-in-e MCSs wrll plovide hr-eher payoffs than will focusing on improving strategy. AFortrttte studlr 5hsw.d that sevetr out of 10 CEOs who fail do so not because of bad strat-egy but because of bad executior-r.r5 Another study of Finrtncial Times 1,000 cornpanies

Chapter 1 . Management and Control

found that 80o/o of directors thou-eht their compan), had the ri-qht strate-qies. but or-iy l4Vo

thou-qht that theil companies were implementing the stlategies ,,r,eil.ro The tenns exectt-

tiott and str(tteS!\, intplenrentatrori have the same meaning as mana-qement control the waythat telm is used in this book.

From a lranagement control perspective, strategies should be viewed as useful, butnot absolutely necessary, guides to the plopel design of an MCS. As lve show later, whenstlategies are formuiated more clearl1,, more controi altematives become feasible. and it1..-^'-'.. r',.i:- t. i-rll:rrert erch fbnr of management control effectivel;,. Managets

:-..,.; i.1 ilc) oi controls u'ithout havin-e any clear strateg-

k'5 1ll lllll1(1.

Behavioral emphasis

As mentioned eallier, nanagemeut contlol invoives managel's taking steps to help ensurethat the eniplol,es5 do what is best for the or'-canization. This is an impol'tant functionbecause it is people in the organization who make things happen, N4anagement controlsal'e necessaty to guard against the possibilities that people will do something the organ-ization does not want then to do or fail to do solnethin-q they should do. It makes littlesense to talk about cost control, for example, without let-erence to people because costsdo not control thernselves; people contlol them.

Tliis behavioral orientation is not only an alea of agreement in the recent nanagementcontrol literature; it also has long been recognized by lnanagers and controllers. Forexample, when Bill McElroy, fnance directol and boald mernber of the Toyota MotorColporation of Australia, was asked what he would stLrdy if given the oppoltunity forsome formal learning, he replied:

I rvould like to know more abont psl,chology -in terms of why people ar'e the way they are and

rvh1, they behave the way they behave. If I had studied this in my university days, I thinkI rvould have gained significant benetrts all the u,ay through nry career.'t

If all employees could always be relied on to do what is best for the organization, therewould be no need fbr an MCS. But employees ale sonetimes unabie or unwilling to actin the organization's best interest, so nranagers must take steps to guard against theocculrence, and particularly the per'sistence, of undesir'able behaviors and to encouragedesirable behaviors.

CAUSES OF MANAGEMENT CONTROL PROBLEMS

Given the behavioral focus of contlols. the next logical question to ask is: What isit about the employees on whom the organization mnst rely that creates the need toimplement MCSs? The causes of the needs for contlol can be classified into thlee maincategories: lack of dilection, motivational problems, and personal limitations.

Lack of direction

Some employees perform poorly simply because they, do not know what the organiza-tion wants fi'om them. When this lock o.f clireclioru occurs. the likelhood of the desiredbehaviors occurring is obviously smail. Thus, one fnnction of management controlinvolves informing employees as to how they can maximize tl-reir contributions to thefirlfilhnent of organizationai objectives.

Causes of management control problems

Lack of dilection is not a trivial issue in many organizations. For example. suwey evid-ence collected in 2005 by KPMG flom approxirnately 4,000 US enployees spanning alllevels of job responsibility across a wide range of industries and olganizational sizes

revealed that 557o of the sampie respondents had a lack of understandin-q of the standards

that apply to theil jobs.rN Moreover', a 2004 study of 414 World-at-Work members inmostly managelial positions at large Norlh-Arnerican companies shorved that 81% of therespondents believe that senior lnanagers in their olganizations understaud the valuedrivers of theil business strategy; 46Vo say that middle managelnent undetstands these

dlivels, but just 13oz'c believe uonfiranagement enployees understand them. This indic-ates that or'-eanizational goals are not cascacling down to all levels in the organization.And, while 79a/o of the respondeuts in this study beiieved tliat theil employees' goals are

aligned with or-eanizational goals, 44Vo also stated that employees set goals based ontheir own views lather than dilection fi'om leadership.''

Motivational problems

Even whele employees understand what they are expected to do, some choose not to per-form as the organization would have thern perfotm because of ntotit,cttional problents.Motivational problems al'e comrr-ron because inclividual and organizational objectives donot naturaily coincide; individuals are self-interested.r0

Most, if not all. empioyees sometimes act in their o\\/n personal intelest at the expenseof their organization's interest, Flederick Taylor. one of the n-rajor' figures in the scientiJicntcutogetnent ntovenletlt that took place il the early-2Oth century wlote: "Hardly a com-petent workel can be found who does not devote a considerable amount of time to study-ing just how slowly he can work and still convince his employer that he is going at a

good pace."2t Effot't aversiott and other seu-interestecl belnviors are still a problemtoday, however. For example, recent sun/ey evidence suggests that wasting, mismanag-ing. falsifyin-e, stealing, and abusing organizational lesources, among other types ofemployee misconduct, ale prevalent in most organizations.:: Even ostensibly inconse-quential forms of wasting time on the job can have high costs. Surflng the Intemet whileon the job, for example, was estimated in 2001 to have cost US employers $63 billionper year."

Extreme forms of employees' nisdirected behaviors, such as employee fraud, canhave seveLe. detrimental impacts. includin-e deteriorated employee molale, impairedbusiness relations, lost levenues from damaged reputations, investments in improvingcontlol procedures, legal fees and settlements of litigation, fines and penalties to regulat-oly agencies. and losses from plummeting stock prices.r' While some of these impactsmay seem far-fetched, they are not. One recent sun/ey of practice found that nearly threeout of four employees leported that they had observed misconduct in theil olganizationsin the prior l2-month period, with half of the lespondents stating that what they hadobserved could cause "a significant loss of public trust if discoverecl."r5 Focusing on

employee flaud, a 1997 study by the Association of Certified Fraud Examiners estimatedthat the cost of workplace fraud in the US alone was $400 billion per year, or an averageof $9 per day per employee tor the average compan1,.16 These losses were the equivalentof 6Vc of the total US gross domestic product. Similarly, a 2003 sur\/ey of executivesfrom 459 US publicly held firms and govemment agencies shoq,ed that 36Vo of these

organizations incured $1 million ol more in costs due to fraud.r? But small companiesare also affected b), fraud; perhaps pl'oportionally even more in terms of the losses

they incur because of it. Estimates by the Association of Certifled Flaud Examiners iu1995 sug-eest that businesses with fewer than 100 employees that were victimized by

Chapter l Management and Control

fiar.rd lost an avel'age of $120,000. an amount that sometimes threatens the life of these

small or-eanizations. r8

These huge fraud costs can be traced back to hurnan weaknesses (and, probably, to the

iack of effective MCSs). Brian McNally, the rnanager of the fashionable "44" r'estalrrant

in the Royaltorr Hotel in New Yolk, said, "Every single person in your restaurant istrying to steal fi'om yoLr,"re Randolph D. Brock, president of Block International SecurityCorporation, estimatecl, more conservatively, that:

Between 10 and 20Vo of a companl,'e employees will steal anything tliat isn't nailed down.Another'2O% wiII never steal; the,v would sa1, i1 is morally wlong. The vast majolitS' of peopleare situationalll, honest; the1, lvsn', steal if there are plopel contlols.tt'

These estimates al'e consistent with research findin-gs and the sur\/eys of plactice.rl Aspecial form of "stealin-e" occurs when employees manipulate their performance reports.either by falsifying the data ol by taking decisions that artificially boost perfolmance,with the intention of earning higher', but undeserved, per'fbrmance-dependent rewards.The surveys ofpractice invariably shorv that financiai repolting fraud has the hi,qhest costper incident (u,ith cost estimates varying from anywhele between $10 and S100 millionin sorne studies to as high as $250 rrillion per incident in othel studies), even though itocculs relatively infrequently.tt MCSs are obviously needed to plotect organizationsa-eainst these behavicls.

Employees, particularly marlagers, are also prone to rnake decisions that serve their'interests, but not those of their organization. They tend to overspend on things that maketheir lives more pleasant, such as on office accoutremeltts and other perks. They oftenengage in gamesnrcursftp such as "earnings management" to make their performancereports look good even when they knou, the actions they are taking have no economicvalue to the company and, in some cases, are actually harrnful. And they sometimes tendto be excessively risk averse and reluctant to make even good investments because offear that if the investments do llot pay off, they may lose their job. We discuss these andrelated plobiems in detail in Chapter 5 and in the chaptels in Section IV of this book.

But in addition to focusing on how MCSs can be employed to avoid ol mitigate these"negative" behaviors, this book's emphasis is also, even primaril),, on how MCSs can be

employed to motivate "positive" behaviors; that is, hou, the), can encoul'age employeesto work consistently hard to accomplish organizational objectives. The role of MCSs inmotivating employees to perform rr,,eli involves, among other subjects, the studlr efirtcetttives (Chapter 9) in a results c'orttrol context, wlrich we introduce in Chapter 2.

Personal limitations

The final behavioral problem that MCSs must address occurs where ernployees lvhoknow what is expected of them, and are highly motivated to pelforn weil, are simplyunable to do a good job because of certain personal limitations. Man1, s1 15rre limitationsare person-specific. They may be caused by a lack of requisite intelligence, trainin-e,experience, stamina. or knowled-ee fol the tasks at haud. An exarnple is the too-conmonsituation where employees are promoted above their level of competence. When employ-ees are "over theil heads," problems are nearly inevitable.

Another colnmon personal limitation is lack of knowledge or information. Manymanagement control problerns occur because key personnel do not have the informationnecessary to do a good job. Management at American Bakeries desclibed the companyas being in a "state of disarray" because of the absence of cliticai information aboutdelivery routes, depots, bakeries, and divisions.rt Some major firms in the romance novel

10

Characteristics of good management control

publication industr'1, u'ere desclibed as being "out of control" because the key mana-qel's

in the firn-rs did not have the ir-rtbtrnation uecessary to make good publication decisions,

aud costly uristakes wele being urade.ilSome jobs just are not desigued properly. These jobs ma)/ cause even the most physic-

a111,61 people to get tired or stlessecl ar-rd, in turn, lead to on-the-job accidents and

decision en'ol's.Son-re jobs require employees to perfolm duties or make judgments that e\fen the most

talented amon_q us are unable to perfornr. A significaltt arld -erorvin-u body of psycholo-gical resealch has deu-ronstrated that all indivicluals - even vely intelli-eent, well-trained,experienced rndividuals - have solne sevel'e limitations on their abilities to pelceive newproblems, to remember'important facts, and to pl'ocess infbmration propell1,.r5 In look-ing at the future. it has been shorvn that people tend to overestimate the likelihoodof commor-r e\1ents and events that have occurred relativell, recently (both of which aleeasier' to lemember') as compared with relatively rare events and those that have notoccumecl recentll,. Sometirnes trainin_q can be used to reduce the severity of these limittr-tions, but in most situations nultiple biases and limitations renlain.i6 These lirnitationsare a probiem because the1, 1s4r'r.. the probability that employees ivill rnake tlle cor'r'ect

decisions 01'even that they will obsen,e the problems about rvhich decisions should be

made. Researchels ale just beginning to explol'e the mauagemeut control implications oftliese lirnitations.''7

These tluee lnana-sement control probler.ns - lack of direction, motivational problems,and personal lirnitations - can obviously occur simultaneously and in any combination.An ernployee may not understand wirat is expected. n-tay not be motivated to pefibrnilvell, and may uot be capable of perfoming well even if slre both unclerstands what isbein-e asked for and is liighiy rnotivated to achieve it.

CHARACTERISTICS OF GOOD MANAGEMENT CONTROL

To have a hi_sh probability of success, organizations urust maintain good managementcontrol. Goocl cotttt'ol means that management can be reasorabl)/ confident that no major'unpleasant surprises will occur, The label out of corttrol is used to describe a situationwhere thet'e is a high plobability of poor perfornance. either overall or in a specificperformance area, despite having a leasonable stl'ategy in place.

Good managerrent control stiil allorvs for some probabiiity of failure because perfectcortrol does not exisl except perhaps in vely unusual circumstances. Perfect controllvould require corlplete assurance that all physical contlol s),Stems are foolproof and allindividuals on lvhom the organization must rely always act in the best way possible.Perfect control is obviousll, not a lealistic expectation because it is viltually irnpossible

to instail MCSs so rveil designed that they guarantee goocl behaviors. Furthennore, as

MCSs are costl1,, it is larely, if ever, cost effective to tly to irrplement enough controlse\/el] to approach perf'ect control.

Tlre cost of not having a pet'fect controi systen'r can be called a c'ontt'ol io.ls. It is the

diffelence between the performance that is theoretically possible given the strate-qy

selected and the performance that can be reasonably expectecl with the MCSs in place.

More or bettel MCSs shoLrld be irnplemented only if the amount by which they wouldleduce the contlol loss is greatel than their cost. Optinral c'orttrol can be said to have been

achieved if the contlol losses are expected to be smaller than the cost of implementingrnore controls. Because of control costs. perfect cot.ttt'ol is rarely tl-re optimai outcome;what is optimal is control that is good enough at a reasonable cost.

11

Chapter 1 . Management and Control

Assessing whether good control has been achieved must be future-oriented and

objectives-driven. It must be .firure-orienlerl because the goal is to have no unpleasantsurprises in the future; the past is not relevant except as a -euide to the future. It must be

ob.jectit'es-ch'it,eli because the objectives l'epresent what the organization seeks to attain.s*

Nonetheless, assessing whether good control has been achieved is difficult and subject-ive. It is difficult because the adequacy of management contl'ol must be measuredagainst a future that can be very difficult to predict. Good control also is not establishedover an activity or entity with rnultiple objectives unless performance on c// signific-ant dimensions has been consideled. As difficult as this assessment of managementcontrol is, however, it should be done because organizational success depends on a

good MCS.Organizations that have not achieved good control, either because they have not

implemented an MCS or because they have not implemented one well, are likely toface severe repercussions. As the examples plovided at the beginning of this chapterillustrate, they can suffer loss or impairment of assets, deficient l'evenues, excessivecosts, inaccurate records and leports that can lead to poor decisions, legal sanctions, or'

business intenuptions. At the extreme, if they do not control performance on one or morecritical perfomance dimensions, these organizations can fail.

CONTROT PROBTEM AVOIDANCE

Implementing some combination of the behavior-influencing devices commonly knownas MCSs is not always the best way to achieve good control; sometimes the problemscan be avoided. At'oiclance means eliminating the possibilitl, that the control problemswill cause the organization harm. Organizations can never avoid all their controlproblems, but they can often avoid some of them by limiting exposure to certain types ofproblems and problem sources, or by reducin,e the maximum potential loss if the prob-lems occur. Four prominent avoidance strategies are activity elimination, automation,centralization, and lisk sharing.

Activity elimination

Managers can sometimes avoid the control problems associated with a pafiicular entityor activity by tuming over the potential risks, and the associated prolits, to a third partythrough such mechanisms as subcontracts, licensing agreements, or divestment. Thisform of avoidance can be called actrllrl' elinunatron.

Managers who are not able to control certain activities, perhaps becanse they do nothave the required lesources, because they do not have a good understanding of therequired processes, or because they face legal or structural limitations, are those mostlikely to eliminate activities. General Motors (GM) turned its Clark, New Jelsey. rollerbearing opelations over to the plant's ernployees. The plant had not performed rvithinlimits acceptable to GM managers. The GM managers hoped that the emplol,ees wouldsoon understand an important message that they had been unable to get them to under-stand - that productivity improvements were necessary for the plant to sun ive.se

When managers do not wish to avoid completely an alea that they cannot control well,they ale wise at least to limit theil investments, and hence their risks, in that area. ChaseManhattan Bank was left with a potential $135 million after-tax write-off because of itsinvolvement in the -government-securities lending business with Drysdale GovemmentSecurities. In retrospect, bank executives admitted they had not understood this business

tz

Control problem avoidance

.,i.i1 its lisks ver1, well and that they had not been wise to become so heavily involved iu:,- Limiting risk is partial avoidance of problems that might arise.

The economics-oliented literature that focuses on wirether specific activities (transac-.:ons ) can be controlled nore effectively thlough ntarkets or throu-9h or-Qanizational ltier-.,'t /iics is krrowr-t as trctttsctctiotl cost econoriirc's. A detailed exaurination of the theories-,nt1 evidence in this field of stud1, is outside the scope of this book.'' We just note that.ire fact that all olganizations of any size struggle with mana-eement control issues is tes-inlent to the limitations of arms-length, market-based transactions r.vith entities extemal.tr the firm to solve all control problems satisfactorily. As such, olganizations will always.r.ri'e to rel1, on MCSs, which have been found to be effective in a br oad range of settings.The r,vorldwide glowth and success of lalge diversified organizations has depended to a

irirge extent on -qood MCSs.

Automation

)trtontntiott is a second avoidance possibility. Mana-{ers can sometlmes use computers.lobots, expert systems, and other means of automation to reduce their organization'sJ\posule to sonle control ploblems. These automated devices can be set to behave appro-pliately, and u,hen they ale operating propelly. they usually perform rnole consistentlythan do hnmans. Computers eliminate the human problems of inaccurac),. inconsistency.lnd lack of motivation. Once progralnmed, computers are consistent in their treatmentsof tlansactions, and they never have dishonest or disloyal motivations.

As techuolo-ey has advanced, organizations have substituted machines and expert\)'stems for people who have been perfonnin-e quite complex actions and makingsophisticated judgments and decisions. In hospitals, artificial intelligence systems are

rible to perform many of the tasks doctors and nurses previousll, had to perform. Theses)'stems monitol the patients' conditions and tlends and alert the medical staff of pos-sible ploblerns; they assist in making diagnoses; they order the needed drugs; and theycheck for potential dru-e interactions and allergic reactions.r2 These systems allow hospi-tiils to avoid one of the behavioral problems - the personal limitations of the rnedicalstaff. In the vast majority of situations, these systems are more likely than are the mern-bers of the medical staff to recall all the details of evely condition. medication, and pos-sible complications to initiate the propel lesponse. The system makes it more likely thatno major, unpleasant surprises u,ill occur', in this case, avoidable medical elt'ors.

Mrs. Fields' Cookies provides another good example of an MCS dominated by auto-mation.| Ti-ris company is the largest retail operator of gourmet cookies through a fran-chise system with over 3,000 points-of-retail worldwide in over 30 countries, many ofrvhich are located in shopping rnalls. The 8,000 stole-level emploi,ees are mostly youngand irrexperienced. The company has a small headquarters staff.

The Mrs. Fields' Cookies' MCS is built around a cornputer application that directsand assists the store rnanagers. The computer makes hourll' sales projectiorts and tells thernanagel's what to bake - how nrany batches of cookies. of what t1rpe, and when. It gives thenranagers volume-increasing suggestions. If customer counts are down, it might su-sgest

-eiving away cookie samples to shoppers in the mall. If customer counts are nomral butsales are down, it rtight present ideas for suggestive selling. The computer schedules thecrerv. It helps intewiew applicants by having applicants type answers to questions intothe computer, and an expert system analyzes the responses. It assists rvith personneladministration by generating the personnel folder and remindin-s the manager of paper'-

work requirements. It helps with naintenance b1, making suggestions regarding repairs,and if the suggestions do not work, b1, preparing a wolk order and selecting a vendor. The

13

Chapter 1 Management and Control

computer system is designed to make it possible tor even inexpelienced store managers

to run their stores just as Debbi Fields, the successfnl founder of the company. woulcl.In most managerial situations. horveveL, automatiol'r can provide only a paftial control

solution at best. One limitation is feasibili{r,. Hnmans have many talents - particularlythose involving complex, intuitive judgments - that no machines ol decision modelshave been able to duplicate. A second limitation is cosr. Automation often requires majorinvestments that may be justifiable only if improvements in productivity, as well as incontrol, ale forthcoming." Finaliy, automation may just leplace some contlol problemswith others. Computel automation often increases control risks. The elimination ofsource documents can obscure the audit tlail; the concentration of information in one

Iocation can increase security risks; and placin-e ,qreater reliance on computer programscan expose the company to the risks of progratnmel effors or fraud.

Centralization

Cenn'ali:cttion of decision-making is a third avoidance possibility, and it can eveu be oneof the centlal elements of an or'-eanization's MCS. Extrente forms of centralization inwhich all the key decisions are made at top management levels ale comlron in smallbusinesses, particularly when the1, ale nrn "like a family store, with an iron _erip onauthority" by a strong leader who is otlen the founder ol owneL.ts Stlong fomrs of cen-tralization also exist in some Iarge businesses rvhose top managel's have repntationsfor bein-q "detail oriented." Data General Corporation's ex-president. Edson de Castro,maintained centralized control. A former lnana-qer in the {ilrn observed that "all the realdecisions in that company go to one desk - de Castlo's."tt' Mr'. de Castro leserved theimportant, and sometimes the not-so-irnpoltant, decisions fbr himself because he appar-ently did uot tmst subordinates to take the propel actions. In so doing, he avoided somecontrol ploblems.

Some nranagers centralize decision-rnakin_s in some al'eas of theil organizations at

specific points in their histories to implove conttol. N4any companies that became con-cemed about the multimillion dollar losses u'ith complex delivatives suffeled by Procter& Garnble, Gibson Greetings, Metallgesellschafi. and Oran-ue County California, toname a 1-ew, have responded by centlalizing lisk ntanasement activities. In the bankingindustry alone, Morgan Stanle1,. Citicorp, and Su'iss Bank are alnong the companies thathave appointed corporate risk managers to pelform this important activity.rT

Centralization exists to some extent at all levels 01'mana-eement. as managers tend toreserve for themselves many of the rnost critical decisions that fall within their autholity.In fact, one sttrdy found that (1) identification ofthe key'r'isk aleas, and (2) centralizationof decision-rnaking in these areas ale charactelistics of the MCS used by "excellent"Canadian companies.*8 Common candidates fol centralization are decisions regardin-ernajor acquisitions and divestments, major capital expenditLu'es, negotiation of pivotalsales contracts, organization changes, and hiring and fir'ing of executives. However,in most organizations of even rninimal size, it is not possible to centralize all criticalactivities, and other contlol soltttions are necessarl'7.

Risk sharing

A final, paltial avoidance possibility is risk shalins. Sharin-s risks with outside entitiescan bound the losses that could be incurred by inapplopriate employee behaviors. Risksharing can involve buying ittstu'curt'e to protect against celtain types of potentially largelosses the organization rnight not be able to afford. N'Ianv companies pulchase fidelity

14

Control alternatives

bonds on employees in sensitive positions (such as bank tellers) to reduce the fitm'sexposure, These insurance contracts pass at least a portion ofthe risk oflarge losses and

errors to the insurance providers. Another way to share risks with an outside party is toenter into a joittt yenture agreement. This shares the risk with the joint venture partner.

These avoidance altematives are often an effective partial solution to, or bounding of,many of the control ploblems managers face. It is larely possible to avoid all risksbecause firms are rewarded for bearing risk, but most firms use some forms of elimina-tion, automation, centralization, and risk sharing in order to limit their exposure to the

management control problems.

CONTROT ALTERNATIVES

For the control problems that cannot be avoided, and those for which decisions have beenmade not to avoid, managers must implement one o[ more control mechanisms that are

generally called nunagenrcnt cotltt'ols. The collection of control mechanisms that are

used is generally referred to as a mqtngemetlt control D)stem (MCS).MCSs vary considerably among organizations and among entities or decision areas of any

single company. Tables 1.2 and 1,3 show some of the controls used in a manufacturing

Tnete 1.2 Examples of controls used in a manufacturing firm

1. The cash payment and cash receipt functions are segregated.

2. A check protector is used, and signature plates are kept under lock and key.

3. The accounting department matches invoices to receiving reporls or special authorizations prior topaymenr.

4. Checks are mailed by someone other than the person making out the check.

5. The accounting department malches invoices to copies of purchase orders.

6. The blank stock of checks is kept under lock and key.

7. lmprest accounting is used for payroll,

8. Bank reconciliations are to be accomplished by someone other than the one who writes checks andhandles cash.

9. Surprise counts oi cash funds are conducted periodically.

10. Orders can be placed with approved vendors only,

'1 1. All purchases must be made by the purchasing department.

Sourcet K A lvlerchant, Modern Management Conlrcl Systems: Text and Cases (Upper Saddle River, NJ: Prentice Hall, 1 998), p 1 3

Tnetr 1.3 Examples of controls used in a computer facility

'I Written standards exist for documentation of systems, operations, and administration.

2. Access to the computer system and all online data terminals is restricted at all times to authorizedpersonnel only.

3, Data are secured through tape file prolection rings, file labels, cryptographic protection, duplicationprocedures, and requirement of storage of duplicates at a remote site.

4. Hardware controls include duplicate circuitry, dual reading, echo checks, preventive maintenance,and uninterruptible power systems.

5. Major risks are insured against.

6. Backup systems and procedures are developed.

Soutce:K A. Merchant Moden Manegemenl Conlrcl Syslems: Texl and Cases\Upper Saddle River, NJ: Prenlice Hall, 1998), p. 14

1F

Chapter 1 . Management and Control

firm and a computer facility. respecfively. The MCSs of some organizations consist

primarily of trying to hire people who can be relied upon to serve the organization well.

Other or_sanizations provide modest performance-based incentives, and still others offer

incentives that are hi-uhly leveraged. Some organizations base incentives on the accom-

plishment of tar-eets defined in terms of accounting numbers, others use nonfinancialmeasul'es of perfonnance, and still others evaluate performance only subjectively. Some

organizations have elaborate sets of policies and procedures that they expect employees

to follow, lvhereas others have no such procedures or they allow the procedures that were

once in place to get out of date. Some organizations make extensive use of a lalge pro-

fessional internal audit staff, while others do not have a separate intemal audit function.These are just examples. The distinctions that can be made among the MCSs in use are

nun]erous.Managers' control choices are not random. They are based on any of a numbet of

factors. Some controls are not effective, or are not cost-effective, in certain situations.Some types of controls are better at addressing particular types of problems, and differ-ent organizations and different areas within each or-eanization often face quite differentmixes of control problems. Some types of controls have some undesirable side effectsthat can be particularly dangelous in some settings. And some controls met'ely suit par-

ticular managels' styles better than others. A major purpose of this book is to describethe factols affecting management contlol choice decisions and the effects on the organ-ization when better ol'worse choices are made.

OUTLINE OF THIS BOOK

Tlie book discusses MCSs from sevelal different angles, each the focus of one majorsection of the book. Section II distinguishes controls based on the object of corttrol,which can focus on the results produced (results cotltrol), the actions taken (crctiort con-trol), or the types of people employed and their shared norms and values (pet'sonrtel ctncl

ctlttn'ctl control).re Chapters 2-6 in Section II discuss each of these forms of control, the

outcomes they produce (which can be both positive and negative), and the factors thatlead rnana-eers to choose one object of control ovel' allother.

Section III focuses on the majot' elernents of firruncicrl results control ,t),JleillJ, an

important type of results control in rvhich results are defined in financial terms. Thissection includes discussions of financial responsibility stluctules (Chapter 7), planningand budgeting systems (Chapter 8), and incentive compensation s)/stems (Chapter 9).

Section IV discusses some major problems managers face r,vhen they use financiairesults control systems and, particularly, the performance measurements that drive them.These problems include the tendency of accounting measul'es to cause managers to be

excessivell, shofi-telm oriented (myopic), the tendency for return-on-investment meas-

nres of performance to cause bad investment and perfomrance evaluation decisions, and

the likelihood of negative behaviolal reactions from managers who are held accountablefor factors over which they have less than complete control. Throu-ehout Chapters 10, 11,

and 12, u,e also discuss several approaches organizations can rely orr to mitigate these

problems.Section V of the book discusses sonre key olganizational control roles, including those

of contlollers, auditors, and audit committees of the board of directors. It also discusses

lecent deveiopments in corpolate govemallce, as well as common control-reiated ethicalissues and how to analyze them.

16

The final sectiou of the book, Section VI, discusses some of the contextual factors that

have significant effects on either the choices of MCSs or their effectiveness in specilic

settings. Chapter' 16 discusses the effects of thlee of the uost important factors that cause

control systems to be different: environnental uncettainty, organizational strategy, and

multinationatity, Chapter 17 focuses on some control problerns unique to not-for'-profi1

organizations.

Notes

1. A. Zinimerman and A. Raghavan. "Diarnond GroupWidens Probe of Bribe Char-ees." Tlte Wall Sn eet Jottr-ral (Malch 8, 2006), p. B 1.

2. "Report of the Board of Banking Supervision Inquirl'into the Circumstances of tlre Collapse of Balirrgs,"Bank o.f Englatrd (Ju11' 18, 1995).

3 C. Karnrin and G. Fields. "Lax Contlols May ExplainTrading Loss at Allied h'ish," The \|'all Street Jourtnl(March 8. 10021. p. A8.

4. ''Keystroke Errol Causes Tum-roil in UK N4arket," Iftel4/ctll Sn'eer Journctl (May 16,2001), p. Cl4

5. "Man Adn.rits Theft fi'orn US Archives," The LosArtgeles frrras (lt4arch 14, 2002), p. ,\21.

6. lvl Molin, "Tivo Accused of INS Shreddin-u Sptee,"

Tlte Los Angeles files (Jat.ruary 31,2003), p. 857. "The Road to Perdition." Tlte Ecortonist (July 24,

2003). p. 39.

8. See. fol exarnple, R. Simons. Levers of Control HottMattaget s Use Irrnot'cttit'e Contol Sls/ents to DrireStotegic Renetrol (Boston, MA: Harvard Bustttess

Scliool Press, 1995); R. Simons, Pet.fonttctrtce Merrs-

ut emeti ancl Control Slstents Jor Intplententittg StrateSy(Upper Saddle River, NJ: Prentice Hall. 2000): D. J.

Galloway, "Control Models in Perspective." IrttetttalAuditor,51, r'Lo. 6 (December' 1994). pp. 46-52;E. G. Flarnholtz. T. K. Das and A. Tsui, "Toward an

Inte,erative Frarneu,ork of Organizational Contro1,"

Accottntittg, Ot gctni:cttiorts ctrtcl Society, 10, no. 1

(Januar'1, 1985), pp. 35-50; and K. A. Merchant.Cotttrol irr Bttsirtess Organi--atiotts (Marshfield, N{A:

Pitrnan, 1985).9. Some key leferences in the area of olganizations and

marlagement include: H. Mintzberg. Structtrre irt Fives'D e s i g n i n g EJfe c: r i v e O r g o r i : a t i o r t s, 2nu edtt (En-uleri'ood

Cliffs, NJ: Prentice Hall, 1992)l H. Mintzberg. f/reStructuting of Organizatiorrs (Engleu'ood Clitfs, NJ:

Prentice Hall, 1979): and H. Mintzberg, ldirtt:berg on

Mattagenrert (Nerv York: The Free Press, 1989).

10. Some key references in the area of stlategic nlal1a-qe-

ment include: R. M. Grant. Corttentporars' SucttegtAnaltsis, 5'r' edn (Oxford: Blackwell, 200,1); H.Mintzberg, J. B. Lampel, J. B. Quinn and S. Ghoshal,

The Sn'arcgt Process.4'r' edn (Engleu,ood Clifts, NJ:

Prerrtice Hall,2003); Ir4. E. Polter, Cotttpettttt'eSn'nteg,r', Techrtiques fot Artctlt':ing Industries ancl

Conuetitors (Ne,'v York: The Free Pless. 1980); and

Notes

tr4 E. Porter. Contpetitive Aclvantttge: Creatitrg cttttl

Sustainirtg, Superior PetJot nrutlce (Neu' York: The Flee

Pless. 1985).11. See H. Nfintzberg. "Clafting Strategy," Hartctrd Busi-

iress Ralierr,. 65, no. 4 (Jul1,-{u-ou51 1987), pp. 66-75.12. H. Mintzberg, "The Strateg)' Concept I: Five Ps for

Strate-sy," Ctlifornia Managemertt Reviev', 30, no. I(Fall 1987), pp. 11-24

13. R. Henkoff. "Horv to Plan lor 1995," Forturte(December 31, 1990), p. 76.

I4. The strate-qic control task has been discussed b1', among

others. I. Tavakoli and K. J. Perks, "The Developmentof a Strategic Control Systern tbr the Managetneut ofStrategic Change." Strategic Clrctnge,10, no. 5 (August

2001), pp. 29'r,-305; M. Goold, "Stlategic Control inthe Decentralized Fir:m." Sloart X4anagentettt Relie\',32, no. 2 (Winter 1991), pp. 69-81; J. F. Preble,"Torvards a Comprehensive System of StrategicCorrtlol." Jortrttctl of Mcrrtogentent Stu(lies, 29, no. 4(July 1992), pp. 391-409; G. Schreyogg and HSteinmann, "sttategic Control: A New Perspective,"Acarlen4' o.f Managentent Ret'iev', 12, no I (1987),

pp. 91-103; and J H. Holovitz, "Stlategic Control: ANeq,Task for Top Management," Long Range P[nrtttirtg.12 (Jnne 1919), pp.2-1.

15. R. Charan and G. Colvin, "Why CEOs Fail," Fortune,139, no. 12 (June 21, 191)9).

16. Businessballs,corn ivebsite.17. W. Birkett, "The Changing Role of the CFO: An

Irltervie\\, u'ith Bill McEh'o),," A \/iew oJ Torttorrov'The Seniot Fittcttttial OJficet in tlte Year 2005 (New

York: Intelnational Federation of Accountants, 1995).

18. KPMG 200512006 Integri4' Srrr.r'e,r' (KPMG LLP,200s).

i9. Wolld-at-Wolk, Sibson, and Synygy, The Stare oJ

Pe'rformance ll4artagenrett (Survey Report, August

2004); and J. Kochanski aud A. Sorensen, "Managing

Perforrnance Management," Workspatr (September'

2005), pp. 2l-6.20. Man5, managentent accountittg and management

control textbooks refel to lttck of goal ('ongrue]t('e as a

-9eneral problem category rvhich subsumes both lack

of direction and lack of rnotivatiort.21. F. Taylor', The Prirtciples o.f Scietttifit'Mctttogienrent

(New Yolk: Harper, I929).22. KPI,IG 200512006 lttegt'i.r'Sirn'e,i' (KPMG LLP, 2005).

11

Chapter 1 Management and Control

23. S. PlLritt. "Are E,rnployees Wastrng Time Orrline?''PCIUot'ld Colr (August 2. 2001).

24. Prit'ewaterhouseCooners 2005 Globctl EconontttCrinte Survet' (PricewatethouseCoopels LLP. 2005).

25. KPMG 2005t2006 Integrin'Srrr ler' (KPMG LLP. 2005).26. l. T. We]rls. Occttttcrtiottal Fratrd trrtcl Abuse (Associ-

ation of Certified Fraud Exauriners, 1997).21 . KPMG 2003 Fraud Srrn'e,r' (KPMG LLP. 2003).28. J. R. Ernshwiller, "Small Business is the Bi_egest

Victim of Theft by Ernployees, Sulvey Shorvs." I/rr,Wnll Srreet Jr,urrral (October 2, 1995), p B2.

29. K. B. Leu,is. "Thou Better Not Steal,'' Forbes(Novernber 1 , 1994), p. 110.

30. D. Gillmor'. "Crinre Is Headed Up - And So IsBtrsiness," Bostr,tu Globe, Febrrrary 15, 1983, p. 47.

31. M. Liprrran. Steulittg How Anterit'o's Etrytlot'ee.s treSteulirtg Tlteir Contparties BIirtd (Nerv York: Harper'sMagazine Press. 1973); PricewoterlnuseCoopers 2005Global Ecottotttic Crinte Srl'r'e,t (PlicervateLhouse-

Coopels LLP, 2005); KPMG 200512006 IttregrirvSrrt e_r (KPN4G LLP. 2005): KPI,IG 2003 Fraud Sttt'et'(KPMG LLP. 2003).

32. PricevaterltotseCoopet's 2005 Global Et'ott,ttttit'Crinte Sun'et' (Pricewaterhousecoopers LLP. 2005):ard KPMG 2003 Fratrd Srrr le-r (KPMG LLP. 2003).

33 "Anrelican Bakeries: A Nerv Chef Cleans Up the

Kitchen," Brr.rilres.r W'r,al (June 27, 1983), p. 52.

34. "Why Book Publishers Are No Longer in Love withRornance Novels." Busittess lTeeA (Decenrber 5, 1983).p,157.

35. Seminal ref'elences are R. E. Nisbett and L. Ross. HnttcrttInlerettce' Strategies cutcl Slrortt'orttirtgs of SocittlJutlgntent (Englervood Clifft, NJ: Prentice-Hall, 1980);R. E. Nisbett and L. Ross. Tlte Per.sott and tlte Situcttiorr(Nerv York: McGlaiv-Hill, 1991); and D. Kahne rnan andA. Tvetsky. "Plospect Theory: An Analysis of DecisionsUnder Risk." Ecottorrtett'ictt. 11 (1919), pp. 313-27.See also R. H. Ashton and A. H. Ashton. .lutlgnrertt atrlD eci s i o n-M ctki t t g Resea rch i rt Accttu ttt i n g a ncl Audi|i rt g

(Cambridge Selies on Judgrnent and Decision Makin-e.Carnblidge Univelsity Press, 1995); R H. Ashton.Htrrttatr Ittfortttatiort P rocessirtg itt Ac'coutttittg, Studiesin Accounting Research no l7 (Sarasota. FL: AmelicanAccorrnting Association, 1982): R. Libby. Act'outtrirtgartcl Huntcttr Irt.fortrntiott Ptor;essittg: T!teot v crrttl Ap-pliccttions (Englewood CtitTs, NJ: Prentice-Hall, 198 I );

A. Tvelsky and D, Kahneman, "Advances in ProspeclTheoly: Cumulative Repr:eseutation of Uucertainty,"Jourrrql o/Rr.rA urrcl Utttettctinn', 5 (1992), pp.297-323; and D. Kahneman, "A Pelspective on Judgmentand Choice: ivlapping Bounded Rationality.'' ,{rrtet tr'enPst'cltoktgist,58, no. 9 (2003), pp. 691-120.

36. M. H. Bazerman, Jrrtlgnteur irt Nlartctgeriul DecisiottMaking.6'n edn (Nes, Yolk: John Wiley & Sons, 2005).

37. S. L. Sclrneider, Entergittg Pet qtectit es ort .ltrclgutettt(td Decisiou Re secu't'lt (Camblidge Series on Judgrnentalld Decision Making, Cambudge Universitl, Pless.

18

2003): and R. H. Ashton and A. H. Ashton. .luclgntctttand Decisiort-NIakirtg Re.scarclt itt Accoutttirtg crtrcl

Auditittg (Canrbridge Series on Judgment and DecisionNlaking, Canrbridge Urrivelsitl, Press, 1995),

Confusion has long reigned as to how to use the termsobjectitas and goals. In this book. objectives ret'er tobload things tl-re or-eanization \\,auts to achieve, such as''be a leader in the infblrnatioll sel'\,ices indnstry "Goals refer to specific thin-es the or-uauization \\'auts toachieve in a specifiecl tirne period. such as "earln a2O7cretullr on net assets in the corning 1'ear'." Objectivesale relativell, stable, Goals rnay clrange every planninepeliod.A. Sloan, "Go Forth and Conlpete." Forhes (Novenrbet'

23. 1981), pp.41.-2.J. Salarnon, "Horv Nerv Yolk Bank Got Itself Entangledin Drysdale's Dealings." Tlte lI;ull Srret't Jourttol (ItneI l. 1982), p. A1.Olivel Willianrson is _geneLally lecognized as the nlostpronrinent theoretical contl'ibutol' in the area of tlansac-tion cost econorlics. FoL an overvierv of this litelature.see O. E. Williamsor-r, "Tlansaction Cost Econornics."irr R. Schmalensee and R. Willig (eds). Hartdbook oJ

Irrclrtstrictl Et'otrcntics (Neu, York: North Holland.1989), ChapteL 3l and O. E. Williarnsor.r, "Tlansaction

Cost Econourics: Hon, ir WoLks: Whele it is Headed."De Ecottotttist. 146. no. 3 (1998). pp. 23--58.S. Oliver. "Take Trvo Aspirin: rhe Compulel Will Callin the Moruin_s.'' Fot'bes (N4arch 14. 1994). pp. 1 l0-11.This exanrple is taken flonr T. Richman. ''N{r's. Fields'Secret Irrgredient." /iir'. (October 19871. pp. 61-72.See also S. H. Haeckei aud R. L. Nolan. "Mana-eing

by Wite." Hcttt'arcl Brrsilcss Rc,r'1crr,,71. no.5(Septeniber-October 1993). pp. 122-31.See. tbr example. C. Olnstein. "Hospital HeedsDoctors; Strspends Use of Sotirvare." Tlte Los ArtgelesIirras (January 22.2003), p. Bl.S. NlLrfson. "Amerada Hess Chief Keeps ControlsTiglrt, Emphasizes Nlarketin_s." The lYall Street Jout-tnl(Janualy I I, 1983). p. l."Data GeneraI's N{ana-eement Tlouble," Brtsiness 11/eek

(Febnrary 9. l98l). p.58."Managin-e Risk." Brrsirrcss lI/eek (October 31. 1994),p 92.T. Carvsey, G. Deszca and H. D. Tetl|, ilIutta,qenctilCortn'ol St's/ents irt E.rc'cllcttt Cqnatlion Conrpctniet.i\,lana-Rernert Accounting Issues Paper #5 (Hamilton.ON: The Society of iVlanagetneut Accountants ofCanada. i994).This fi'arnewolk rvas discussed b1, W. Ouchi, "A Con-ceptual Flarnes,ork fbl the Design of OrganizationalContlol Mechanisnts." ful ct tta g e nr e rt St' i e nc e, 25. no. 9

(S€ptenrber 1979), pp. 833-48. It rvas elaborated on byK. A. Merchant. Coutt'ol irt Business Or.z,ani:tttiotts(Cambrid-ue, IVIA: Ballin-ser'. 1985). Section II of thisbook presents a rehned and expanded discussion of thisfiamervork.

38

39

40

4l

+L

+J

44

46

48

19

RESULTS CONTROLS

ff asked to think about powerful wa1's to influence behavior in organizations, most

Ipeople would ptobably think first about pay for pet'fotmauce. And, no doubt, pay forpelformance is an effecti\re motivator. For example, at Thor Industries, the world'slargest recreational vehicle manufacturer, Wade Thompson, the CEO, attributes much ofthe company'S succesS to its compensatiotr s),stem. Among other things, the company

shales 15olc of each division's pretax profits with the division managers, because,

\{r. Thompson explained, "I want every one of olu' company heads to feel like it is their

business, in their control. If they don't perform, they don't get paid very much. If they

do, thele is no cap to what they can make."r Pay for perfotmance is a prominent exam-

ple of a type of control that can be called results control because it involves re'v\i arding

employees for generating good t'esltlts.

But the lewards that can be usefull;' linked to results -eo fal beyond monetarl' ssm-pensation. As Vicky Wright, managing directol' at the Hay Group, a personnel and

compensation consultancy firm, put it:

The companies on the Most Aclmired list [a list of cornpanies produced annually by FortLtne)

have chief executives who undelstand what perfolmance measurement is all about. It's about

leaming how to motivate people - how to link tliose performance measures to rewards.2

Other rewards that can be linked usefully to measured perfonnance include job security,

prornotions, autonom)/, and recognition.'Results controls create meritoct'acies.In meritocracies, the rewards are given to the

most talented and hardest working employees, rather than those with the longest tenure

or the right social connections. The combinations of rewards linked to results inform or

remind employees as to what result areas are imporlant and motivate them to produce the

results the organrzation rewalds. Results controls influence actions becattse they cause

employees to be concerned about lhe consequences of the actions they take. The organ-

ization does not dictate to employees what actions they should take; instead employees

ate enlpoy,ered to take those actions they believe will best produce the desired results.

Results controls also encoutage emplol'ees to discover and develop their talents and to

get placed into jobs in which they will be able to perform well. For all these reasons, a

well-designed results control system should help produce the desired results.

There are many examples of dlamatic perfomance improvements following the

introduction of anew results control system. In 1995. US West Communications Group

implemented a pay-for-pelformance system that promised sales reps compensatlon

increases of 20Vo or lnore if they met peformance goals. In the following three years,

revenues per emplol,ee had increased by 4lVo,] This is consistent with a meta-analytic

25

Chapter 2. Besults Controls

review of 45 field and laboratory research studies on the use of incentives to motivatepelformance, which found that the overall avera-qe effect of all incentive prograrns in allwork settings and on all work tasks was a22Vc gain in perfot'mance'

Like all othel forms of controls, however, results controls cannot be used in everysituation. They are effective only where the desiled result aleas can be contlolled (to a

considerable extent) by the employee(s) whose actions are being influenced and wherethe controllable result areas can be measuled effectively.6

PREVALENCE OF RESULTS CONTROTS

Results controls ale courmonly used for controlling the behaviors of employees at manyorganizational levels. They ale a necessary element in the employee entpov'ennett ap-proach to rnanagement, which became a major mauagement trend startin-e in the 1990s.7

Results controls are particularly dominant as a means of controlling tlie behaviors ofprofessional entployees; those rvith decisiort ctuthority, like managers. Reengineeringguru Michael Hammer even defines a professional as "someone who is responsible for'achieving a result rather than [for] performin-e a task."o

Results controls are consistent with, and even necessary for, the implementation ofdecentralized forms of organization with largely, autonomous entities ol responsibilitycenters (which we discuss in more detail in Chapter 7). Fol example. business pioneer'Alfied P. Sloan observed that he sought a way to exelcise effective control over thevvhole corporation yet maintain a philosophy of clecentralization.' At General Motors(and numerous other companies that followed), the solution under Sloan's leadershipand beyond was tesnlts controls built on a retuln-on-investment (ROI) performancemeasure. By using this type of control systeln, corporate management could review andjud-ee the effectiveness of the various olganizational entities while leaving the actualexecution of operations to people lesponsible tbr the perfbrmance of those entities; theentlty managers.

DuPont, Menill Lynch, Boeing, Coca-Cola, Alcoa, and Snnrise Medical are amongthe many cornpanies that have gone through the process of instituting rnore decentralizedforms of organization with a concurrent increased emphasis on results control. In 1993,DuPont's CEO teplaced a complex management hierarchy by sptitting the compan)/ into21 strategic business units (SBUs), each of which opel'ates as a free standing unit. TheSBU managers wel'e given greater responsibilit;, und asked to be more entrepreneurialand more customer-focused. They u,ere aiso asked to bear more risk, because a lar-ee

portion of SBU managers' compensation was based on SBU performance (sales andprofrtability). The managers noticed the change. For exarnple. one SBU managel said,"When I joined DuPont [2 I 1'ears ago], if you kept )/our l-lose clean a1d worked hard, yottcould work as long as you wanted. [But today] job security depends on resuits."'0 Thechange was perceived as being successful: A Brrslness Weck article noted that, "Theima-ee of DuPont has morphed fi'om giant sloth to gazelle."rl

In 2003 and 2004. Merill Lynch placed a new emphasis on financial results.Thel, 6su.,1 billions of dollars of central adrninistrative costs into their business units'profit-and-loss statements, used perlormance talgets and metrics to inclease operatiltgdiscipline. and rcinforced the chan-qes with more transparent. more luclative perfomrance-dependent bonuses. About this process, Ahrnass Fakahany, Merrill's CFO, said, "OnceaccoLrntability of the fully integrated P&L was in place, people started to make tradeoffs."Managers started paying more attention to their technolo-ey bud-eet, and the businessunits were pushed towards more outsollrcing. Now, Mr, Fakahany continued, "A famill,

26

Prevalence of results controls

culture has been replaced by a perfonnance culture with operating discipline and a sense

of nrgenc1,."rrPhil Conclit, Boeing's CEO, sirnilarly derided the family culture in his company,

which was all "about seniority, not performance. In a family cultule, you never throw out

a bad performer." Among the first changes he made after assuming the CEO offrce was

to change Boeing's paternalistic culture into one focused on accountability for perform-

ance.'t Improvements were almost immediately apparent.

Coca-Cola's plesident explained his company's inteut in decentralizing as follows:

We're giving our division mana-qers alonnd the world a lot of authority, atrd we't'e holding

them responsible. We're going to reu,ard thern for rneeting objectlves that thel' have agleed to.

If the1, meer them, rhe1,'r's going to have rnoney jingiing in theil pocket; if they dou't, some-

body else u,ill be given that opportunitl'."

Paul H. O'Neill. Alcoa's chaiman. said:

We cannot succeed if we persist in our use of the traditional command-and-contlol system ofrnanagement where many thousands of people believe their only responsibility is to do what

they are told to do.r5

And Richard H. Chandler', CEO of Sunrise Medical, a medical products company head-

quartered in Carlsbad, California, defended his company's decentralized olganizationancl lucrative performance-based bonuses as an effort to "replicate the entrepreneurial

model" within a multifaceted corporation. He said,

People want to be relvalded based on theil own effolts. [Without divisional accountability] youend up with a system like the US Post Office. There's no incentive [for wolkels to excel].'n

Indeed, many conlpanies have found that managers will act in the entrepleneurialnanner necessal'y to thlive in fast-moving markets only if they are subjected to the

sarne market forces and pressures that drive independent entrepl'eneul's and if they are

promised commensurate rewards for the risk they must bear,

Thus, decentralization or the delegation of authority ol decision rights to managers,

on the one hand, and the design of incentit'e s-l)'s/eir?s to ensule that these managers do not

misuse their discretion and are appropriately rewarded commensurate with the risk they

bear, on the other hand, are two critical organizarional design choices in a results-

control context, Both decisions are part of what some olganizational theorists refer to as

orgctnizcttional erc'hitectu'e. This litelatule maintains that both orgauizational choices ale

linked and that concentt'ating on one element to the exclusion of the othel leads to poorlydesigned organizations; in other words, decisions about decentralization and iucentivesystems should be made jorrrrly'. Fulther', this literature argues that a firm's choice of the

organizational architecture is context-specific', depending ou factors such as the market

structure, the or-eanization's strategy, the ploduction or sel'vice process, and the extent ofinformation asynmetry, Factors identifled as supporting decentralization include molelocal information, constrained upper nanagement time, greater need for training oflower-level managers, feasible incentive costs, productiou or service processes that

require little coordination across olganizafional units, and low levels of centralized

information needed for local units to function.rT (We discuss the influence of contextualfactors on MCS design in more detail in ChapteL 16.)

At middle organizational levels, results controls ale often irnplernented under the

framer.vork of a martctgentenf -b1'-objecrn,es (MBO) system. In its most basic fotm, MBO is:

A process u,heleby the superiol and subor:dinate mallagers of an organization jointly identifycommon goals, define each emplo1'ee's rnajor areas of responsibilities in terms of the results

21

Chapter 2 Results Controls

expected of thern. and use these rneasuLes as guides for operating the unit and asse.ssin_g thecontribution of each of its rnembels,rv

Results controls can also be emphasized down to the lowest ievels in the organization,as many companies have done with good effects. In a survey of mid-sized nanufactur-ers (with annual sales between $10 million and $500 million) sponsored by professionalservice firm Grant Thornton, 80% of the l'espondents repofied they were working on pro-grams to give their workels mole power and responsibility on the shop floor'.re

For example, it is common for delivery personnel to be paid on a commission basis.At the Frito-Lay division of Pepsi Coia, deliverymen receive only a small weekly salarybut are paid a 107c commission on all the chips they sell. Studies have found that thissystem encourages them to serve the company's interest better: the drivers do not merelydeliver the chips; they also "stop to talk with sttpermarket managers, an-eling for an extrafoot of shelf space."ro

Porsche, the German automobile manufacturer'. and Cleveland-based Lincoln ElectricCompany are among the compar-ries that use results controls down to the lowest organ-izational levels in their manufacturing areas. Porsche, which is known for high-qualityproducts, enters the name of the worker who installs each majol engine component inthe engine's log so if a fault (a result) appears later, it can be ttaced back to the personresponsible.rr Lincoln Electric provides wa-qes based solely on piecervork for mostfactory jobs and lucrative perfbrmance-based bonuses that can rnore than double anemployee's pay.2r This incentive system has cleated such high procil-rctivity that someof the industry giants (Genelal Electlic, Westinghouse) found it difficult to compete inLincoln's line of business (arc welding) and exited from the market.r3 A Business \Ueekarticle observed that "in its reclusive, iconoclastic way, Lincoln Electric remains one ofthe best-managed companies in the United States and is probably as good as anythingacross the Pacific."1 And even though Lincoln's legenclarl, Incentit,e Perf'orntanc'eS_),s/ent has essentialll, r.tnu n"O the same since it was installed in 1934, the companyis still acclaimed for its systems and perfom-rance today, such as in a recent book titledThe Modern Firm.'5

Frattchisirtg is another approach for implementin-e results controls. With franchising,business ownership, with all of its risks and lewards, is passed to a franchisee. So are the

decision rights, although these are constl'ained by the franchise contract to ensure thatfranchisees do not deviate from the franchise concept. For example, McDonald'shambulger franchisees mnst offer Big Mucs, lvhich are an important element of the

menu concept used around the u,orld. But the control advantage of franchising is that

franchisers can spread the use of their concept and earn fees and royalties with minimumcontrol risk because franchisees' r'ewards stem directly fiom the profits they earn. The

rewards motivate the franchisees to be hardwolking, efficient, r'esponsive to ctlstomers,

and entrepreneurial.

RESULTS CONTROTS AND THE CONTROL PROBLEMS

Results controls are preventive-type controls that can address each of the majot cate-e-

ories of control problems. Well-definecl results inform employees as to u'hat is expected

of them and encourages them do what they can to produce the desired results. In this way,

the results controls alleviate a potential lack of direction'

Results controls also are often particularly effective in addressing motivational prob-

Iems. Even without upper-level manager supervision or intervention, tlle lesults contlols

28

Elements of results controls

induce employees to behave so as to lrraximize their chances of ploducing the results the

organization desires. This desirable motivational outcome occurs because the olganiza-rions' desired results are also. not coincidentally, those that will maximize the employ-ees' own pelsonal rewards.

And results controls also carr address pelsonal limitation problems. Because resultscontrols usually promise high rervalds for good performels, use of results controls can

help lirms attract and retain employees who are confident about their abilities.26 Andlesults controls can encourage all employees to address their limitations and to developtheir talents to position thenselves to earn the results-dependent rewards.

The perfonr-rance ureasut'es that al'e a part of the results controls also provide somenonmotivational, detection-type control benefits of a c1'bernetic (feedback) nature,as was rlentioned in Chapter 1. The results measures help managers answer questionsabout how various strategies, organizational entities, and employees are perfotming. Ifperformance fails to meet expectations, managers can consider chan-ees of the strategies,the managers, or the operational processes.=t Investigating and intervening only whenpelfonnance is lagging is the essence of a ntanagenrcnt-b\,-esception approach to man-agement, which is widell, u5e6.

ETEMENTS OF RESUTTS CONTROLS

The implementation of results contfols requires four steps: (l) defining the dimension(s)on which resr,rlts are desired (or not desired), such as profitability, customer satisfaction,or product defects; (2) measuring perfolmance on these dimensions; (3) setting perform-ance targets for employees to strive for; and (4) providing lewards to encourage thebehaviors that will lead to the desiled results. Each of these steps has pitfalls.

Defining performance dimensions

Defining the right performance dimensions is critical because the goals that are set andthe measurements that are made shape employees' views of what is important. Or, in theterms of a '*,r,idely known business adage: "What )i ou measure is what you get." What iswonisome is that employees work to improve the areas that are measuLed regardless ofv'ltefher or not tlte nteasrtrentent clintensions are cleJtned correctl\,.If the measurementdimensions are not defined conectll,; that is, if the1, al'e not congruent with the olgan-ization's objectives or agreed-upon strategies, the results contlols will actually encour-age employees to do the wrong things. We discuss this probleni further in Chapters 5, 10,

and 1 1.

Measuring perfolmance

Measurerneut, which involves the assignment of numbers to objects, is a critical elementof a results control system. The object of importance is the perfbrmance of an employee(or a group of employees) in a specific time period.

Many different results measures can be linked to rewards. Many objective financialmeasures, such as net inconre. eamings per shale, alld leturn on assets. are in commonuse. So are some nonfinancial measures, such as market share, growth (in units), cus-tomer satisfaction (as measured, for example, by repeat sales or a mailed survey), and

the tirnely accomplishnent of certain tasks. Some measurements involve subjectivejudgments. Evaluators may be asked to judge whether a manager is "being a team

29

Chapter 2 BesLrlts Controls

player" or "developing employees effectivell," and to record their judgrnents on a crlrde,

oldinal measurer.nent scale (e.-s. frorn I (unsatisfactory) to 3 (excellent)).The measures used tl,pically vary across or-eanizational levels. At hi,sher organiza-

tional levels, most of the key results linked to reu,ards are defined in financial terms. The

measures ma)/ be either market-based pertonnance indicators (such as stock price orreturns) ol accounting plofits or 1'eturns (such as return on equity). Lower-level rnan-

agers. ou the other hand, are typically evaluated in tenns of operational data that are lrtore

controllable at the local level. The key result areas for a lnana-qer in charge of a tnanu-

facturir-s site, fol example. rnight be a combination of efficiency (such as labor hours per

units produced), inventory contlol (such as days sales on hand), quality (such as avelage

number of defects per unit produced). delively time (such as the percentage of orders

shipped on time). and batch setup time.The lack of symrnetr'), in the nses of financial and opelational performance measures

between top mana_qetnent anci iower-level management creates a critical pivotal point inthe management hieralchy, which is sornetitnes called a hinge or linkin-e pin.tt At some

critical middle organizational level, often a pl'ofit center levei. managers must tt'anslate

financial goais iuto opelational _eoals. These managel's'goals are primarily defined infinancial terms. so their communications rvith theil superiols are primarily in financialterms. But because their subordinates' -qoals are primarily opelational, their downwardcommunications ale primalily in operational terms.

If mana-qers identify mole than one resnlt lleasule fol a given entployee, theymust attach lelative inportance wei_uhtings to each lneasure so that the judgments

about perfolnlnance in each |esuit area can be a-eg|e-eated into an overall evaluation. The

weightings can be additive. For example .60c/c of the overall evaluation is based on return

on assets and 40c/c is based on saies growtll. The wei-ehtin-qs can also be multiplicative'For example, Browning-Fenis Industries multiplies a score on achievement of profitand levenue goals by a score assessecl based on enviLonrnental lesponsibility.r') If the

environmental lesponsibility score is less than 107c,the multiplier is zelo, and so is the

resulting bonns.Sornetirnes. organizations make the weightin-es of perfomance lneasures explicit

to the emplo)/ees, as in the examples just presented. Often, however, the weightings are

partially ol totally implicit. sr-rch as rvhen the perfonnance evaluations are doue subject-ively. Leaving the weighting implicit blurs the commnnication to employees about rvhat

results are important. Employees ale left to infer what results will most affect their over'-

all evaluations. We discuss the causes and effects of performance evaiuation explicitness(i.e. objectivefformulaic vs. subjective/discretionary) further in Chaptels 9 and 12.

Setting pedormance targets

Pertbnnance targets, or standalds, are another important results control systern element.In a lesults contlol system, tal'_sets should be specified fot'every perforrnance dimensionthat is measured.

Perfornance tar_sets aff'ect behavior in two basic ways. First, they stimulate action(inrplove motivation) by providing conscious goals for employees to strive for. Mostpeople pref'er to be given a specific target to shoot for', rather than merely being ,eivenvague statements like "do the best yon can" or "rvork at a reasonable pace."r0 Second,

pelforrnance talgets allow employees to interplet theit'olrt perfbrmauce. People do not

lespond to feedback unless they are able to intet'pret it, and a key palt of interpretation

involves comparing actual performance lvith the predetetmined perfbrmatrce targets.''

The targets distinguish strong from poor perfot'mance. Failure to achieve the tar,qets

30

Elements of results controls

provides managels with a signal that they should probably chan-ee their actions. We dis-cuss perfonnance targets and target setting processes in mole detail in Chapter 8.

Providing rewards

Rewards or incentit,es are the final important elernent of a results control system. Therewards included in incentive contracts can be in the form of anything employees value,such as salary increases, bonuses, promotions, job security, job assignments, trainingopportunities. freedom. recognition, and power. Punishments are the opposite of rewards.Thel, 219 things employees dislike, such as demotions. supelvisor disapploval, failure toget lewatds eamed by peers or. at the extlelne, the threat of dismissal,

Organizations can derive motivational value from linking any of these valued rewardsto results that employees can influence. For example, organizations can use any of a

number of e-rtt'insic rev'qrcls. They can grant additional monetary rewards, such as in theform of cash or stock. They can use non-monetary rewards, such as by glarrting high per-forming employees public recognition and additional decision authority. Alternatively,in entities where perfomrance is mediocre or pool', they can threaten to reduce the deci-sion authority and powel managers derive from rnanagirrg their entities by lefusing tofund ideas for expenditures.

Results measures can provide a positive motivational impact even if no rewards are

explicitly linked to results tneasures. People often derive theil orvn internally-generatedittrinsic rev,arcls through a sense of accomplishment for achievin-s the desired results.For example, when William J. Bratton became the New York City Police Cornmissionerin January 1994, he gave department pelsonnel one clear. simple goal: cut crime,sr(Previously the thinking had been that crime was due to societal factors beyond thedepartment's control, so the police were largely measured by how quickly they re-sponded to emelgency calls.) He also implemented a results corrtrol system. He decen-tralized the department by giving the 76 precinct commanders the authority to make mostof the key decisions in their police stations. including the light to set pe1'sonnel sched-ules, and he started collecting and leporting crime data daily. Even though CommissionerBratton legally couid not award good performers with pay raises or merit bonuses, thesystem was tremendously successful. In 1994, majol felonies in New York fell by 12%,and in the first three quartel's of 1995, thel, fs11 another 18% below 1994 levels. This suc-cess clearly was not attributable to pay for performance in the strictest sense; it wasinstead due, at least in part, to providing officers with clear goals and empowering themto go about lighting crime. Seeing the r'esults of theil initiatives gave police officers a

sense of accomplishment, and thus, an intrinsic motivation to perform well.The motivational strength of any of the extrinsic or intrinsic rewards can be under'-

stood in tenns of several motivation theories that have been developed, such as

expectoncv theorl'. Expectancy theory postulates that individuals' motivational force, oreffolt, is a function of (l) their expectattcies or their belief that certain outcomes willtesult frorn their behavior (e.g. a bonus for increased effort); and (2) their valence.s or thestrengtlr of their pleference for those outcomes, The ralence of a bonus, however, is

not always restricted to its monetary value, it also nray have yalettc'e in securing othervalued items. such as status and prestige.'r

Organizations should promise their ernployees the rewards that provide the most pow-erful motivational effects in the most cost effective way possible, But the motivationaleffects of the various rewatd forms can vary widely depending on individuals'personaltastes and circumstances. Some peopie are greatly interested in immediate cash awards,whereas others al'e more interested in increasin-e their retirement benefits, increasing their

31

Chaoter 2 . Besults Controls

autonolny, or improving their promotion possibilities. Rewald tastes also vary acl'oss

countries for a number of reasons, including diffelences in cultr.rres and income tax laws.

However, if organizations can tailor their reward packages to theil employees' indivi-dual preferences, they can provide meaningful tewalds in a cost efficient manner'. Buttailoring rewards to individuais or small groups within a large olganization is not easy

to accomplish. A tailoled system will likely be complex and costly to administer. Butwhen poorly done, it can easily lead to employee perceptions of unfaimess and poten-tially have the opposite effects of those intended: demotivation and poor employeemorale. In Chapter 9 we discuss issues related to choices of incentives in more detail.

CONDITIONS DETERMINING THE EFFECTIVENESSOF RESULTS CONTROTS

Although they are an important form of control in many organizations, results controlscannot always be used effectively. They work best only when a// of the following con-ditions are present:

1. organizations can determine what results are desired in the areas bein-e contlolled;

2. the employees rvhose behaviors are being controlled have significant influence on theresults for which they are being held accountable: and

3. organizations can measure the lesults effectively.

Knowledge of desired results

For results contlols to work, organizations lnust know what results are desired in the

aleas they wish to control, and they must communicate those desiles effectively to the

employees working in those areas. Resrr/ts clesirabilin' means that more of the qLrality

represented by the resuhs measure is preferred to less, evelything else being equal.At a general level, most people agree that the primary objective of for'-profit organiza-

tions is to maximize shareholder (or owner) value.! It does not follow, however, thatbecause this overall objective is known, the desired lesults are then also known at all inter-mediate and lower levels in the organization. The disaggregation of overall organizationalobjectives into specific expectations for all employees lower in the hierarchy, is otien dif-ficult. Different needs and tradeoffs may be present in different parts of the organization.

For example, purchasing lnanagers create value by plocuring good-quality, low-costmaterials when needed. These three result areas (quality, cost, and schedule) can oftenbe traded off against each other. and the overall organizational objective to maximizeshareholder value does not provide much help in making these tradeoffs. The irnpot'tanceof each of these result areas may vary over time and among parts of the organizationdepending on differing needs and strategies. For example, a company (or entity) short ofcash ma1, want to minimize the amount of inventory on hand, which may make schedul-ing the dominant consideration. A company (or entity) with a cost leadership strategymay want to emphasize the cost considelations. And a company (or entity) with productswith a quality image or differentiation strategy may emphasize meeting or exceedingthe speciflcations of the materials being purchased. Thus, to ensure proper purchasingmanager behaviols, the importance orderin-es ol weightings of these three resnlt areas

must be made clear.If the wrong lesult areas are chosen, or if the right areas ale chosen but given the

\\i rong importance weightings, the combination of results measures is not corrgrrrent with

32

Conditions deternrining effectiveness

:_lmization's true objectives.35 Using an incongl'uent set of results measures will,.r r.notivate ernployees to take the wlong actlons.

llility to influence desired results (controllabilityl

.r'iond condition that is necessar)i for lesults controls to be effective is that the

:.iovees Lvhose behaviors are being controlled must be able to affect the lesults il a

,.:r'ial way in a given time period. This controllabiliS' pt'inciple is one of the centrai- :rs of responsibility accounting (u,hich we discuss in more detail in Chapters 7 and

- Hele are some representative expressious of this perennial principle:

It is alrnost a self-evident proposition that, in appraising the perforrnance of divisional man-:rqerneut, no account should be taken of lnattel's outside the division's coutrol.t6

\ manager is not norrnally held accountable fol unfavorable outconres or cledited withilr olable ones if they, are cleally dLre to causes not under tris control.r?

The main lationale behind the controllability principle is that results measures ale...efr.rl only to the extent that they provide information about the desirability of the

-:.-tions that were taken. If a results area is totaily uncontrollable, the results measures:eveal nothin-q about what actions were taken. Paltial controllability makes it difficult tornt-er frorn the lesults measures whether ol not good actions \ /ere taken.

In most organizational situations, of course, nurlerous uncontrollable or paltiallyLrncontrollable factols affect the lreasures used to evaluate perfomance. These uncon-rlollable influences hinder efforts to use results measures for control purposes, If theeft-ects of the uncontlollable factors cannot be sorted out and eliminated, often allrhat car be measured is a broad band within which perforrnance probabll, Iies. rathertl.ran a precise lrreasure of performance. In this case, it becornes difficuit to deteminervhether the results achieved are due to the actions taken or to uncontrollable factols.Good actions will not necessarily produce good results. Bad actions may siurilarlybe obscured.

In situations rvhere urany, large uncontrollable influences affect the availabie resultslneasures, results control is not effective. Managers cannot be lelieved of their respons-ibility to respond to relevant environmental factors, but jf these factors are difficult toseparate from the results measures, results controls do not provide good infbrmationeither for evaluating perfolmance or for motivating good behaviors. We discuss themethods or-eanizatior-rs use to cope with uncontrollable factors in lesults control systelnsin rnore detail in Chapter' 12.

Ability to measure controllable results effectively

Ability to measule the controllable results effectivell, is the final constraint liuriting thefeasibiiity of lesults contlols, Often the controllable results the or-ganization truly desires,and the employee involved can affect, cannot be measured effectively. Measurementitself is lareiy the problem; in virtually all situations sontethittg can be nreasuled as, bydefinition, measurement requires onl1, that numbers be assigned to events or objects. Butsonretimes the key results areas cannot be measured e.ffectfteh'.

The key cliterion that should be used to judge the effectiveness of results measuresis the ability to evoke the desired behaviors. If a measure evokes the right behaviors ina -eiven situation, then it is a good control measure. If it does not, it is a bad one,

even if the measure acculately reflects the quantity it purpolts to lepresent. To evoke theright behaviors, in addition to being con-sruent, results measures should be (1) plecise,

33

Chapter 2 . Results Controls

(2) objective, (3) timely, and (4) understandable. If any of these rneasurement qualitiescannot be achieved. results control will not be effective in evoking the desired behaviors.

Precision

Measurernenl precision refers to the amount of t'andomness in the measure. For plecisionto be high, the dispersion amon-q the values placed on a -9iven result by multiple inde-pendent measurements must be small. For example, if l0 independent measurements

show that the quantity being measured is exactly 120.3, then the measure is precise.

If one can conclude only that the quantity is between 100 and 130, the measure is less

precise. Some aspects of perfonnance (such as social responsibility, pelsonnel develop-

rnent) ale difficult, ol'even irlpossible, to measure precisely.Precision is an inportant quality because without it the measure loses rnuch of its

information value. Imprecise measules inct'ease the risk of misevaluating performance.En'rployees wili react ne_eatively to the inequities that will inevitably arise when equallygood pertormances al'e rated differently.

0biectivity

Ob.lectit,itt,, which means fieedom frorn bias, is another desirable measurelrent quality.Measulement objectivity is low (i.e. the possibility of biases is high) where either the

choice of measurement rules or the actual measuring is done by the persons whosepelfomances are being evaluated. Low objectivity is likely, for exampie, where per-formance is self-reported or where evaluatees are allowed considelable discletion in the

choice of measnrement methods, such as is true to some extent with the measurement ofaccounting income.

There are two main altematives to inclease measurement objectivity. The first altem-ative is to have the measuring done by people independent of the plocess. such as bypelsonnel in the controller's depaltment. The second alteurative is to have the measule-ments verified by independent palties, such as auditols.

Timeliness

Tintelirtess refers to the lag between the employee's pelformance and the measulementof results (and the provision of reu'ards). Tirneliness is an important measurement qual-ity for trvo reasons. The first is motivational. Employees need consistent, short-tern per-formance pressure to perfom at their best. The pressure helps ensule that the employeesdo not becorne complacent, sloppy, ol wasteful. Measures, and thus rewards, that are

delayed for significant peliods of time lose most of their motivational impact. Shorl-telmpl'essure can also stimulate creativity by increasing the likelihood that employees will be

stimulated to searcl.r for new and better ways of improving results.A second advantage is that timeliness increases the value of interventions that might

be necessary. If significant problems exist but the perfomance measllres ale not timely,it might not be possible to intelvene to flx the ploblems before they cause seveLe harm.

Understandability

Two aspects of unclerstanclabiliry,are important, Filst, the employees whose behaviorsare being controlled must undelstand what they are being held accountable for. Thisrequires communication. Training, which is a fonn of communication, may also be

necessary if, for example, employees are to be held accountabie for achieving goals

expressed in new and different terms, such as rvhen an organization shifts its measure-

ment focus from accounting income to, say, economic value added.

3+

S:cond, employees must uudelstand what they must do to influence the measul'e, at: ,.t in bload terms. Fol exarnple, purchasing managel's who ale held accountable for

'.erin-u the costs of pnrchased materials will not be successful until they develop strat-

-'--.:: fol accomplishing this goal, such as imploving neqotiatious r.r,ith vendols, incleas-: competition anong vendors, or u,orking with en-eineerin_e personnel to ledesign

-:':.titin pat'ts. Simiiat'ly, employees who ale held accountable for custorner satisfaction. i.t Lrnderstand rvhat theif cr-lstolneLs value and what the1, g2rl do to affect it.In tnost situations, understandability is not a limiting factor. When employees

-.;1e rstand what a measure l'epresents. they rvill fi-9ure out what they can do to influenceIn fact, this is one of tl-re advantages of results controls: -eood control can be achieved

,ihout knorvin-q exactly horv employees will produce the results.\ltrn1, rlsurules cannot be classified as either cleally effective ol clealll, ineffective.

lriierent tradeoffs arnong the evaluation criteria create some advantages and disadvant-.:_:.\. Fol' example, measul'es can often be made lnore collgluent, contlollable. precise.- ,ul objective if tin-reliness is corrpromised. Thus, in assessing the efi'ectiveness of resultsI lilsures, many difficult jud-ements are often uecessary. These judgtnents are discussed. nrore detail throughout several chapters of this book.

CONCLUSION

- :ris chapter described an important torn of control, results control, which is used at:t,trtv levels in most olganizations, Results cor.rtlois al'e an indilect form of control:-citllse they do not focus explicitly on the employees'actions. However, this indilect-:s\ provides some important advantages. Results controls can often be effective when. is t'tot clear what behaviols are most desirable. In addition, results coutrols can yield:.',od contLol u,hile allowing the emplol,ees whose behaviors are being contlolled hi-gh.,-rtollorl!, Manl,people, pafiicularly those highel in the or_qanizational hierarchy, value: .sh autonoml' and respond Lvell to it. Hi_eh antonomy often breeds innovation.

Results controls are not effective in every situation, however. Failule to satisfy all-ree effectiveness conditions - knowledge of the desiled results, ability,to affect the

--.::iled leslrlts, and ability [o measure controllable results effectivell, - r,vill render.results,'DIrtl'ols impotent. It will also probably, precipitate any of a number of dysfunctional side.'ltects, man1, sf which are discussed in later chapters.

Results controls usually are the major element of the MCS used in all but the snallestrrganizations. However, r'esults controls often are supplemented by action and person-:-.:l/cultural controls, which we discuss in the next chaptel.

Notes

J. Fahey. "Lord of the Rigs." FolDes (March 29,2004),p. 68."Measulin_s People Porver'." Forttne (October 2, 2000).p. 186.

See. for example, S. L. Rynes, B. Gerhar-t and K. A.N4inette. "The Lnpoltance of Paf in Enrployee Mot-ivalion: Disclepancies betu,een What People Sa1, ur.,,l

What Thel, Do." Hunrutt Resourge tr4attctgeltert. 13.no. 4 (Winter 2004): pp. 381-94.C. Palmer.i, "A Gazelle. nor a Godzilla." Fot'bes(September 21. 1998), p. 64,

l

N otes

5. S. J. Condll', R. E. Clark and H. D. Stolovitch, "TheEffects of Incentives on Workplace Performance:A Meta-Analytic Revieu' of Research Studies,"Perforntorrce Intprot'entattt Quarterl\'. 16. no. 3 (2003),pp. 46-63. For exarnples of en-rpirical resealch studieson the eft'ects of incentives on pelforntance. see R. D.Banker. S. Y. Lee. G. Potter and D. Srinivasan. "AnEmpirical Analy,sis of Continuin-9 ImproverlentsFollou,in-e tlie Implementation of a Perfolmance-BasedCornperrsation Plan." Joru nct I oJ' At'r'o un t i n,e, nncl Eco-ttotrtits. 30. no. 3 (December' 2000). pp. 315 50: and

35

Chapter 2 . Results Controls

R. D. Banker'. S. Y, Lee and G, Potter, "A Fielci Studyol the Impact of a Perlblmance-Based Incentive Plan,''Jourtul of Accouttting ctttd Ecottotttics, 21, no. 3 (Aplil1996), pp. t95-226.

6. As an exarnple of ser,eral lesults control issues that can

arise rvlien these conditions al'e not rret. see S. Kerr."The Best-Laid Incentive Plans," Hnrlnrd BustttessRetietr, 81, no. I (Januar1, 2OO3), pp. 21-40

7. Fol references on the empowerment movement see.

tbr exanrple, K. H. Blanchald. J. P. Carlos and W. A.Randolplr. Tlrc 3 Kets to Errtltorrerntelrt (San Francisco,CA: Benett-Koehlel Publishels. 1999): B. Manville and

J. Ober', "Beyond Ernpowelment: Building a Com-pany of Citizens," Haryortl Brr.iirre.ss Reyielr,8l. no. I

(Jarruary 2003), pp. 48-56; and D. E. Bou,en and E. E.Larvler'. "The Empowerment of Service Workels: What.Why. Horv. and When," Slotttr Monagentent Rerie\,,33, no. 3 (Spring 1992), pp. 3 l-9. Fol a l'ecent enrpir-ical resealch study irl this alea. see S. E. Seibelt. S. R.Silvel and W. A. Randolph, "Taking Enrpowennent tothe Next Level: A Multiple-Level Model of Emporver-ment, Pedbmance, ancl Satisfaction," A caclerny of Marragentett Jourrnl,47, no 3 (June 2004), pp. 332-49.Fol a theoretical snrd1, on the f'easibility artd limitatiousof ernpo'nvemrerrt and decentralization. see G. Baker.R Gibbons and K. J. N4urphy, "Blinging the MalketInside the Firm," Alrelic'att Etottonric Ret'iev,9l. no. 2

(May 2001), pp.2l2-18.8. M. Hanrrner. Beyottcl Rcertg,itteerittg Hov tlrc Prot'e ss-

Certterecl Orgutti:cttiou is Cltougirtg Our lVot'k arul Otrtln'cs (Nerv Yolk: Harper Business, 1996).

9. A. P. Sloan, Ml Years v'itlt Getreral Motot's (Nerv York:Doubleday, 1964).

10. J. Weber', "For DuPont, Christmas in Aplil," BusinessWeek (April24, 1995). p. 130.

11. Ibid.. p. 129.

12. "Radical Surgerl, Saved Melrill." F irruncial lirrres (July2,2004), p. 8,

13. K. Labich, "Boeing Finally Hatches a Plan," Fortarc(Malch 1, 1999), p. 102

14. Don Keough quoted in J. Huey. "New Top ExecutivesShake up Old Order at Soft-Drink Giant," The l4lallStreet Jourtml(November 6, 1981), p. 17.

15. D. Milbank, "Changes at Alcoa Point Up Challengesand Benefits of Decentralized Authority," The \UullStreet Jourrrctl (November'7, 1991), p. B7.

16. R. H. Chandler', qr-roted in T. Petluno. "Sunlise ScamThlorvs Light on Incentive Pa1, Proglams," Tlte LosAngele,s lrires (January 15, 1996), p. D3.

17. J. Brickley, C. Srnith and J. Zimmerman, MtrnagerialEconontics ancl Ot'gotti:ntiortal Arcltite.'rrrl'e (Boston.

MA: McGrarv-Hill Ilvin, 2001); D. A. Nadler, M. S.

Gerstein and R. B. Sharv, Or'.gali--otion(tl Arcltitecture:Desigrts for Cltaugiug Ot'gcrtti:rttiotts (Nerv York:Jossey-Bass: 1992); and P. Milgrom and J. Roberts,Ec otton t i c s, O t' gct t t izct t i on a n tl M a r t ct g e rnerrl (Englervood

Cliffs, NJ: Prentice Hall, 1992).

36

21.

G. Odiorne, Monagentartt-bt-Objectites: A St'stent ttfMatragenrertt Laaclcrship (Behnont, CA: Pitman Leam-ing. 1965). pp. 55-6. See also H Levinson, "Manage-merlt b)' Whose Objectives." Hotyarcl Busittes.s Retiey,.81, no I (JanuaLy 2003), pp. 101-17."Employee Autonomy Results in Enhanced Profitabil-ity." Martufactming & Distribution Issttes,7 (Suntmer1996). pp.3-4.J. Guyon. "The Public Doesn't Get a Better Potato Chiprvittrout a Bit of Pain." The IUall Street Jotu'rrnl (March25, 1983), p 1.

"Autontaking on a Human Scale." Foltrlia (April 5.

1982), pp. 89-93.See N. Fast and N. Belg, "The Lincoln ElectrrcConrparry," Case no. 9-376-028 (Bosron. MA: HBSCase Services. 1975); and M Mrowca, "Ohio FirmRelies on Incentive-Pay System to Motivate Workersand Maintain Products," Tlte lI/oll Street.Iournal(August 12, 1983). p.23."Lincoln Electric: Whele People Are Never Let Go,"i"irrre (June I8. 2001), p. 40.

24. "This Is the Answer." Brrsrness l4!eek, Jll1, 5, 1982,pp. 50-2.

25. J. Roberts. Tlte Moclertt Finn. Organi:ational Desigrt

fot Pe(brntartce arrcl Grottlt (New York. OxforclUniversit), Press, 2004).

26. A. Arya and B. Mittendorf, "Offering Stock Optionsto Gauge Nlana_sel'ial Talent," Journol oJ Accourttittgcutcl Ecouornit s, 40, no. 1-3 (Decernber 2005), pp. 189-210.

27. D, Campbell, S. Datar, S. L. Kulp and V. G. Narayanan,"The Strategic Infonnation Content of Non-FinancialPerfolrnance IVIeasules." Wot king Paper (HarvardBusiness School. 2006).

28. K. J. Euske. M. J.. Lebas and C. J. McNair', "Pelfolm-arrce Measurement in an International Setting," &,[arr-

agenrctlt Accourttittg Researclr, 4 (1993) pp. 275 -99.29. Institute of Managelrent Accountants. Implententing

Corporote Ertirorttrrcntal Strategies, Statement of Man-agement Accounting no. 4W (Montvale. NJ: Instituteof Managemenr Accountanrs, July 31. 1995).

30. E. A. Locke and G. P. Lathan, A Theorr of Goal Settingancl Task Pet'forntartce (Englervood Cliffs. NJ: PrenticeHall. 1990): G. P. Latham, "The Motivational Benefitsof Goal-Setting," Acatlenty of Managenrcnrt Erecutire,18, no.4 (Novernber 2004). pp. 126-9 andE. A. Lockeand G. P. Latham, "Building a Placticalll, UsefulTheory of Goal Setting and Task Motivation: A 35-yearOdyssey," Antericctrt Psl,t'ltologisr, 57 (2002), pp. 705-t7.

31. K. A. Merchant. Returdittg Results Motit'ating ProfitCenter Matwgels (Boston. MA: Harvard BusuressSchool Press, 1989).

32. E. Lesly, "A Sat'er New York Cit),," 6,,rt,,"rs Week(Decembel I l. 1995). p. 81.

33. V. H. Vloonr, l'l/ork and Motilnliorr (New York: Wiley.1964).

l8

19

20

22

23

ACTION, PERSONNEL, AND CULTURALCO NTROLS

p esults controls are not the only form of controls. Organizations can suppletnent orI\replace t'esults controls with other forms of controls that serve the same purposes;that is, to make it more likely that employees will act in the organization's best interest.

One such important category of controls, actiotr cotttrols, involves ensuring thatemployees petform (do not perform) certain actions known to be beneficial (harmful)to the organizalion. Although action controls are commonly used in organizations, theyare not effective in every situation. They are feasible onl1, y,1l.n managers knou, whatactions are (un)desirable and have the ability to ensure that the (un)desirable actions(do not) occur.

Another control category, personnel controls, are designed to make it mole iikely thatemployees will perform the desired tasks satisfactorlly on their ov,rt because, for example,the employees are experienced, honest, and hard working. A flnal category of contr.ols iscalled cLtltLu'al controls. Cultural contlols exist to shape organizational behavioral nolmsand to encourage employees to monitor and influence each other''s behaviors,

Action, personnel, and cultural controls are a part of virtually every MCS. In sorneorganizations, they at'e so important they can be said to be the dominant fon11 of control.

ACTION CONTROLS

Action controls are the most direct form of management control because theyinvolve taking steps to ensule that employees act in the organization's best interestby making their actions themselves the focus of control. Action controls take any offour basic forms: behavioral constraints, preaction leviews, action accountability. andredundancy.

Behavioral constra ints

Behavioral constraints are a negative form of action control, They make it impossible, orat least nore difficult, for employees to do things that should not be done. The constraintscan be applied physically or adminisrrarively.

Most companies use multiple forms of p/r,r,sical constrcurzts, including locks on clesks.computer passu'ords, and limits on access to areas lvhere valuable inventories and sensit-ive information are kept. Some behavioral constraint devices are technically sophistic-ated and often expensive, such as magnetic identiflcation card readels ancl fingerpriltor eyeball pattern readers. In situations u,here a high degree of control is desired, such

76

Action controls

as in facilities where radioactive materials al'e processed, secret service agencies whele

classified information is gathered, or casino count l'ooms whele cash is handled, the

benefits of such sophisticated contt'ols may outweigh their costs.

Aclntirtistratit,e constroints can also be used to place limits on an ernplol,ee's abilities

to perform all or a portion of specific acts. One common form of administtative control

involves the restriction of clecision-making authority. Managers at a low organizational

level may be aliowed to approve expenditures of up to $ 1,000; those at a higher level up

to $5,000, and so on. Above those limits, the pulchasing department is instructed not to

place the order. The senior managers who restrict the decision-making authority in this

way al'e trying to minimize the risk that untrained or uninformed employees will make

major mistakes.Another common form of administrative control is generaily referred to as separation

o.f cluties. This involves dividing up the tasks necessaly for the accomplishment ofcertain sensitive duties, thus making it impossible, or at least difflcult, fot one pelSon to

complete certain tasks alone. Separation of duties comes in rnany forms. One commonexample involves n-raking sure the employee who makes the payment entries in the

accounts receivable ledger is not the employee who receives the checks. If an employee

who is diverting company checks to a personal account only has the payment-entry

duties; that is, opening the mail and listing, endorsing, and totaling incoming checks,

custonrers rvill eventually complain about being dunned for amounts they had already

paid. But a person with both check-receiving and payment-entry duties could divert the

checks and cover the action by making fictitious entries of returns of goods or, perhaps,

price adjustments.Separation of duties is described by auditors as one of the basic requirenrents of

what they call good internol control . The effectiveness of separation of duties is limited,however, in that it does not plevent negative actions produced by collusiott between two

or more employees, such as those with the check-receiving and payment-entry duties.

Although collusion requires devious employees to reveal their bad iutentions to other

employees whom they seek to engage in the scheme, 63Vo of the respondents in the

KPMG 2003 Froucl Srn,e-v indicated that fraud in their organizations took place in this

way, either by collusion between employees and third parties (489a) or by collusionamong employees or management themselves (157o).1

Sometimes physical and administrative constraints can be combined into what has

been labeled as poka-.tokes that ale designed to make a process ol system/oolproof.l Apoka-yoke is a step built into a process to prevent deviation from the correct order ofsteps; that is, where a certain action must be completed before the next step can be

performed. A simple mechanical poka-yoke example is the inclusion of a switch in the

door of a rlicrowave oven so that the oven cannot be operated with the door open.

Similar mistake-preventing poka-yokes can also be built into sorne production and

administrative processes. For example, some airlines have recently switched to laptops

in the cockpit to repiace manual preflight calculations that were elror-prone. The soft-

ware is idiot-proof as it does not slip up on the math and flashes a waming if a seriouslywrong number is entered, such as a 10-ton mistake in the weight of the plane or fuelload.3 Similarly, it mi,eht be possible to have a signature-verifying computer generate thepaperwork necessary for making a cash distribution only after all the approvals for thatdistribution have been secured.

It is often difficult to make behavioral constraints foolproof, especially when the organ-

ization is dealing with disloyal, deceitful employees. For example, despite teasonable

safeguards, a fonner secretary at Bear Stearns, a global investuent firm, used disappear--

ine ink to write checks that her boss requested. After the manager signed the checks, she

77

Chapter 3 Action, Personnel, and Cultural Controls

wouid elase the name of the payee and rewrite the checks for cash. In hel eight rnonthswith the firm. she made more than $800.000 vanish from her boss's bank accounts.t

Preaction reviews

Preqctiort reyiews involve the scrutiny of the action plans of the employees bein,q

controlled. Reviewers can approve or disapprove the proposed actions, ask formodifications, or ask for a more carefirlly considered plan before granting final approval.A common form of preaction leview takes place during planning and budgetingplocesses characterized by multiple levels of reviews of planned actions and budgets

at consecutively higher organizational levels. We discuss the plannin,e and budgetingprocess in more detail in Chapter 8.

Action accountability

Actiott accoruttctbilin'involves holding emplol,ees accountable for the actions they take.The implementation of action accountability controls requiles; (1) defining what actionsale acceptable or nnacceptable, (2) communicating those definitions to employees, (3)observing 01' othelwise tlacking what happens, and (4) rewardin-e good actions or pun-ishing actions that deviate from the acceptable.

The actions for which employees are to be heid accountable can be communicatedeither administlatively or socially. Aclnirtistrcttile modes of communication include theuse of work rules, policies and procedures, contract plovisions. and company codes ofconduct, It is common in chains of fast-food franchises, such as McDonald's, to prescribeand communicate in writing, and clalify and reinforce through trailin-q, how virtuallyevelything should be done, including how to handle cash, how to hire personnel, whereto buy supplies, and what temperature to keep the oil while cooking fries.' Similarly,nurses use preoperatit'e checklisrs to help ensure that they prepare patients thoroughlyfbr surgery. These checklists remincl them to check on the patient's allergies, ciru-q-takin-s

history, and time of last meal. Department store personnel also colnrnonly have sets ofprocedures they are expected to tbllow. At a lar-ue retailer', store managers are rebuked ifempty merchandise cartons are not broken do'*,n before they are sent to the trash loombecause ernployees could use the cartons to steal merchandise.6

The desired actions do not have to be communicated in written form, however. Theycan be communicated face-to-face in rneetings or in private. Fol example, AndrewGrove, Intel's fonner chief executive officer' (CEO), reco-enized that to keep "his gen-elals and troops malching in the same direction requires constant cajoling and quarrelingup and down the ranks."l

Sometimes the actions desired afe not comrnunicated explicitly at all. In many opera-tional audits, post audits of capital investment decisions, and peer leviews of auditors,lawyers, doctoLs, and managers, irdividuals are heid accountable for their actions thatinvolve pt'oJessiortctl juclgntent. The desilability of the actions of professionals generallycannot be clearly delineated in advance. Nonetheless, these individuals are held account-able for their actions under the premise that they should act professionally.

Although action accountability controls are most ef'fective if the desired actions are

well communicated, cornmunication is not sufficient by itself to make these controlseffective. The affected individuals must understand rvhat is lequired and feel reasonablysure that their actions will be noticed and rewarded or punished in some significaut way.

Actions can be tracked in several ways, Employee actions can be obsen,ed directiyand neariy continuously as is done by direct supervisols on ptoduction lines. They can

78

Action controls and the control problems

be tracked periodically, such as retail stoles do when they use nD'ster-y shoppers tocritique the service provided by store clerks.s They can also be tracked by examiningevidence of actions taken, such as activity reports or expense documentation. Auditors,particularly internal auditols. spend much of their time examining evidence about com-pliance with preestablished action standards.

Action accountability is usually implemented with negative reinforcements. Thatis, the actions defined are more often linked with punishments than with rewards.Steelmaker Nucol links se'r,eral contract eleurents to actions as palt of the productionwolkforce's incentive compensation agreement. Anyone late for a shift loses aday's bonus, and anyone rvho misses a shift loses the bonus for the week.e At HomeDepot, managers are required to use an in-house personnel-screening system whenhiring new emplo;,sss. But recently five managers failed to use the system, and theyu'ere fired.Io

Redundancy

Reclundartcy, which involves assigning more employees (or machines) to a task thanis strictly necessary, or at least having backup employees (or machines) available, alsocan be considered an action control because it increases the probability that a task willbe satisfactorily accomplished. Redundancy is common in computer facilities, seculityfunctions, and other critical operations. However', it is larely used in other areas becauseit is expensive. Fulther, assigning more than one employee to the same task usuallylesults in conflict, fi'ustration, and/ol boredom.

ACTION CONTROTS AND THE CONTROL PROBTEMS

Action controls work because, like the other types of controls, they address one or moreof the thlee basic control problems. Table 3.1 shows the types of problems addressed byeach of the action controls.

Behavioral constraints are primarily effective in eliminating motivational problems.Employees who might be tempted to engage in undesirable behaviors can be preventedfrom doing so.

Preaction reviews can address all three of the control problems. Because they ofteninvolve communications to the employees about what is desired, they can help alleviatea lack of direction. They can also provide motivation, as the threat of an impendingreview of an employee's actions usually prompts extra care in the preparation of an

Tnarr 3.1 Control problems addressed by each of the action control types

Control problem

Type of action control Lack of direction Motivational oroblems Personal limitations

Behavioral constraintsPreaction reviewsAction accountabilityElar{r rndennrr

X

X

X

X

X

X

X

Sourcq K. A lvlerchant, Modern Managemenl Contral Syslems: Texl and Cases (Upper Saddle Fliver, NJ: Prenlice Hall. 1 998), p 30

7A

Chapter 3 Action, Personnel, and Cultural Controls

expenditure ploposal, a budget, or an action plan. Preaction reviews also mitigate thepotentially costly effects of the personal limitations. since a good reviewel can add

expertise if it is needed. The revieu,s can prevent mistakes or other harmful actions fromhappening.

Action accountability controls can also address all of the control problems. The pre-scriptions ofdesired actions can help provide direction and alleviate the types ofpersonallimitations due to inadequate skills or experience. And the rewards or punishments helpprovide motivation.

Redundancy is relativeiy Iimited in its application. It is primarily effective in helping toaccomplish a particular task if there is some doubt as to whether the employee assigned

to the task is either motivated to perform the task satisfactorily or capable of doing so.

PREVENTION VERSUS DETECTION

Action controls can also be usefully classified according to whether they serve ro pre\tetltor to detect undesirable behaviors. This distinction is important because controls that pre-vent the undesired enors and irreguladties flom occurrin,q are, when they are effective,the most powerful form of control because rtone of the costs of the undesirable behaviorswill be incuued. Detection-type action controls diffel frorn prevention-type controls inthat they are applied after the occurrence of the behavior. Still, they can be effective ifthe detection is made in a timely manner and if the detection results in a cessation of thebehavior and a comection of the effects of the harmful actions. Also, the promise ofprompt detection of harmful actions is itself preventative; it discourages individuals frompurposefully engaging in such behaviols.

Most action controls are aimed at preventing undesirable behaviors. The exception isaction-accountability controls. Although action-accountability controls are designed tomotivate employees to behave appropriately, it cannot be verified whether the appropri-ate actions were taken until evidence of the actions is gathered. Hou,ever', if the evidencegathering is concurrent with the activity, as it is with direct supervision, then actionaccountabilitl, control can approach the desired state of prevention of undesired actions.Table 3.2 shows examples of common forms of action controls classified according torvhether their purpose is to prevent or detect problems.

Tner 3.2 Examples of action controls classified by purpose

Control purpose

Type of action control Prevention Detection

Behavioral constraints

Preaction reviews

Action accountability

Redundancy

Locks on valuable assetsSeparalion oJ duties

Expenditure approvalsBudget reviews

Prespecified policies linkedto expectations of rewardsand punishments

Assigning multiple peopleto an important task

N/A

N/A

Compliance-oriented internal auditsCash reconciliationsPeer reviews

N/A

Source:K A l,4erchant, Modern Managemenl Conlrcl Syslems: Text and Cases (Opper Saddle River, NJ: Prentice Hall, 1998), p, 31

80

Conditions d eterminino eff e ctiveness

CONDITIONS DETERMINING THE EFFECTIVENESS OFACTION CONTROLS

Action controls cannot be used effectively in every situation. They are effective onlywhen both of the following conditions exist, at least to some extent:

1. organizations can determine what actions are (un)desirable; and

2. organizations are able to ensure that the (un)desirable actions (do not) occur.

Knowledge of desired actions

Lack of knowledge as to what actions are desirable is the constraint that most limits theuse of action controls. This knowledge is often difficuit to obtain. Although it may be easyto define relatively completely the actions required of employees on a production line,the deflnitions of preferred actions in highly complex and uncertain task environrnents,such as those of salespeople, research engineers, or top-level managers, cannot be ascomplete or precise. Most organizations do not have a good idea as to how employees inthese roles should best spend their time.

Knowiedge of the desired actions can be discovered or leamed in either of trvo basicways. One is by analyzing the actions/results pattems in a specific situation or similar situ-ations over time to leam what actions produce the best results. For example, (mortgage)loan approval decisions aLe now highly structured. Over tirne, lenders obser"ve whichbonowels are likely to fail their loan payments. In so doing, they can develop a loanapproval protocol, delegate the decision to lower-level employees, and control employeebehaviors by monitoring their adherence to the desired decision protocol. Another wayorganizafions can leam which actions are desirable is to be informed by others, espe-cially for strategic decisions. Indeed, this is a major role played by consultants who havedetailed knowledge of best practices.

It is important that the actions for which employees are to be held accountable be, infact, the actions that will lead to the highest probability of accomplishment of one ormore of the or-eanization's goals, or at least the proper implementation of the strategy thatis being followed. As with results controls, many organizations have actually foundthemselves holding employees accountable for taking the v,rong actions. We discuss thisproblem in detail in Chapter 5.

Ability to ensure that desired actions are taken

Knowing what actions are desilable is not enough to ensure good control; organizationsmust have some ability to ensure or obsen,e that the desired actions are taken. This abil-ity varies widely among the different action controls.

The effectiveness of the beltat,iot'al constraints and preaction ret,ievts varies directlywith the reliability of the physical devices or administrative procedures the organizationhas in place to ensure that the (un)desired actions are (not) taken. In many cases, thesedevices and procedures are not effective. For example, a rogue currency trader at AllfirstFinancial, who had lost about $700 million in foreign exchange trading, was said tohave "targeted every control point in the system and systematically found a way atoundthem," When called aside by managers for going over his trading limits, the trader com-plained that the computerized risk-monitorin-e system he used to check his risk exposureduling the day was too cumbersome. He got away with it.rL ro cover up his losses, thetrader allegedly started selling bogus option contracts. This practice was not detected in

81

Chapter 3 Action, Personnel, and Cultural Controls

a timely manner either, in part, because the responsibility fol the monitoring and report-ing of the trader's foreign-exchan-ee risks was,eiven to a junior', relatively inexperienced,staff mernber,rr

At Lelrman Brothers. a star stockbroker rnana,ged to keep a pelsonal computer on

his office desk, despite rules prohibiting this practice, and allegedly despite the fact thatsome senior executives were awal'e of it. The blokel used his personal computer, ratherthan a Lehrnan office computer, to create fake account statements rvith inflated stockprices. He then divelted the leal account statements to post-office boxes that he con-trolled, rather than to his clients who instead leceived the phony statements he generated.

Clients say the brokel forged their autholizations to send their statements to the post-office boxes. What's more, the broker supervised the compliartce staff whose job it was

to help police the office brokers. Over a l5-year period. this bloker stole $125 n-rillionflom his unsuspecting customers.'l

Exarnples such as these are consistent rvith the findin-ss of the KPMG 2003 FrquclSrrl'r'e-r'. In this survey,39Vc of the respondents stated that the frauds in their or'-ganizationsoccuned because of poor itrternal cottil'ols, many of which fall into our categor)/ ofbehat,ioral cottstraints, and 3lVo because of ntattagentettt ot'erride of internal controls.r+

Action tracking often provides a significant challenge that must be faced in makingac'tiott ctc'coutttctbilin' controls effective. Even where employees' actions cannot beobserved directly, usually some actions can be tracked. But this tlacking is not alwayseffective. The cliteria that should be used to jud-ee '"vhethel the action tlacking is eff-ect-ive at'e plecision, objectivity, tirneliness. and understandabiliti, (as we also discLrssed inChapter 2 in a results control context). If any of these measurernent qualities cannot be

achieved, action accountabiliti' control rvill not be effective in evoking the desiredbehaviors.

Precision refers to the amount of enor in the indicatols used to tell what actions havetaken place. If action trackin-e involves direct supervision, can the snpelvisors accuratelydistinguish good fi'om bad actions? 11' action tracking involves exanrination of trans-action records. do those records reliably tell whethel the ploper actions u,ere taken?For example. an initiative aimed at tlacking rvhether salespeople spend enough time onmarket development activities, as opposed to direct sales activities, is doomed to failuntil precise definitions can be developed as to rvhich actions tall into each of thesetwo areas.

Auother precision failure of an action control is the US Foreign Corrupt Practices Act.This act was intendecl to rnake bribes to foreign officials ille-eal, but it allowed.facilitat-itrg ltatntettts to Iower'-level officials. The distinction betrveen bribes and facilitatingpelntetlts was not made clear, however. The va-sueness of this law has caused rnuchconcern among corporate officials who cannot be snre that their real-time interpreta-tions of the act would match those rlade by independent observers (such as a jury) at alatel date.r-' Precision problems also limit the effectiveness of many organizations' codesof conduct.

Olt.jectit,itt'. or freedorn from bias, is a concern because repol'ts of actions plepaled bythose whose actions are being controlled camot necessarily be relied upon. Project- and

sales-orier.rted pelsonnel are tiequently asked to prepare self-r'eports of horv they spendtheir tine. Lr most cases. these l'epol'ts ale precise, as the allocatious ma1, be in units oftime as small as by the miuute. But the reports are not obiective. If the pelsonnelinvolved want to obscnre the true time patterns, perhaps to co\i er a bad perfomnance orto allow some personal time, it is relativell, eas1, fbr them to report that most of theil timewas spent on productive activities. Most companies use direct supervisors and internalauditors to provide objectivity checks on such reports. Without objectivity, management

82

Personnel controls

cannot be sure whethel the action l'eports reflect the actual actions taken, and the reportslose their value fol control pulposes.

Timeliness in tracking actions is important. If the trackin-e is not timely, interventionsare not possible before harm is done. Further, much of the motivational effect of the feed-back and rewards is lost when the tlacking is significantly delal'ed,

Finally, it is important that the actions for q,hich individuals are to be held accountablebe uttderstatdable. Employees can easily understald prescriptions to "show up to workon time" or "don't steal." But undet'standability does become a problem where the actionis defined in aggregate tetms and the individuals involved do not understancl everythingirnplied b), the aggregate prescriptions. For exarnple, auditors who are held accoultablefor "testing an accounts receivable balance" may not understand that their tests will be;udged based on the satisfactory accomplishment of a series of generally accepted steps,including inspections of documentation, confirmations, computations, reconciliations ofgeneral-ledger balances, and clelical checks. If the employees do not understand thedetailed procedures, the overall behavioral effect rvill be unsatisfactory even though thea-qgregate action is defined con-ectly and the tracking of whether or not the steps havebeen performed adequately can be done precisely. objectively, and on a rimely basis.

Implementing action controls where one of these action-tracking qualities cannot beachieved will lead to some undesirable effects. These, too, are discussed in Chapter 5.

Like results control systems. action control systems usually cannot be made near-petfect, or at least they are prohibitively expensive to make near-perfect. As a consequence,organizations use personnel and cultural controls to help fill in some gaps. These controlsmotivate employees to control their own behaviors (by means of personnel c'ontrols) orto conh'ol each other's behaviors (by means of cultural controls).t6

PERSONNEL CONTROTS

Personnel controls build on ernployees' natural tendencies to control and/or motivatethemselves. Pet'sonnel controls serve any of three basic purposes. First, some of themclarify expectations. They help ensure that each employee understands what the organ-ization wants. Second, some of them help ensure that each employee is able to do a goodjob; that they have all the capabilities (e.g. experience, intelligence) and resources(e.g' information and time) needed to do a -eood

job. And third, some of the personnelcontrols increase the likelihood that each employee will engage in self-monitoring.Self-ntoniton ng is the naturally present force that pushes most employees to want to doa good job, to be naturally committed to the organization's goals. Self-monitoring iseffective both because most people have a conscience that leads them to do what is rightand are able to derive positive feelings of self-respect and self-satisfaction when theydo a good job and see their olganization succeed. The phenomena under.lying self-monitoring have been discussed in the management literature under a variety of labels,including self-control, intrinsic ntotit'ation, ethics ancl ntoraliA,, tntst, and loyalrt,.

Some organizations rely heavily on personnel controls. For example, JohnMcConnell, chairman of columbus, ohio-based worthington Industries, a superiorpeformel in the steel processing industry, said, "You have to trust the wolkforce. If youdon't, you've done a bad job."r7 Trust is a substitute for other, more fomal forms ofcontrol. worthington Industries does not have time clocks or plant supervlsors.

Three major methods of implementing personnel controls are through (1) selectionand placement of employees, (2) training, and (3) job design and provision of necessaryresources.

83

Chapter 3 .Action, Personnel, and Cultural Controls

Selection and Placement

Finding the right people to do a particular job and giving them both a good work

envlronment and the necessary resources can obviously increase the probability that

a job will be done properly. Organizations devote considerable time and effort toemployee selection and placement, and a large literature describes how these tasks

should best be accomplished. Much of this literature describes possible predictors ofsuccess, such as education, experience, past successes, and personality and social

skills.rREmployee selection often involves refelence checks on new employees,le which many

organizations haye stepped up in response to the heightened wonies over workplace

security.rO But beyond screening new employees to mitigate security issues, organ-

izations primarily focus on matching job requirements with job applicants' skills. For

example, Horne Depot has an in-house computer system that contains the names ofprescreened candidates who have the right skills and experience. This allows managers

to find qualified candidates quickly when the need arises. But the automated system also

provides cues about what interview questions to ask, what answers to listen for, and even

what advice to give the interviewees.rrMore exotic employee-seiection techniques have also been developed and used. Some

organizations have lesofted to analyzing potential employees' handwriting or using poly-

graph tests ro rry to weed out high-risk individuals. Dell Computer, General Electric,

Motorola, and other companies require job candidates to undergo lengthy interviews

with outside human resource service providers, or to take paper-and-pencil tests, or both.

While these evaluations are expensive, their cost is far less than the costs associated withhiring someone who is a "poor fit" with the company.rr

Training

Training is another common way to help ensure that employees do a good job. Training

can provide useful information about what actions or results are expected and how the

assigned tasks can best be pelformed. It can also have positive motivational effects

because employees can be given a greater sense of professionalism, and they are often

more interested in performing well in jobs they understand better.

Many organizations use formal training programs. such as in classroom settings, to

improve the skills of their personnel. The Los Angeles Unified School District warlted

to decentralize and give school principals more decision-making authority. Districtmanagers concluded, however, that the principals would not know how to use their

increased authority. They decided to put the principals through a formal mini-MBAprogram to teach them how to improve the educational process and manage school

costs. The principals attended classes over an 18-month period and several follow-upworkshops. The program was judged

-so successful that it was expanded to the San

Francisco Bay area and the East Coast."Much training takes place infotmally, such as through employee mentoring. Jerry

Reinsdorf, a successful entrepreneur and owner/chairman of the Chicago White Sox

basebail club, noted the importance of his role as a mentor:

My management style is to hire good people and develop a relationship with them sothat95Va

of the time they'll knor,v what decision I'd make and go ahead without askiug me.2l

His control system could be described as being dominated by selection and

training.

84

Cultural controls

Job design and provision of necessary resources

Another way to help ernployees act appropriately is siniply to make sure that the job isdesi-ened to allow motivated and qualified employees a high probability of success. Someolganizations do not give all their employees a chance to succeed. Some jobs are toocomplex. Salespeople may be assigned too many accounts to handle effectively.Ernployees also need a particular set of resources available to them in order to do a goodjob. Resource needs are highly job-specific, but they can include such iterns as informa-tion, equipment, supplies, staff suppolt, decision aids, or freedom from intenuption. Inlarger or-eanizations, pafiicularly, there is a strong need for tlansfer of information amongorganizational entities so that the coordination of lvell-timed, efficient actions and deci-sions is maintained.

CULTURAL CONTROTS

Culttrral cotttt'ols are designed to encourage mutual monitoring; a powerful form ofgroup pressure on individuals who deviate from gloup norms and values. Cultulal con-tlols are most effective where rnembers of a group have emotional ties to one another.Iu some collectivist cultules, such as Japan, incentives to avoid anything that woulddisglace oneself and one's famill' are paramount. Similarly, in many communities, suchas the Hasidic Jewish community in New Yolk City, and in many countries, notably thosein Southeast Asia, many business deals are sealed by verbal agl'eement only. The com-munities' social and moral pressures are stronger than legal contracts. But strong culturalcontrols produced by mutual-monitoring processes also exist within single organizations.

Cultures are built on shared traditions, norms, beliefs, values, ideologies, attitudes,and u,ays of behaving.25 Olganizational cultrlres remain relatively fixed over time, evenwhile goals and stlategies necessarily adapt to changing business conditions.26 The cul-tural nonns are embodied in written and unwritten rules that govenr employees' beha-viors. To understand an organizalion's culture, ask long-time employees questions like:What are you proud of around here? What does it take to get ahead? How do you stayout of trouble? If a strong organizational culture exists, the vast majority of long-timeemployees will have consistent answers to these questions even when the answers are notwritten down. When that is the case. strong, functional organizational cultures pronptemployees to wolk together in a synergistic fashion.

Managers attempt to create and shape organizational cultures in many ways, both inwords and by example. Codes of conduct and group rewards are among the most importantmethods of shaping culture, and thus effecting cultural controls. Other approaches includeintraorganizational tlansfers, physical and social anangements, and tone at the top.

Godes ol conduct

Most organizations above minimal size attempt to shape their organizational culturethrough what are knorvn, variously, as codes of conduct, codes of ethics, organizationalcredos, or statements of mission, vision, or management philosophy.rT These formal,written documents provide broad, general statements of organizational values, commit-ments to stakeholders, and the ways in which management would like the organizationto function. Each of these codes or statements is designed to help employees understandwhat behaviol's are expected even in the absence of a specific rule or principle. Thesestatements may include important messages about dedication to quality or customer

85

Chapter 3 . Action, Personnel, and Cultural Controls

satisfaction, fair treatment of staff and suppliers, employee safety, innovation, risk tak-ing, adherence to ethical principies, open communications, and willingness to change.For maximum effect. the messages included in these statements should be reinforcedthrou_eh formal training sessions, or at least through some discussions among employeesand their superiors.

The vadous codes and statements differ considerably in form. As an example, Figure 3.1

shows the code of conduct nsed at Provident Mutual, which includes a general policy andguidance on specific issues. Provident's general policy statement is aimed at influencingthe organization's culture. But the code _qoes on to provide behavioral guidance onspecific issues. The detailed behavioral prescriptions provide action accountabrlity con-trol because emplolrgs5 who violate these prescriptions will be reprimanded.

One survey of 264 companies (707c from the US, and the rest from Europe, Canada,and Mexico) found that 84Vo of the US respondents and 58Vo of the non-US respondentshave a code of conduct.rS The codes are drafted, most commonly, by top management,the corporate iegal department and, to a lesser extent, the board of directors. This surveyshowed that where codes exist, the vast majority of them define "fundamental guidin-qplinciples of the company." The only specific issties cited by 507o or mole of the

Frcunr 3.1 Code of Conduct of the Provident Mutual 0rganization

General PolicyThe Provident Mutual organization is committed to

achieving high standards of business and personalethics for itself and its personnel, Through performancein accordance with these standards, the Organizationand all its employees will merit and enjoy the respectof the public, the business community, policyholders,customers, and regulatory authorities

It is the personal responsibility of all employees toacquaint themselves with the legal and policy standardsand restrictions applicable to their assigned duties andresponsibilities, and to conduct themselves accordingly.Over and above the strictly legal aspects involved, all

company personnel are expected to observe high

standards of business and personal ethics in the

discharge of their assigned responsibilities,

Employee ConductEach member of the Organization must avoid any action,relationship or situation which could jeopardize or impairthe confidence or respect in which the Organization is

held by its customers and the general public, or whichappears to be contrary to the interests of ProvidentMutual or its policyholders.

Employees shall comply fully with all applicable statutesand regulations. Willful and knowing disregard ot thelaw may result in severe penalties to the Organizationln its many business activities, Provident Mutual and

its affiliated companies engage in vigorous, tair and

ethical competition. Discussions and agreements withcompetitors concerning pricing or other competitivepolicies and practices are strictly prohibited.

Conflict of InterestProvident Mutual annually circulates a policy statementof Conflicts of Interest. The basic policy states thatevery employee must avoid any interest that conflictswith the interests of Provident Mutual. The documentprovides detailed examples and explanations ofsituations and types of transactions which can giverise to conllicts of interest, In order to implement theconJlict of interest policy of Provident Mutual, all o{ficersand other a{fected persons are required to submitannually a completed disclosure statement to theChairman and Chief Executive Officer of ProvidentMutual. Each affiliated comoanv has a similarrequrrement,

Gifts to or by EmployeesEmployees may not give or receive anything of more thantoken value to or from any individual or organization withwhom Provident Mutual or its affiliates does business, orwho is seeking to do business with Provident Mutual orits affiliates. "Token" is defined as havinq a value of $50or less

Certain business courtesies, such as payment for a lunchor dinner in connection with a business meeting, normallywould not be a gift within the contexl of this policyHowever, such activity shall be limited in frequency.Employees shall endeavor to avoid any situation wherea gift or activity might appear to iniluence businessjudgment or relationships. Any question as to whethera gift might appear to be improper or questionable shallbe addressed in writing, with a statement of all relevantfacts, to the office of the General Counsel.

86

Cultural controls

Frcunr 3.1 continued

Po I itical Contri butionsNo funds or assets of the Company shall be used forfederal, state or local political campaign contributions.These prohibitions cover not only direct contributionsbut also indirect assistance or supDort of candidates orpolitical parties through purchase of tickets to specialdinners or other fund raising events or the furnishing ofany other goods, services or equipment to political partiesor committees,

No funds or assets of the Company shall be used directlyor indirectly for political contributions outside the UnitedStates, even where permitted by applicable law, withoutthe prior written approval of the Chief Executive Officer orGeneral Counsel,

The above prohibitions apply only to the direct or indirectuse of corporate funds or assets for polltical purposesand are, of course, not jntended to discourage employeesfrom making personal contributions to the candidates,parties or commiltees of their choice, through theCompany's Political Action Committee. Under nocircumstances shall employees be reimbursed inany way for personal contributions

Confidential lnlormation and lnsider TradingEmployees frequently have access to confidentialinformation concerning the Organization's business andthe businesses ol present and prospective customers,policyholders and other employees. Safeguardingconfidential information is essential to the conduct ofour business. Caution and discretion must be exercisedin the use of such information, which should be sharedonly with those who have a clear and legitimate need andright to know

No employee shall disclose confidential informationot any type, to anyone, except persons within theemployee's company who need to know. Informationregardrng a customer may not be released to thirdparties, government, or other organizations, withoutthe consent of the customer unless required by law.

Any requests for information arising through a legalprocess (e.9. subpoena or court order) must f irst bereferred to the office ol the General Counsel before therelease of information and belore the client is conlacted.

Selling or acquiring stocks, securities or otherinvestments, on the basis of non-public information isprohibited Securities include stocks. bonds, notes,debentures, or any other interests, instruments,documents or rights which represent securities.Questions concerning the definition of non-publicinformation or a security shall be referred to the officeoI the General Counsel before anv transaclions areunderlaken.

Service and Customer ConcernsThe foundation of the Organization is to providehigh quality service to our existing and prospectivecustomers Each company endeavors to giveprompt, courteous and accurate response to inquiriesand complaints received from customers. Whenappropriate ad.iustments are warranted, employeeswill make them promptly and courteously, Equallyimportant, we seek to add or improve policies,procedures and products that contribute to customersatisfaction.

lntegrlty of Records and Compliance withAccounting ProceduresAccuracy and reliability in the preparation of all businessrecords is mandated by law. lt is of critical importanceto the corporate decision-making process and to theproper discharge of Provident Mutual's financial, legaland reporting obligations. Ail business records, expenseaccounts, vouchers, bills, payroll and service records andother reports are to be prepared with care and honesty.False or misleading entries are not permilted in thebooks and records of Provident Mutual or any affiliatedcompany. All corporate funds and assets are to berecorded in accordance with applicable corporateprocedures. Compliance with accounting procedures andinternal control procedures is required at all times. lt isthe responsibility of all employees to ensure that both theletter and the spirit of corporate accounting and inlernalcontrol procedures are strictly adhered to at all timesThey should advise the responsible person in theirdepartment of any shorlcomings they observe in suchprocedures.

Administration of the CodeEmployees are encouraged to seek guidance regardingapplication or interpretation of this Code of Conduct andare expected to cooperate fully in any investigation of apotential violation. The statements set forth in this Codeof Conduct are intended as guidelines for employees.Routine questions of interpretation regarding the Codeshall be directed to the employee's supervisory otficer.and if necessary, relerred to the oflice of the GeneralCounsel. lf any employee belreves the code may havebeen violated, the matter shall prompily be reported tothe Director of Internal Audit. Violations of the Code ofConduct may be disciplined by the Organization, up toand including dismissal. However, the Code of Conductdoes not set forth all of the reasons or situations in whichemployees may be disciplined.

The Code of Conduct is not an employment contract,and the Organization may at any time modify theprovisions of this Code of Conduct as it deemsappropriate.

Source: K A l'.4erchant Moden Managemenl Controt Syslems: Texl and Cases lupper Saddle River, NJ: prenlice Hall 1 998), pp I 26-7

87

Chapter 3 Action, Personnel, and Cultural Controls

companies deal with purchasing gLridehnes (56%) and security of proprietar)' informa-

tion (53%). Other statenents commonly included relate to responsibilities regarding

the environment, marketing' product safetl" workplace safety, and conlidentiality ofemployee records. The surve), also found that the codes are dynamic documents: 597o

hacl been changed lvithin the three years prior to the survey. The most frequent reasons

fo1change were specific incidents either within the cornpany or the industry, new lead-

ership, new Ialvs, or a change in business strategy.

Most organizations flnd that the adoption of codes stimulates ciiscussion as to what

constitutes clesirable behavior and forces development of a cousensus. The adoption ofwritteu cocles also enhances comrrunication of expectations and the Ieasons for the

expectations. One study found that pressule to achieve perfotmance tal'gets was greatest

in companies witl-r forn-ral codes of conduct. But this perhaps only indicates that the per-

formance plessures create a need for the codes:re consistent with findin-ss that "pl'essure

to clo whatever it takes to meet business tar-qets" is the most common drivel of employee

miscondr-tct.3o

Do codes of conduct u,ork? The evidence is equivocal. One sulvey found that employ-

ees who work fol companies with codes of ethics were much more likely to rate the

commitment to ethical conduct by others in their'fitm as "about light." They were also

as much as 88% more iikely to rate their f,rm's fulfillment of its ethical obligations as

"exceptional."-t' However, a study that compared 202 firrns with codes of conduct with

104 firms without codes found that those with codes were just as Ikell' to be convicted

of illegal acts as those without thent.rr Moreover, as many as 527o of the 4,056 employ-ees lecently sun,eyed tnthe KPMG 200512006 Integt'ih' Srn't'e-l'believe that the fact that"their company's code of conduct is not taken seriously" was a root cause of ernployee

misconduct.rrSome codes of conduct indeed fail becar.rse they are not supported by stron-q leadership

and ploper torte Jront the top. Top managers do not always appear committed to then. or

worse, set bad examples themselves throngh inappropriate conduct, One stt"rdy founcl that

one fourth of the codes of conduct sttidied were clorntan l, meaning that employees per-

ceived the codes as simply public relations and not something to be taken seriously."Perhaps a case in point. Enron managers were proud of theil company's code of ethics,

but it failed to prevent the majot problems that led to the company's batrkluptcl'.r5

Group rewards

Providing rewarcls based on collective achievement also encoura-ses cultural control.Reward plans based on collective achievement come itt rnany forms. Common examples

are bonus, proflt-shaling, or gain-sharing plans that plovide compensation based on cor'-

porate or entity perfonnance in terms of accor,tnting returns, proflts, ot' cost reductions.

Encouraging broad employee ownelship of company stock, with effective colporatecommnnications to keep employees infonned and enthusiastic, encourages ali employ-

ees to think like owners. Research evidence shows that such plans work, seemingiy

because they create an ownelship culture.r6 A leview of 70 studies done over the past

25 years found that botl-r erlployee ownership and profit-sharing programs improvedemployee productivity, company performauce, ancl cotnpany sumivol'rates.rt

Gloup rervards are discussed here as a type of cultural control rather than as a tesults

conrrol (as r,ve discuss in Chapter 9) because they are quite different in chalacter fiomrervalcls given for iudividual accomplishment. With group l'ewards, the link between

individual efforts and the results being lewarded is weak, perhaps near zero for most

88

Cultural controls

-.oups other than small work teams. Thus, motivation to achieve the rewards is not.::rong the primary forces affected by group rewards; instead communication of-'\pectations and mutual monitoling (social control) at'e. Gloup rewards, however, can

ork, even in cor.rntlies iike the US with a culture oriented towards individualism and:rsonal accountability.

Evidence suggests that group rewards can have a positive effect on motivation and-:r'fonnance.r' Gloup rewards can encoulage teamwork, on-the-job training of new

crkers by rnore experienced ones, and tl-re creation of peer pressure on individualr:.rU)loyees to exelt themselves for the good of the group. Panhandle Eastem Corporation,,. natlu'al gas company, installed a _eain-sharing plan that called for all employees to-:ceive a borus if corporate earnings exceeded $2 per share. Tliis gloup-pelformance'.rur created a cost-cutting cultule throughout the organization and turned "employees:rnr top to bottorn . . . into cost-cutting vigilantes."3e

\4ichael Armstrong, then-chairman of Hughes Electronics, used group incentives to-:rrurge his cornpany's culture. Before Annstrong, Hughes'culture was a regimented,rp-heavy hier archy that:

. . . rnirrored its miiitary clients. Managers had iittle accountability. And the engineers'cultulereu,arded those who came up with the most sophisticated inventions - whether or not the rnar-ket rvanted them.ro

irr change the culture, Armstron-e instituted a new bonus program for all employees with':tvments based on the profits of their business unit, He required that engineers attend-rance classes, and he opened the company's books for all employees to see the results.i their efforts. Steven Dorfman, then-president of Hughes' Satellite unit said, "Now.:\.el'yone [is] walking the floors talkin-s about leturl on net assets."rr

Other evidence of the success of group rewards comes from the litet'ature that describes.ornpanies'experiences with plograms known as open book ntarngentent (OBM), of'. hich group rewards are an important ingredient. The goal of an OBM program is to cre-

-:te a cleal line of sight between each employee's actions and corporate financial perform-,'.trce and an incentive for the employees to behave in the corporation's best interest and-r'r lr?k€ useful suggestions for improvement. OBM programs involve: (1) regular shar-rrs of the con'rpany's financial information and any other information that will help the

;n.rployees work together with managenent towards organizational success; (2) training,:o that employees undelstand both what that information means and how they them-.elves can contribute to company performance; (3) rewards linked to company perform-.rnce; and (4) if necessary, a cultural change away fi'om a top-dorvn culture to ensure:hat employee ideas are both encouraged and considered fairly. Most commonly, OBM.ncentives involve tying a portion of each employee's compensation to key coryorateinaucial indicators, usually in the fom of an employee stock ownership plan (ESOP) oL

.r profit-shaling plan.The earliest program given the OBM label was that implemented at the Springfield

Remanulacturing Company in the early 1980s.4t This program was credited with tumingrround a near-failing company. The idea has spread and the business literature now con-iains a number of examples explaining how OBM programs have yielded significant:nrplovements in productivity and profits.'l

In conclusion, group rewards essentially delegate the monitoring of employees' activ-ities to employees' coworkers. This is the essence of mutual monitoring. Managers knowiheir ,eroup rewards are working u,hen they hear hard working employees urying on their

'luggish collea_eues ivith statements like, "You're hurting my profit sharing."

89

Chapter 3 . Action, Personnel, and Cultural Controls

Other approaches to shape 0rganizational culture

Other common approaches to shape organizational culture include intraorganizationaltransfers, physical and social arran-qements, and tone at the top.

Introorgarizational trctnsfers or entplo:'ee rotcttiott help transmit culture by improvingthe socialization of employees throughout the organization, giving thern a better appre-

ciation of the problems faced by different parts of the olganization, and inhibiting theformation of incompatible goals and perspectives. Transfels also potentially rnitigateemployee fraud by preventing employees frorr becoming "too" familiar with certainentities, activities, colleagues. and/or trausactions.rl

Pln,sical on'ongenletlts, such as offlce plans. architecture, and interiol decor, andsocial anangements, such as dress codes and vocabulary, can also help shape organiza-tional culture, Some organizations, such as technology fimrs in the Silicon Valley, havecreated informal cultures, with open office anangements and casual dress codes thatdeliver messages about the importance of innovation and employee equalitl,.

At Disneyland, ernployees are called cctst ntentbet's; being on the job is being onstage(off the job is offitage): a work shift is a petforntortce: and a job description is a scril:t.This vocabulary, which is imparted imrnediately on joining the company and is rein-forced through training, separates Disney employees from the rest of the world, bringsthem closer to-eether, and reminds them that they are pelfomers rvhose job is to helpfulfill the company's mission: to make people happy.

The largest Japanese flrms find it easier to uraintair.r a strong culture because they tendto retain tlrcir entplol'ees for long periods of time, until lecently often an entire lifetime.This stability in the employee base increases the hornogeneity of perspectives in theorganization. The errployees become socializecl to their organization's values and its"walr of doing things."

Finally, management can shape culture by setting the proper totrc ot the top. Theirstatements should be consistent with the type of culture they are trying to cleate and,importantly, theil behaviors should be consistent with their statenlents. Managers selveas role models and ale often cited as a determining factor in creating a culture of integrityin theil organizations.at Management cannot say one thing and do another.

Management sometimes sets the v'rcng tone by not responding appropliately to mat-ters brought to their attention, sucir as ethics concerns or reports of malpractice. All toocommon, wltistle blotvers (ernployees rvho draw attention to suspected malpractice) aleignored, such as Sherron Watkins was initially at Enron.*o To correct such a situation,Abbey National, a Blitish bank, set the right tone by producing a booklet about whistleblowin-e and by providing contacts inside and outside the firm fbr employees who aleconcemed about malpractice.rT In so doing. management set the tone that honesty andintegrity are valued and rer.varded by the organization. Several studies, hor.r,ever, painta rather gioomy pictule of tone at the top. Fol example, a study by the TreadwayCommission that exarnined acconnting scandals that brought dou,n companies tound thatfraud started at the too in7jc/c of the cases.r*

PERSONNEL/CULTURAL CONTROLS ANDTHE CONTROL PROBLEMS

As a group, the personnel/cultural controls are capable of addtessing all of the controlproblems although, as shown in Table 3.3, not each type of control is useful in address-

ing each type of problem. The lack-of-direction problem can be minimized, for example,

90

Effectiveness of personnel/cultural controls

TneLt 3.3 Control problems addressed by the various ways of effecting personnel andcultural controls

Lack ofdirection

MotivationalproDtems

Personallimitations

Ways of eflecling personnel controlsSelection and placementTrainingJob design and provision of necessary resources

Ways of elfecting cultural controlsCodes of conductGroup-based rewardsIntraorganizational transfersPhysical arrangementsTone at the top

X

X

X

X

X

X

X

Source:K A lvlerchant ModenlvlanagementControl Sysfems:TextandCases\tJpperSaddleRiver.NJ: prenticeHall. lgg8),p 130

by hirin-e expelienced personnei. by providing training programs, or by assigning newpersonnel to work groups that will provide good direction. The motivational problems,which ma1, be minirnal in organizations with strong cultures, can be minimized in otherorganizations by hiring highly motivated people or by assigning people to work groupsthat will tend to make them adjust to group nolms. Personal linitations can also bereduced through one or rnore types of personnel controls, parlicularly selection, training,and provision of necessarv resources.

EFFECTIVEN ESS O F PERS ON N Et/CU LTU RAt CO NTRO tS

Personnel/cuitural controls are adaptable. All organizations rely to some extent on theiremployees to guide and motivate themselves. Even in plisons whele administrators facegeneral inmate hostility and have few cor.rtlol options available other than physical con-straints, administratol's screen inmates so as to not assign dangerous inrnates to high-riskjobs, such as in a machine shop.

Some corporate control systems are dominated by personnel controls. William F.Cronk, president of Dreyer's Grand Ice Cream, said, "We consider hiring the mostimpoltant decision we can make. We hire the smartest. most inspired people we can find,give them the resources they need, then get out of their way."re Cultulal controls canalso, by themselves, dominate a control system. The best chance to create a strongculture, however, seems to be early in an or-eanization's life when a founder can imbuethe organization with a distinctive culture.50 But stron-q leaders and mana-eement policiesadded latel in an organization's history also can have an impact. Regardless of thedifficultS, in implementing them, cultural controls should sel've some positive purpose ine\/ery organization.

Cultural controls often have the advantage of being relatively unobtrusive. The limitsof acceptable behaviors may be prescribed in terms as simple as "the way we do thingsaround here." The people whose actions are being controlled may not even think of theshared norrns as being part of the MCS. As such, organizational cultules (shared values)can substitute for other fonnal tvpes of controls. Or. as Peters and Waterman observed:

q1

Chapter 3 Action, Personnel, and Cultural Controls

"the stronger the culture . . . the less need thele is for policy nanuals, olganizationcharts. or detailed procedures and rules,"sl

Personnel/cultural controls thus have several important advantages over results and

action controls. They are usable to some extent in almost every setting; their cost is often

lower than more obtn-rsive forms of controls; and they usually produce fewer hannfulside effects. Personnel/cultural controls perhaps even make good "economic sense" as

some recent evidence suggests that "itpa,l,s to be nice to employees."5r In an 800-store

study, Sears, the giant US retailer, found that if employee attitudes (such as about work-load and trearment by bosses) improve by 5Vo, customer satisfaction will jump by | .37o,

driving a half percentage point increase in revenues.''3 This lo-eic is echoed by Elizabeth

Rose, vice president at Northem Telecom of Toronto, Canada, tvho states that "they came

up with conclusive evidence that improving employee satisfaction will satisfl, customers

better and, in turn, improve financial results "5r

At the Raleigh, North Calolina-based SAS Institute, the world's largest privately held

software company, cofounder and chairman James H. Goodnight says that "he likes

happy people." He instills employee loyalty with an unusual anay of perks for his 2,700

headquarter employees, such as profit-sharing; a fi'ee health clinic; daycare centers;

private offlces for everyone; flexible 35-hour weeks; free sodas, ftesh fruit, and pastries

in the coffee-break rooms; and even a pianist in the subsidized lunch and rect'eation

room. SAS'5 tuntover late has been about 47o for years, conpaled to an industry

average of about 20Vo. Stanford University professol Jeffrey Pfeffel concluded, "The

roughly $50 million per year that SAS saves with its low turnover pays for all the

family-friendly stuff. And, while the free company clinic costs $1 million per year to

oDerate. that is $500,000 less than what it would cost the compan)/ if employees were

triated elsewhere."55But, the degree to which personnel/cultural controls are effective can var)/

significantly across individuals, groups, and societies. Some people are mol'e honest

than others, and some groups and societies have stronger emotional ties amon-9 their

members.

c0Nctusl0N

In this chapter we provided an overview of the most direct type of controls, ctction cotttrols,

which take any of several different forms: behavioral constraints. pleaction reviews,

action accountability, and redundancy. Action controis are the rnost dilect type of man-

agement control because they ensure the proper behaviors of the people on whom the

organization must rely by focusing directly on their actions.

We also described personnel and cultru'ctl corttrols, which managers implement to

encourage either or both of two positive forces that are normally pt'esent in organizations:

self- and mutual-monitoring. These forces can be encouraged in a number of ways.

including effective personnel selection and placement, training, job design atld provision

of necessary resources, cocles of conduct, group rewards, intra-organizational transfers,

physical and social arrangements, and tone at the top.

Personnel and cultural controls, which are sometimes referred to as sol controls. have

become more important in recent years. Organizations have become flatter and leaner.

Managels have wider spans of control and elaborate hierarchies and systems of action

controls (bureaucracies) have been dismantled and replaced with empowered employees'

In this environment, shared organizational values have become a lrlore important tool for

ensuring that everyone is acting in the organization's best interest.

92

i'lotes

' . KPMG 2003 Fraud Sin er' (KPMG LLP, 2003).L D. Ster,r,art and R. Chase, il[istake-Pt ooJtrtg: De sigttittg

Errors Out (Poltland. OR: Productivitl, Press. 1995).Pokct:oke is the Japanese ternt for .foolproof. It *,asintroduced to the managernent litelature b), the Japan-ese qua1it1, gur-u, Sigeo Shingo.

-r. S. Carey and D. Michaels. "At Sorne Air-lines, LaptopsReplace Pilots' 'Blain Bags','' The Woll Stt.eet Jout nal(Malch 26, 2002), p. B 1.

l, "Beal Stearns Ex-Staffer Pleads Guiltl, Ts 1.pitltfunds rn Check Scheme." Tlte lVall Street Jomnal,(Februaly 26,2002), p. Cl4.

5 ''Menrorable Memo: McDonald's Sends Operations toWar on Fries," I/re Woll Street .Iourncrl Intercrt-titeEdirion (December 18, 1997).

5 V. Govindalajan and J. G. San Miguel, "Sears,Roebuck aucl Co. (C): The Internal Audit Function,"Case uo. 9-179-125 (Boston: HBS Case Sen,ices. 1979)

I R. Henkoff, "Hou,to PIan for 1995," Forrttne(Decernber 31, 1990), p. 74.

3, S. Leung, "McDonald's Asks Myster.y Shoppers WhatAils Sales: Undercover Customers to Rate Service andFood Quality." The Wall Sil eet Journal (December. 17,2001), p. Bl

9 F. K. Iverson, "Effective Leader.ship: The Key isSimplicity," rn Y. K. Hsetty and V. M. Buehler (eds).Tlte Quest .for Cotnpetitir,ene,rs (Ner.v York: euorum,199 I ).

0 C. Daniels. "To Hire a Lumber Exper-t, Click Here,"Forturte (April 3, 2000), p. 268.

.1. A Galloni and M. R. Sesit, "Allfirst Officials RaisedConcern about Trader in a 1999 Merno," The WattStreer Jorrntctl (Feblualy 25 , 2002), p. C 1 6.

'1, A, Galloni and M. R. Sesir. "Controls at Allied L.ish'sAlllirst Likely Failed in Imporranr Wa1,s," The lVattStt'eet Journal (Febluarl, 20. 2002), p. C 1 .

i3. C. Gaspalino and S. Claig, "Lehman Broker in AllegedSwindle Also Supen,ised the Compliance Officer." I/reWall Sneet Jounlol (Februar;, 20,2002). p. Cl; andC. Gasparino. ''Rogue Broker. Costs Lehrnan. Coivena Bundle; Re,sulatol's Pt'ess Firns to Pay up to Sl00Million to Investors Wlio Were Gruttadauria Victinrs.''The V'all Srreet.lournal (December 4,2002), p. C1.

+. KPMG 2003 Fraud.Srrn.a.r, (KPMG LLP. 2003).15. See R. N. Hoit and R. E. Fincher, "The For-ei-sn Conupt

Plactices Act," F inancial Attalt'.sts Joru-rtal, 3j (March-April 1981): pp. 73-6: and L. Landr.o, .'Analysis ofITT's Report Shorvs Pr-oblems in Halting euestionableFoleign Payments." Tlrc Vlall Street Journal (June 3,1982), p. ?7.

16. MCSs dominated b1, personnel and cultural contr.olshave been called ot grutic (as opposed to nrcchanistic)b), BLrrns and Stalker': a professional bureatrcrac\. (asopposed to a rtmthirte bureattctctcti) by Mintzber.g;and belief s.t'.rrel/r.r (as opposed to dia,qnostit' c,otttrol

Notes

slsteirrs) by Simons. See T. Burns and G. M. Stalker.Tlte Matrtgentetrt of Intlot'aliorr (London: Tavistock.1961); H. Mintzbelg, The Structut'ing of Organi:atiorts(Engleu,ood Cliffs, NJ: Prentice-Hall, 1979); R. Srrrons,Lerers of Cottilol: Hou' Managets Use Intk-^,atieCortn'ol ,9-r,,rlclrs to Driye Strotegic Renev'al (Boston,MA: Harvard Business School Pless, 1995); andR Sirnons, Petfornrurtce Measurentertt and Ccntrol5-1,s/errs for Inrplententing Stt oteg\' (Upper SaddleRiver, NJ: Plentice Hall, 2000).

17. H. Rudnitskl,, "You Have to Tlust the Wolk Force,"Forbes (Itrly i9, 1993), p. 78.

18. See. fol example. B. D. Smart. How Leading Cont-panies IYin bt Hiriug, Cooching attd Keeping the BestPeople (Nett, York: Portfolio Haldcover, PenguinGroup USA,2005): and M. T. Brannick and E. L. Levine,Job At,zh'sis. Methods, Researclt, and Applications forHurnotr Resource Marrugentertt in the New Millettttiunt(Thousand Oaks. CA: Sage Publicarions, 2002).

19. KPX,IG 2003 Fraud Srrn,e,r. (KPMG LLP, 2003).20. See J. S. Lublin, "Check. Please." The Wall Sn-eer

.lour rtctl (March 11, 2002). p. R11; A Davis,"Employers Dig Deep Into Workers' Pasts, CitingTenolisrn Feals." Zte lI/all Street Jouructl (March 12,2002), p. A1.

21. C. Daniels. "To Hire a Lumber Expert, Click Here."Fortutte (April 3. 2000), pp. 261-10.

22. C. Daniels, "Does This Man Need a Shrink" Fortune(February 5. 2001). pp,205-1.

23. R. Wartzrnan. "School Inc.: Principals Taught to ActLike CEOs." The Wttll Street .lournal (Octobel 16,1996). p. CAl.

21. "Jeny Reinsdorf Pulls a Double Play in Chicago,"Busirtess WeeA' (October 10, 1983), p. 53.

25. An extensi\/e literatur-e exists on the benefits and meth-ods of shaping corporate cultules See R. M. Kilmann,Martctgitrg Betond the Quick Fn (San Francisco, CA:Jossey-Bass. 1989); R. H. Kilmann and M. J. Saxton(eds), Gaittirtg Control oJ the Corporcrte Culture(San Flancisco. CA: Jossey-Bass, 198-5); E. H. Schein,Orgctrti:atiortctl Cultute ancl Leoclersltip: A Dvranttcl/ieu' (Sar Francisco, CA: Jossey-Bass, 1985); andR. E. Walton, "Toward a Strategy of Eliciting Em-ployee Commitment Based on Policies of Mutuality,"in R. E. Walton and P. R. Lawlence (eds), HRM Tt endsatrd Cltrtllenges (Boston, MA: Harvar-d Business SchoolPless. i985). pp. 35-65.

26. J. C, Collins and J. I. Porras. Built to Last; successJulHabits of \tisionary Contpctnies (Nerv York: HarperBusiness. i994).

27. The Conference Boarrl, Corporate Ethics prartices(Nerv Yolk: The Cont-erence Board. Inc., 1992).

28. The Conference Board. Corporute Ethics Prauic.es(Nerv Yor'k: The Confelence Board. Inc., 1992). Theplevalence of codes of conduct reported in this study is

93

Chapter 3,Action, Personnel, and Cultural Controls

consistent with that reported in a difi-erent and rnorerecent sanrple of firnrs fi the KPlvlG 200512006 Irtreg-i'lt1' 9111'1,s,t (KPMG LLP. 2005).

29. A. I. Rich, C. S. Snrith and P. H. Mihalek. "AreCorporate Codes of Conduct Eft'ective?," MatngenretttAccourttirtg (1990), pp. 34-5.

30. KPMG 200512006 Integritt Srn'e,v (KPMG LLP.2005).

31. "Employees Say It's Hard to be Ethical in SomeOrganizations." Internal Aucliror (Febluary 1995), p. 9.

32. See R. Berenbeim. "An Outbleak of Ethics." Acr.oss tlrcBoard (May 1988), pp. 15-19.

33. KPMG 200512006 lrttegritt' .!rl.r'c.r' (KPMG LLP,2005).

34. S. Landekich, Corporare Codes of Cottclrtct: ArtExanitntiort atrcl Intplenrcntation Guide (Montvale, NJ:National Association of Accountants, 1989).

35, "Why Honesty Is the Best Policy: Corporate Deceit Is a

Slippery Slope." A Survey of Mana_eentent. The Eco-irorrrisl (Malch 9,2002), pp. 9-13.

36. C. Rosen, J. Case and M. Staubus, Equirt': lVln'Employee Ownership is Gootl Jbr Busittess (Boston.MA: Harvard Business School Press, 2005).

37. J. R. Blasi, D. Kruse and A. Bernstein, In tlte CotnpartyoJ Orners. The Trust ctbotrt Stock Optiotts (ctnd Wl:'l.

Even, Enrylotee Should Hat'e fftarrrJ (Neiv York: BasicBooks. 2003).

38. D. L. Kruse, "Plofit Shalin-e and Productivity:Microecononric Eviclence from the United States." ilre,Economic Jout ttctl (January 1992), pp.24-36.

39. E. Nelson, "Gas Company's Gain-Sharin-s Plan TurnsEmployees into Cost-Cutting Vigilantes," Tlte WallStreet Jourrrul (September 29, 1995), p. Bl

40. E. Schine, L. Armstrong and K. Kenvin, "Liftoff:Michael Amstrong Has Made Hughes an Electronicsand Telecom Contendet'," Business Week (Aprll 22,

1996), p. r42.41. E. Schine, L. Armstrong and K. Kerwin, "Liftoff:

Michael Armstrong Has Made Hughes an Electronrcsand Telecom Contender." Brtsirtess Week (Aprtl 22.

1996D, p. r42.42. J. Stack, The Great Game of Busirrcss (New York:

Doubleday, 1992).

43. See C. Lee, "Open Book Management," Trctitrirtg (Jttly

1994), pp, 66-80; J. Fierman. "Winning Ideas from

Mavelick lt{anagels." Fotttttte (FeblLrary 6, 1995),pp.2l-1; J. Case, Open-Book Mcutagettrent;Tlte Con-irtg Btrsittess Reyolution (Ne'uv Yolk: Haqper, 1995);A. Adelson. "Casual, Worker-Friendly, and a Money-maker. Too," Tlte Nev York Tintes (June 30, 1996),p. F8l and J. P. Schuster. J. Carpenter and M. P. Kane,Tlte Poter o.f' Open-Book Matragenent (New York:John Wiley. 1996).

44. KPMG I99B Frautl Srrr le,r (KPNIG LLP, 1999).45. KPMG 200512006 Itrtegrirr Srur,er, (KPMG LLP,

2005 ).

46. M, Swartz and S, Watkins. Potet Fctilure:Tlrc InsicleStort oJ the Collapse of Enron (Neu, York, Doubledal,.2004).

47 "Why Honestl, Is the Best Policl,: Corporare Deceit Isa Stippery Slope," A Sulvey of iVlana_sement, TheEcottontist (March 9, 2002), pp. 9-13.

48. H. W. Jenkins, "One More Dilty Job fol Accountants:Take the Blanre." Tlte \l'ull Street Jountal (March 20,2002).p A23.

49. Quoted in D. Fer'-euson. "Do Enrlepreneurial Com-panies Lose Their Innovative Spalk as Thev GrorvLar-eer?" Cttl Business (Fall 1995). p. 12.

-50. M. Trottnran, "New Atrnosphere: Inside SouthrvestAillines. Storied Cultule Feels Stlains; Spirit of Funand Hard Wolk is Clouded by Picketing and EmployeeComplaints; No Longer the Underdos," The Wctll Sn eetJourttal (July I 1, 2003), p. Al: E. H. Schein. "The Roleof the Founder in ttre Creation of Olganizational Cul-ture," Orgatti:otionel D\'nonrics. l2 (Summel 1983).pp. 13-28.

51. T. J. Peters and R. H. Waterman. Irt Seatclt ofE.rce Ilenc'e (Nerv York: Harper & Rorv, 1982). p. 75.

52. "Cornpanies Are Finding It Really Pa1,s ro Be Nice toErnployees," Tlte lYall Streer Jountql (July 22, 1998).

53. A. J. Rucci, S. P. Kirn and R. T. Quinn, "The

Enrployee-Custolrler-Pl'ofit Chain at Sears.'' HctnardBrrsirress Review (January-FebruaLy, 1998), pp. 82-97.

54 "Companies Are Finding It Really Pays to Be Niceto Errployees," The Wall Steet Jottrnal (Ittly 22,

l 998).55. T. D. Schellhardt. "An Idyllic Workplace under a

Tycoon's Thumb." Tlte Wall Street Journql (November23.1998).

94